Critical Readings on Global Slavery (4 vols.) 2017960524, 9789004346543, 9789004346567, 9789004346574, 9789004346581, 9789004346598, 9789004346611

Citation preview

Critical Readings on Global Slavery

Critical Readings on Global Slavery Edited by

Damian Alan Pargas Felicia Roşu

LEIDEN | BOSTON

The Library of Congress Cataloging-in-Publication Data is available online at http://catalog.loc.gov Library of Congress Control Number: 2017960524

Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface isbn 978-90-04-34654-3 (hardback, set) isbn 978-90-04-34656-7 (hardback, vol. 1) isbn 978-90-04-34657-4 (hardback, vol. 2) isbn 978-90-04-34658-1 (hardback, vol. 3) isbn 978-90-04-34659-8 (hardback, vol. 4) isbn 978-90-04-34661-1 (e-book) Copyright 2018 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijhoff, Brill Rodopi, Brill Sense and Hotei Publishing. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change. This book is printed on acid-free paper and produced in a sustainable manner.

Contents VOLUME 1 Introduction: Global Perspectives on Slavery 1 Damian Alan Pargas and Felicia Roşu

Definitions and Global Approaches 1 Definition and Distinction from Kindred Phenomena 13 H. J. Nieboer 2 The Causes of Slavery or Serfdom: A Hypothesis 43 Evsey D. Domar 3 The Emergence of a Slave Society 58 Moses I. Finley 4 Authority, Alienation, and Social Death 90 Orlando Patterson 5 The Anthropology of Slavery: The Womb of Iron and Gold 147 Claude Meillassoux 6 Slavery: A Question of Definition 184 Suzanne Miers 7 History as a Problem of Slaving 201 Joseph Miller 8 Historiography and Research Problems of Slavery and the Slave Trade in a Global-Historical Perspective 249 Michael Zeuske

vi

CONTENTS

Antiquity to the Early Modern Period 9

Between Slavery and Freedom 279 Moses I. Finley

10

A Scientific Approach to Ancient Slavery? 297 Niall McKeown

11

A Life-Course Approach to Household Slaves in the Late Third Millennium BC 313 Laura Culbertson

12

Slavery, Freedom and Citizenship in Classical Athens: Beyond a Legalistic Approach 334 Kostas Vlassopoulos

13

Justifications: Barbarians and Natural Slaves 359 N. R. E. Fisher

14

Novel Evidence for Roman Slavery 371 Keith Hopkins

15

Resisting Slavery 394 Keith Bradley

VOLUME 2 16

Body Work: Slavery and the Pauline Churches 427 Jennifer A. Glancy

17

How and Why Ancient Slavery Came to an End 477 Marc Bloch

18

Slavery in Early China: A Socio-Cultural Approach 504 Robin D. S. Yates

CONTENTS

19

Empire, Monotheism and Slavery in the Greater Mediterranean Region from Antiquity to the Early Modern Era 553 Jeffrey Fynn-Paul

20 Medieval Slavery in a New Geopolitical Space 588 Youval Rotman 21

Slavery in Late Medieval Europe 665 William D. Phillips

22

The Identity of the Slave in Scandinavia 699 Ruth Mazo Karras

23

Slavery and Cultural Antipathy 742 David Wyatt

24 An Explanation of Military Slavery 800 Daniel Pipes 25 War, Servitude, and the Imperial Household: A Study of Palace Women in the Chola Empire 841 Daud Ali

VOLUME 3 The Modern World: 1450–1900 26 The Geography of Slaving in the Early Modern Mediterranean, 1500–1800 863 Robert Davis 27

Piracy, Ransom Slavery and Trade: French Participation in the Liberation of Ottoman Slaves from Malta during the 1620s 880 Pál Fodor

vii

viii

CONTENTS

28 Shifting Patterns of Ottoman Enslavement in the Early Modern Period 895 Ehud R. Toledano 29 Ransom Slavery along the Ottoman–Hungarian Frontier in the Sixteenth and Seventeenth Centuries 915 Géza Pálffy 30 The Black Sea and the Slave Trade: The Role of Crimean Maritime Towns in the Trade in Slaves and Captives in the Fifteenth to Eighteenth Centuries 958 Mikhail B. Kizilov 31

The Gypsies in the Romanian Lands during the Middle Ages: Slavery 983 Viorel Achim

32 Serfs, Slaves, or Wage Earners? The Legal Status of Labour in Russia from a Comparative Perspective, from the Sixteenth to the Nineteenth Century 1044 Alessandro Stanziani 33 I Make Him My Dog/My Slave 1069 Brett Rushforth 34 The Process of Enslavement and the Slave Trade 1125 John K. Thornton 35 Europeans and the Rise and Fall of African Slavery in the Americas: An Interpretation 1155 David Eltis 36 The Living Dead Aboard the Slave Ship at Sea 1187 Stephanie Smallwood 37

From Creole to African: Atlantic Creoles and the Origins of AfricanAmerican Society in Mainland North America 1216 Ira Berlin

CONTENTS

ix

VOLUME 4 38 Task and Gang Systems: The Organization of Labor on New World Plantations 1263 Philip D. Morgan 39 Rethinking Palmares: Slave Resistance in Colonial Brazil 1294 Stuart B. Schwartz 40 The “Second Slavery”: Bonded Labor and the Transformation of the Nineteenth-Century World Economy 1326 Dale Tomich 41

Antislavery Debates: Tides of Historiography in Slavery and Antislavery 1350 Seymour Drescher

42 Indian Ocean Slaves in Cape Town, 1695–1807 1368 Nigel Worden 43 Slavery, Forced Labour and Resistance in Indian Ocean Africa and Asia 1397 Gwyn Campbell and Edward A. Alpers 44 Carrying Away the Unfortunate from India and Southeast Asia, 1500–1800 1419 Richard B. Allen 45 “Closed” and “Open” Slave Systems in Pre-Colonial Southeast Asia 1462 Anthony Reid 46 The Structure of Slavery in the Sulu Zone in the Late Eighteenth and Nineteenth Centuries 1486 James Francis Warren 47 The Rise and Fall of the Transpacific Slave Trade 1505 Tatiana Seijas

x

CONTENTS

Contemporary Slavery 48 Illicit Human Cargoes 1541 Eric Tagliacozzo 49 Slavery and the Slave Trade as International Issues, 1890–1939 1571 Suzanne Miers 50 Trafficked into Slavery 1595 Joel Quirk 51

Forced Marriage: Slavery Qua Enslavement and the Civil War in Sierra Leone 1627 Jean Allain

52 Slavery in its Contemporary Manifestations 1660 Kevin Bales Index 1687

Introduction: Global Perspectives on Slavery Damian Alan Pargas and Felicia Roşu Nobody knew slavery like an enslaved person, and few enslaved people experienced the complexities of human bondage in such diverse settings and from as many perspectives as Olaudah Equiano. Born an Igbo prince in the Kingdom of Benin’s Essaka district in 1745 (in what is now eastern Nigeria), Equiano grew up in a society that accustomed him to slavery from early childhood. His father had “many slaves,” and in his region enslavement through warfare, kidnapping, and as punishment for the severest of crimes was common. Considered for all intents and purposes property, enslaved people in his village could be bought, sold, and bequeathed, and many slaves entered elite households as part of a woman’s dowry, along with cattle and other household goods. The practice of slavery was so ubiquitous that, as Equiano later recalled in his now classic 1789 memoir, “some of these slaves have even slaves under them, as their own property, and for their own use.”1 Until the age of eleven, Equiano knew West African slavery exclusively from the perspective of his father’s household—i.e., an elite slaveholder’s household—but that changed one day when he and his sister were kidnapped in a raid and reduced to bondage themselves. Marched on a several-day journey from his home community, Equiano now discovered first-hand what it really meant to be enslaved. Permanently severed from kin and forcibly moved as a tradable commodity, his experiences echoed those of enslaved people in most world societies: “outsiders, rootless and ahistorical individuals who were ultimately held against their will by the threat of force,” in the words of Herbert Klein.2 Equiano was sold to a chieftain in a “pleasant country,” and although he claimed that his master used him “extremely well,” he also feared corporal punishments and concocted desperate schemes to escape and return to his family—a dream that was only definitively dashed when he was sold yet again “and carried through a number of places,” steadily becoming further and further removed from his place of birth. Some six or seven months after

1  Olaudah Equiano, The Life of Olaudah Equiano, or Gustavus Vassa, the African (1789; Mineola, NY: Dover Publiations, 1999), 11–23, quotes on pages 23 and 17 resp. For more on Equiano’s life, see for example James Walvin, An African’s Life: The Life and Times of Olaudah Equiano, 1745–1797 (London: Cassell, 1998). 2  Herbert S. Klein, The Atlantic Slave Trade (New York: Cambridge University Press, 1999), 1.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_002

2

Pargas and Roşu

having been kidnapped, Equiano finally “arrived at the sea coast,” where he was crammed onto a transatlantic slave ship and subsequently endured all the horrors of the middle passage. Confusion, despair, disease, the “groans of the dying,” and even the suicides of fellow captives formed the grim backdrop to his experiences during the tortuous voyage.3 Equiano had entered the world of Atlantic slavery. Although he had already experienced slavery in the hinterlands of the Niger delta, he consistently expressed shock at the slave system in which he now found himself. In West Africa he had known slavery to be a universal yet relatively small-scale practice, confined mostly to elite households. He had known enslaved people to be treated harshly as non-persons, but he had also known them to be treated well and over time become integrated into their masters’ extended families, even whilst remaining outsiders in a theoretical sense (revealing an “institutionalization of marginality” and “rehumanization” process that was common in African cultures of slavery, according to Suzanne Miers and Igor Kopytoff).4 Nothing prepared Equiano for the brutal and inhumane treatment he witnessed on the plantations of the New World. Indeed, he described slavery in the Americas as “nothing but misery, stripes, and chains.” Its strict racial hierarchies as well as the infrequency with which enslaved people exited bondage baffled him. Moreover, the extensive geographic scope of the Atlantic system—which connected three continents and systematically transported millions of enslaved people across the ocean—initially surpassed the young boy’s comprehension.5 Arriving weakened and “exceedingly miserable” in the Americas, Equiano passed through the West Indies and briefly resided on a Virginia plantation before being sold to the captain of a British merchant ship. From then on he found himself almost exclusively in the service of captains of British navy vessels and West Indian slave ships, during which he sometimes had to assist in the process of transporting newly arrived African slaves between the insatiable labor markets of the Caribbean islands. After years of forced exile a process of acculturation took place, and Equiano found himself identifying with both his native land and with the land of his enslavers. Criss-crossing the ocean several 3  Equiano, Life of Olaudah Equiano, 24–36, quotes on pp. 25, 30, and 33, resp. See also Stephanie Smallwood, Saltwater Slavery: A Middle Passage from Africa to American Diaspora (Cambridge, Mass.: Harvard University Press, 2007), 122–152. 4  Suzanne Miers and Igor Kopytoff, “African ‘Slavery’ as an Institution of Marginality,” in ibidem, eds., Slavery in Africa: Historical and Anthropological Perspectives (Madison: University of Wisconsin Press, 1977), 3–78. 5  For his opinions of New World slavery, see for example ibid., 69–81, 69 (quote).

Introduction: Global Perspectives On Slavery

3

times, the enslaved Igbo prince gradually came to adopt elements of his masters’ culture, learning English and ultimately converting to Christianity. He also never lost hope of becoming free one day, and in 1766, when he was in his early twenties, he was permitted to purchase his own freedom with earnings from extra economic activities. This made him extraordinarily lucky compared to the vast majority of his fellow captives, and he knew it. As a free man Equiano continued to live a seafaring life—on one occasion he even assisted an English acquaintance in the purchase and transportation of slaves from Jamaica to the Musquito Coast—but he ultimately settled in England and became a prominent abolitionist, playing a major role in the British abolition of the Atlantic slave trade.6 In the course of his lifetime, Olaudah Equiano developed a broad and multifaceted understanding of slavery. He had been born into a slaveholding culture and become an ardent antislavery activist; he had been the free son of an elite slaveholder and a slave himself; he had been enslaved by Africans in Africa and by Europeans in the Americas; he had traveled the seas chained as cargo in the filthy hold of a slave ship and later again as an assistant to European enslavers; he had been, in the words of Orlando Patterson, a “natally alienated outsider” who later adopted the customs and culture of his enslavers.7 In short, Equiano experienced slavery as both a global and a globalizing phenomenon—global in the sense that it existed in diverse settings around the world, from Essaka to the Caribbean, and globalizing in the sense that it connected world societies, from England to Benin to Virginia. Much like Equiano’s story, this anthology promotes the examination of slavery from global and globalizing perspectives. Building upon a recent surge in slavery research, it encourages students and scholars to view slave systems across time and space as both ubiquitous and interconnected. The study of slavery has grown strongly in the last couple of decades, as scholars working in several disciplines have actively cultivated broader perspectives on enslavement. Two new developments can be identified: one is the intensification of research by scholars specialized in disciplines and regions previously underrepresented in the study of slavery; the other is an increased willingness by scholars to zoom out and consider the existence and interrelatedness of slavery practices outside of their particular areas of focus. Not only 6  For Equiano’s arrival in the Americas, see p. 37 (quote). Under one captain Equiano frequently had “different cargoes of new negroes in my care for sale.” See p. 74 (quote). For his assistance of an English acquaintance in purchasing slaves for a plantation, see p. 157. 7  Orlando Patterson, Slavery and Social Death: A Comparative Study (Cambridge, Mass.: Harvard University Press, 1982), 6.

4

Pargas and Roşu

has interest in slavery and its aftermath reached a high point among historians, anthropologists, and area studies scholars working on the Atlantic world, but scholars of ancient, medieval, North and sub-Saharan African, Mediterranean, Near Eastern, and Asian and Pacific societies have intensified the study of slavery in their respective regions as well. Practices of modern slavery and human trafficking from South Asia to Europe and beyond are likewise receiving more academic attention than ever before. At the same time, scholars specialized in particular slave systems are increasingly looking across borders—temporal, spatial, and disciplinary—to more fully grasp how widespread enslavement has been in world history and how it has connected societies through warfare, trade and power relations, and cultural exchange. While comparative approaches to slavery—especially as a means to better understand the Atlantic experience—date at least as far back as Frank Tannenbaum’s classic work Slave and Citizen (1946), new approaches have done much in the past twenty years to provide wider views of enslavement and its legacies, following in the wake of recent trends in the study of global history as a field of inquiry.8 These developments have had a major impact on the way scholars approach slavery, as can be seen in a wide variety of academic activities. For example, the recent publication of the first two volumes of The Cambridge World History of Slavery—which will ultimately consist of four volumes of original contributions from prominent scholars in the field—represents an important step forward in positioning slavery within a framework that transcends time and space. Similarly, the classic journal Slavery & Abolition has now been complemented by the Journal of Global Slavery (launched in 2016), a platform for studies that approach slavery specifically from transnational, transregional, and global perspectives. Theoretical studies from the past few years, such as Joseph Miller’s The Problem of Slavery as History (2009), have grappled with the consequences of global approaches to the study of enslavement, while a slew of excellent new monographs have situated regional slave systems within a more global context. Matthew Hopper’s Slaves of One Master: Globalization and Slavery in Arabia in the Age of Empire (2015), to name but one recent example, explores the African diaspora and slavery in Arabia in the nineteenth and early twentieth centuries, specifically underscoring the interconnectedness of enslavement, globalization, and empire. Beyond the realm of publications, the academic community has also been enriched by recent international 8  Frank Tannenbaum, Slave and Citizen: The Negro in the Americas (New York: Alfred Knopf, 1946); Frederick Cooper, Thomas C. Holt, and Rebecca J. Scott, Beyond Slavery: Explorations of Race, Labor, and Citizenship in Postemancipation Societies (Chapel Hill: University of North Carolina Press, 2000), 1–2.

Introduction: Global Perspectives On Slavery

5

conferences that have emphasized global perspectives on slavery, as well as the founding of new research centers—such as Brown University’s Center for the Study of Slavery and Justice—dedicated to stimulating the study of slavery in its historical and contemporary manifestations, and from interdisciplinary perspectives.9 Clearly the study of slavery has entered a new phase. Never before have so many scholars been so keen to view slavery through such a wide lens; never before have scholars working on different regions and periods so actively engaged in collaboration and cross-pollination of ideas. The fruits of such open contact are bearing out in the literature. Broader perspectives have led many scholars to rethink their approaches to the study of human bondage, including the very theories, concepts, and language they use in their analyses. Some, for example, have begun advocating for the abandonment of the more “traditional” concepts and ideas that had long been employed by historians of ancient and American societies, while others, conversely, have called for a broader application of concepts applied to the study of slavery in these societies. The charge that many of the models and theories that have been applied to the study of slavery in the ancient world and western hemisphere are inadequate for understanding slavery in other contexts is especially strong. This is not a new development; “all-encompassing” theories regarding slavery have indeed always been subjected to intense scrutiny. One of the oldest explanatory models for the development of slave systems, the Nieboer-Domar hypothesis—which posits that slavery emerges as a result of economic incentives related to situations where land is plentiful and labor in short supply (the so-called “land-labor ratio”)—has been criticized by many scholars during the past half-century for being too simplistic and inapplicable to contexts in which slavery did not revolve around economic gains or the control of agricultural 9  Several works dedicated to a global view of slavery, including monographs and anthologies, attest to this development. See for example: David Turley, Slavery (Oxford: Blackwell, 2000); Keith Bradley and Paul Cartledge, eds., The Cambridge World History of Slavery 1: The Ancient Mediterranean World (Cambridge: Cambridge University Press, 2011); David Eltis and Stanley L. Engerman, eds., The Cambridge World History of Slavery 3: AD 1420–AD 1804 (Cambridge: Cambridge University Press, 2011); Gad Heuman and Trevor Burnard, eds., The Routledge History of Slavery (New York: Routledge, 2012); Joseph C. Miller, The Problem of Slavery as History: A Global Approach (New Haven: Yale University Press, 2009); and Michael Zeuske, Handbuch Geschichte der Sklaverei: eine Globalgeschichte von den Anfängen bis zur Gegenwart (Berlin and Boston: De Gruyter, 2013). For a global view of modern slavery, see Kevin Bales, Understanding Global Slavery (Berkeley: University of California Press, 2005; Matthew S. Hopper, Slaves of One Master: Globalization and Slavery in Arabia in the Age of Empire (New Haven: Yale University Press, 2015).

6

Pargas and Roşu

labor, such as the military or administrative slaves of the Ottoman empire.10 While the Nieboer-Domar hypothesis continues to be challenged and defended by scholars, other overarching models have more recently come under attack as well. The concept of “social death” that was developed by Orlando Patterson—himself an early proponent of comparative slavery studies and an open critic of the Nieboer-Domar hypothesis—to explain the nature of slavery as a relationship based on the domination, natal alienation, and social debasement of the enslaved, for example, has been criticized by scholars of African slavery such as Frederick Cooper for ignoring slave agency and focusing too much on enslavers’ ideas about how they thought slavery should work, rather than how the system actually played out in specific settings.11 Similarly, Kostas Vlassopoulos recently questioned the usefulness of “static” and “ahistorical” typologies of slavery such as Moses Finley’s “slave society/society with slaves” paradigm—often applied to studies of slavery in ancient Rome and the Americas—in favor of viewing slavery as a temporally and spatially changing process, as proposed by global slavery historians such as Joseph Miller.12 In a similar vein, scholars of slavery in non-Atlantic societies have strongly cautioned the academic community against viewing the Atlantic experience of slavery as “typical” in world history. In a recent publication, for example, Christine Sears, a historian of slavery in Ottoman Algiers, called upon scholars to rethink the idea that urban slavery was merely an aberration in global history and to abandon the Atlantic-centric notion that forms of plantation slavery were more “normal.” Sears invites scholars to “dislodge[e] antebellum southern slavery from its pedestal as the quintessential slavery” and explore practices of urban slavery from more comparative and global perspectives. 10  Evsey D. Domar, “The Causes of Slavery or Serfdom: A Hypothesis,” Journal of Economic History 30 (Mar. 1970): 18–32. Orlando Patterson delivered one of the most scathing critiques of the Nieboer-Domar hypothesis, arguing that rather than viewing slavery as an economic system, it should be seen in terms of power relationships. See for example: Orlando Patterson, “The Structural Origins of Slavery: A Critique of the Nieboer-Domar Hypothesis from a Comparative Perspective,” in Annals of the New York Academy of Sciences 292 (1977): 12–34; Patterson, Slavery and Social Death, 1. 11  Patterson, Slavery and Social Death; Frederick Cooper, Colonialism in Question: Theory, Knowledge, History (Berkeley: Univeristy of California Press, 2005), 17. 12  Kostas Vlassopoulos, “Does Slavery Have a History? The Consequences of a Global Approach,” Journal of Global Slavery 1, no. 1 (Apr. 2016): 5–27; Moses I. Finley, Ancient Slavery and Modern Ideology (Princeton: Marcus Weiner, 1988); Miller, The Problem of Slavery, 1–35. For an example of the application of Finley’s concept to the study of the Americas, see for example Ira Berlin, Generations of Captivity: A History of African American Slaves (Cambridge, Mass.: Harvard University Press, 2003).

Introduction: Global Perspectives On Slavery

7

Gwyn Campbell has also warned scholars against viewing Atlantic slavery as a model for understanding various forms of unfreedom in the Indian Ocean world, a sentiment echoed by many scholars who work on slavery in Asia and the Pacific.13 While certain theoretical models and approaches are being challenged, revised, and even discarded in the wake of broader understandings of slavery in various world societies, some scholars have called for the broader application of certain concepts in order to underscore universal patterns of enslavement or the interconnectedness of slave systems across time and space. For example, Ehud Toledano has recently engaged with Finley’s above-mentioned paradigm to argue that enslavement practices in the Ottoman and Islamic world more closely resembled “societies with slaves” than “slave societies.” Another example is Dale Tomich’s concept of “the second slavery”—inspired by historical studies of “the second serfdom” in Eastern Europe and adapted by Tomich to conceptualize the resurgence of slavery in the Atlantic world in the nineteenth century—which has recently been used by Paul Lovejoy to explain the related resurgence of slavery in West Africa at roughly the same time.14

13  Christine E. Sears, “‘In Algiers, the City of Bondage’: Urban Slavery in Comparative Context,” in Jeff Forret and Christine E. Sears, eds., New Directions in Slavery Studies: Commodification, Community, and Comparison (Baton Rouge: Louisiana State University Press, 2015), 202–203; Gwyn Campbell, “Slavery in the Indian Ocean World,” in Gad Heuman and Trevor Burnard, eds., The Routledge History of Slavery (London: Routledge, 2012), 52. 14  See Ehud R. Toledano, “Enslavement and Freedom in Transition: MENA Slave Societies from Empires to National States,” Journal of Global Slavery 2, no. 1 (May 2017): forthcoming; Ehud R. Toledano, “Ottoman and Islamic Societies: Were They ‘Slave Societies?’” in Catherine M. Cameron and Noel Lenski, eds., What is a Slave Society? The Practice of Slavery in Global Perspective (Cambridge: Cambridge University Press, forthcoming). For literature on the second slavery, see: Dale Tomich, Through the Prism of Slavery: Labor, Capital and World Economy (Lanham: Rowman & Littlefield, 2004); Anthony E. Kaye, “The Second Slavery: Modernity in the Nineteenth-Century South and the Atlantic World,” Journal of Southern History 75, no. 3 (Aug. 2009): 627. Also see Tomich, “The ‘Second Slavery’: Bonded Labor and the Transformations of the Nineteenth-Century World Economy,” in Francisco O. Ramírez (ed.), Rethinking the Nineteenth Century: Contradictions and Movement (New York: Praeger, 1988), 103–17; Javier Laviña and Michael Zeuske, eds., The Second Slavery: Mass Slaveries and Modernity in the Americas and in the Atlantic Basin (Berlin: Lit Verlag, 2014). For Lovejoy’s application of the concept to West Africa in the age of jihad, see Paul E. Lovejoy, “Jihad and the Era of the Second Slavery,” Journal of Global Slavery 1, no. 1 (Apr. 2016): 28–43; Paul E. Lovejoy, Jihad in West Africa during the Age of Revolutions (Athens, OH: Ohio University Press, 2016).

8

Pargas and Roşu

A consensus regarding the application of such concepts is unlikely to be reached anytime soon, but it is beyond dispute that our understanding of slavery has benefited immensely from scholars’ increased willingness to cross borders and take into account the global and globalizing nature of enslavement. The present anthology encourages students and scholars of slavery to do just that, and it aims to provide a point of reference for better understanding enslavement across time and space. It contains previously published works by some of the most preeminent scholars of slavery in various regions and time periods, with the intention of introducing those who are interested in the study of human bondage to some of the most important and widely cited research from both past and present. Critical Readings on Global Slavery consists of fifty-eight chapters and is divided into four thematic parts. The first part introduces the reader to academic discussions regarding definitions, concepts, models, and global approaches to slavery, an essential starting point for understanding some of the most important debates in the field. The second part contains eighteen chapters that examine slavery from antiquity to the early modern period—from ancient Greece to early China and from medieval Scandinavia to the Chola dynasty of southern India. With twenty-seven chapters, the third part is the largest in this anthology, and it covers the modern world from 1450 to 1900. Employing a broad scope, this part contains works that explore slavery in a wide variety of contexts, including the Mediterranean, the Ottoman empire, the Black Sea region, Russia, various parts of Africa, the Atlantic world, the Indian Ocean, Southeast Asia, and the transpacific slave trade from the Philippines to Mexico. The final part, containing only five chapters, provides a link between historical slavery studies and contemporary research on modern slavery and human trafficking. It should be noted that this anthology does not pretend to be in any way exhaustive of the truly vast amount of research that has been conducted in the field. Editorial constraints have required us to make decisions based on a number of criteria. First, only works available in English have been included here. In general this means only works that were originally written in English, although we have included a few contributions that were translated from their original language (such as Claude Meillassoux’s classic book The Anthropology of Slavery, which was translated from French in 1991). Second, we were limited in the number of chapters we could include as well as the number of pages per chapter. We have done our best to include as many important articles and book excerpts as possible, but in the end this anthology can only provide a sampling of the work that has been done on slavery. Third, while we have endeavored to provide a balance between diverse spatial and temporal settings, we were

Introduction: Global Perspectives On Slavery

9

unable to provide equal numbers of chapters on each region and time period. Instead, this anthology broadly reflects the uneven popularity of slavery research in certain areas over others. Part 3, which deals with the modern world, is the most voluminous part because this era has been privileged in the scholarship so far. We have nevertheless striven to avoid the often extreme overemphasis on the Atlantic world that has dominated the literature for generations, and to include contributions from as wide a variety of contexts as possible. We encourage students and scholars to employ this anthology as a first step to better understanding slavery from a global perspective. It is our sincere hope that its contents will inspire researchers to delve further into this fascinating field and develop new lines of inquiry for the next generation.

Definitions and Global Approaches

∵

Definition and Distinction from Kindred Phenomena H. J. Nieboer 1

Ordinary Meaning of the Term “Slavery”

In most branches of knowledge the phenomena the man of science has to deal with have their technical names; and, when using a scientific term, he need not have regard to the meaning this term conveys in ordinary language; he knows he will not be misunderstood by his fellow-scientists. For instance, the Germans call a whale Wallfisch, and the English speak of shell-fish; but a zoölogist, using the word fish, need not fear that any competent person will think he means whales or shell-fish. In ethnology the state of things is quite different. There are a few scientific names bearing a definite meaning, such as the terms “animism” and “survival”, happily introduced by Prof. Tylor. But most phenomena belonging to our science have not yet been investigated; so it is no wonder, that different writers (sometimes even the same writer on different pages) give different names to the same phenomenon, whereas on the other hand sometimes the same term (e.g. matriarchate) is applied to widely different phenomena. As for the subject we are about to treat of, we shall presently see that several writers have given a definition of slavery; but no one has taken the trouble to inquire whether his definition can be of any practical use in social science. Therefore we shall try to give a good definition and justify it. But we may not content ourselves with this; we must also pay attention to the meaning of the term “slavery” as commonly employed. There are two reasons for this. First, we must always rely upon the statements of ethnographers. If an ethnographer states that some savage tribe carries on slavery, without defining in what this “slavery” consists, we have to ask: What may our informant have meant? And as he is likely to have used the word in the sense generally attached to it, we have to inquire: What is the ordinary meaning of the term “slavery”?

S ource: Nieboer, H. J., “Definition and Distinction from Kindred Phenomena,” in H. J. Nieboer, Slavery as an Industrial System: Ethnological Researches, The Hague: Martinus Nijhoff, 1900, pp. 1–39.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_003

14

Nieboer

The second reason is this. Several theoretical writers speak of slavery, without defining what they mean by it; and we cannot avail ourselves of their remarks without knowing what meaning they attach to this term. And as they too may be supposed to have used it in the sense in which it is generally used, we have again to inquire: What is the meaning of the term “slavery” in ordinary language? The general use of the word, as is so often the case, is rather inaccurate. “Careless or rhetorical writers” says Ingram, “use the words “slave” and “slavery” in a very lax way. Thus, when protesting against the so-called “Subjection of Women”, they absurdly apply those terms to the condition of the wife in the modern society of the west—designations which are inappropriate even in the case of the inmates of Indian zenanas; and they speak of the modern worker as a “wage-slave”, even though he is backed by a powerful trade-union. Passion has a language of its own, and poets and orators must doubtless be permitted to denote by the word “slavery” the position of subjects of a state who labor under civil disabilities, or are excluded from the exercise of political power; but in sociological study things ought to have their right names, and those names should, as far as possible, be uniformly employed”.1 But this use of the word we may safely regard as a metaphor;2 nobody will assert that these labourers and women are really slaves. Whoever uses the term slavery in its ordinary sense attaches a fairly distinct idea to it. What is this idea? We can express it most generally thus: a slave is one who is not free. There are never slaves without there being freemen too; and nobody can be at the same time a slave and a freeman. We must, however, be careful to remember that, man being a “social animal”, no man is literally free; all members of a community are restricted in their behaviour towards each other by social rules and customs.3 But freemen at any rate are relatively free; so a slave must be one who does not share in the common amount of liberty, compatible with the social connection. The condition of the slave as opposed to that of the freeman presents itself to us under the three following aspects. First, every slave has his master to whom he is subjected. And this subjection is of a peculiar kind. Unlike the authority one freeman sometimes has over 1  Ingram, p. 261. 2  In the second Chapter and in the continuation of this we shall meet with more instances of this metaphoric (sometimes rather dangerous) use of the term “slavery”. 3  Bastian, Rechtsverhältnisse, p. 14.

Definition And Distinction From Kindred Phenomena

15

another, the master’s power over his slave is unlimited, at least in principle; any restriction put upon the master’s free exercise of his power is a mitigation of slavery, not belonging to its nature, just as in Roman law the proprietor may do with his property whatever he is not by special laws forbidden to do. The relation between master and slave is therefore properly expressed by the slave being called the master’s “possession” or “property”, expressions we frequently meet with. Secondly, slaves are in a lower condition as compared with freemen. The slave has no political rights; he does not choose his government, he does not attend the public councils. Socially he is despised. In the third place, we always connect with slavery the idea of compulsory labour. The slave is compelled to work; the free labourer may leave off working if he likes, be it at the cost of starving. All compulsory labour, however, is not slave labour; the latter requires that peculiar kind of compulsion, that is expressed by the word “possession” or “property” as has been said before. Recapitulating, we may define a slave in the ordinary sense of the word as a man who is the property of another, politically and socially at a lower level than the mass of the people, and performing compulsory labour. We shall inquire next, whether this notion is a practical one for the purpose of our investigation, or whether it requires any improvement. But it may be convenient first to examine, what our theoretical authors have to say on the subject. 2

Use of the Term “Slavery” in Theoretical Literature

Several authors we have consulted give no definition at all. We mention only two of them: “Bastian, Die Rechtsverhältnisse bei verschiedenen Völkern der Erde”, and: “Giddings, The principles of sociology”. Books with such titles might be expected to contain something on the subject.4 Spencer remarks: “[The captives] fall into unqualified servitude.... They belong absolutely to their captors.... They become property, of which any use whatever may be made”.5 Although this may not properly be called a definition of slavery, it appears that he uses “becoming property” and “falling into unqualified servitude” (or slavery) as synonymous expressions. 4  Griddings’s book, valuable as it is in many respects, is, as a system of sociology, far from complete. 5  Spencer, Pol. Inst., p, 291.

16

Nieboer

According to Ingram “the essential character of slavery may be regarded as lying in the fact that the master was owner of the person of the slave”.6 Lippert remarks: “The fact, that one man becomes an object of possession by another, characterizes the nature of slavery”.7 Sohm calls a slave “a man who is not regarded as a person, but as a thing. The slave is left to the discretion of the master, who has over him the right of property”.8 Letourneau says: “The rights of the masters over their slaves were always excessive; they were those of a proprietor over his possession”.9 Jhering also remarks that “the master’s potestas may be called property”.10 In the first paragraph three principal features of slavery have been enumerated. We see that our theorists attach most importance to the first feature: “property” or “possession”.11 Whether we can agree with them will be shown in the next paragraph. 3

Definition for Scientific Use

The present investigation is a sociological one; therefore our definition of slavery has to be sociologically relevant. We have to ask: What is the social value of slavery? Slavery is an organ in the social body performing a certain function, and we have to inquire: How is this organ developed, and how, in the various stages of its development, does it perform its function? But then we must know first what this organ and its function are. Thus only can we exclude from our inquiry organs somewhat resembling slavery, but functionally quite different from it, and organs wholly different from slavery, but performing the same function or nearly the same. And this is necessary; for the inclusion of such organs would create a confusion fatal to a right understanding. What then is slavery and what is its function?

6  Ingram, p. 262. 7  Lippert, II p. 534. 8  Sohm, p. 106. 9  Letourneau, p. 492. 10  Jhering, II p. 167. 11  This view is also held by Wagner and Puchta, whose ample expositions we shall make use of in the next paragraph.

Definition And Distinction From Kindred Phenomena

17

The great function of slavery can be no other than a division of labour.12 Division of labour is taken here in the widest sense, as including not only a qualitative division, by which one man does one kind of work and another a different kind, but also a quantitative one, by which one man’s wants are provided for not by his own work only, but by another’s. A society without any division of labour would be one, in which each man worked for his own wants, and nobody for another’s; in any case but this there is a division of labour in this wider sense of the word. Now this division can be brought about by two means. “There are two ways” says Puchta “in which we can avail ourselves of the strength of other men which we are in need of. One is the way of free commerce, that does not interfere with the liberty of the person who serves us, the making of contracts by which we exchange the strength and skill of another, or their products, for other performances on our part: hire of services, purchase of manufactures, etc. The other way is the subjection of such persons, which enables us to dispose of their strength in our behalf, but at the same time injures the personality of the subjected. This subjection can be imagined as being restricted to certain purposes, for instance to the cultivation of the land, as with soil-tilling serfs; the result of which is that this subjection, for the very reason that is has a definite and limited aim, does not quite annul the liberty of the subjected. But the subjection can also be an unlimited one, as is the case when the subjected person, in the whole of his outward life, is treated as but a means to the purposes of the man of power, and so his personality is entirely absorbed. This is the institution of slavery”.13 We have not much to add to this lucid description of slavery and its function. The function is a system of compulsory labour, and slavery is the absorption of the whole personality of the forced labourer to this end. As this absorption is properly expressed by the word “property” or “possession”, we may define the slave as a man who is the property or possession of another man, and forced to work for him. This definition, however, on further consideration will show itself capable of some simplification. For when one man is the property of another, this implies compulsory labour. The right of property in this case, the object of it being an man, is a power over that man’s will too. The Romans recognised this: “The master has not only a right of property over the slave as over a lifeless thing, but

12  Wagner remarks that this is the main function of bondage in general (Unfreiheit). Wagner, pp. 374–376, 382. 13  Puchta, II pp. 82, 83. Wagner (pp. 382, 395) arrives at the same conclusion, but does not state it so clearly.

18

Nieboer

also a power like that over his son, the potestas dominica, that is a power over the slave’s will”.14 The right of property, that is a legally unlimited power over a man, were useless, if the owner did not influence the man’s will; and this influencing is equivalent to imposing labour upon him, labour being taken in the widest sense. A mere physical possession, such as the preserving of captives for cannibal purposes, which Letourneau and Spencer make so much of,15 is socially of little consequence. Possession of human beings, as a social institution, is that which gets hold of the will of its object. Hence it follows, that slavery is the fact that one man is the property or possession of another. This simplification of our formula has this advantage that, in inquiring whether some people has slaves, we need not ask whether there is labour imposed on subjected men. When this does not sufficiently appear, we need not say: We do not know whether slavery really exists here. When we are told that among such a people some men are the property of others (except of course the cases of mere physical possession we have hinted at, which are few and easy to recognise), we may be sure that they perform some kind of compulsory labour, and are justified in calling them slaves. Further advantages of our definition are, that it is the definition given by many theorists, and that it lies within the limits of current speech. In the following paragraphs we shall mark the distinction of slavery from some phenomena which somewhat resemble it. Of phenomena of this kind we shall consider only those that most frequently occur; other questionable cases will be examined in surveying the occurrence of slavery in the several parts of the globe. 4

Distinction of Slavery from Kindred Phenomena

4.1 Wives in an Abject Condition In the first paragraph it has already been noticed, that our feminists make very great use of the term “slavery”. We shall see that this equally applies to some ethnographers and theorists describing the state of women, especially as wives, in some primitive societies. To give one instance of each of them: Bancroft says of the Northern Californians: “Although I find no description of an actual system of slavery existing among them, yet there is no doubt that they have slaves. We shall see.... that women entitled by courtesy wives, are

14  Sohm, p. 106. 15  Letourneau passim; Spencer, Pol. Inst. p. 291.

Definition And Distinction From Kindred Phenomena

19

bought and sold”.16 The theorist we shall quote is Letourneau: “In all very primitive societies woman represents the domestic animals, the beasts of burden which the more advanced societies possess: she is indeed treated as a slave, and this certainly is one of the reasons why slavery has been instituted so late in the course of social evolution”.17 We may say that such authors use the word metaphorically (as Letourneau certainly does); but this does not exempt us from examining, whether the condition of wives in those cases, where according to them it so much resembles slavery, is really slavery. We must not, of course, inquire whether there are instances of female slaves being the wives of their owners, but whether in any case the wives as such are slaves. In doing this, we may confine ourselves to observing the condition of wives in Australia, as this condition is commonly described as a striking instance of an abject one. Letourneau remarks: “In the Australian clans slavery, in the sense in which we use the word, did not exist; but one half of the social group, the weaker half, was reduced to servitude; the Australian woman, an indispensable and despised helpmate, was during her whole life burdened with work, ill-used, and in reward often eaten by those whom her unavailing labour had fed”.18 Schurtz, in his little book, states that the treatment of the Australian wives is bad.19 Ratzel expresses the same view: “The position of the wife in such circumstances is always a low one. That she is positively considered to be the property of her husband (hence in the Adelaide district “owner of a wife” means husband) is not peculiar to Australia. But to this a number of customs are added here, that, more than among other peoples to which the notion of the wife as a commodity is equally familiar, place her in the back-ground of public and even of family life”.20 Now let us cite some particulars about this abject state of the Australian wives, as given by ethnographers. For the purpose of enabling the reader to take a comprehensive view of the matter, we shall arrange these particulars not according to the different tribes each applies to, but according to the several phenomena bearing on the object of our inquiry. This gives the following result:

16  Bancroft, p. 349. 17  Letourneau, p. 27. 18  Letourneau, p. 45. 19  Schurtz, p. 439. 20  Ratzel, II p. 66.

20

Nieboer

A. The wife is acquired by the husband without her consent being asked. So among the Dieyerie: “under no circumstances has a woman any say in the choice of a partner”; Powell’s Creek natives: “After being purchased or captured, the woman is generally taken away to a distance and kept more or less isolated with her husband for some months, until she contentedly settles down to the new order of things”; Queenslanders on Herbert River: wives are acquired by bethrothal as children, by exchange for a sister or daughter or by capture; N. W. Central Queenslanders: the marriage can be proposed by the male relatives of the woman, or a man can exchange his true blood-sister, i.e. by the same mother, for another’s blood-sister; in both cases the consent of the whole camp-council is required; aborigines of N. S. Wales: girls are often betrothed in infancy, or else given away by their father or brother without their wishes being consulted; natives of the Western District of Victoria: betrothal of children is very frequent. A girl when adult can be asked of her father, without any attention being paid to her wishes. When two young men have each a sister or cousin, they may exchange the young women and marry them; the women are obliged to obey; Southern Australians of Port Lincoln: girls are betrothed long before puberty; when adult they must follow their intended husbands whether they wish it or not; S. W. Australians: “In no case is the girl asked for her consent”; natives of King George Sound (W. Australia): a girl is often promised to a man years before her birth, but generally she is acquired by capture; Northern Australians of Port Darwin and the W. Coast of the Gulf of Carpentaria: “Wives are obtained by gifts of parents; in the majority of cases female children when born are promised to men of all ages.... Some men obtain women by stealing them, generally from other tribes, or get them in exchange for a sister”; Tasmanians: the girls are betrothed as children; before marriage they are the property of their father or brother. When the match is broken off, the girl is again betrothed, without her wishes being consulted. Brough Smith, speaking of the Australians in general, remarks: “Men obtain wives by a convenient system of exchange, by conquest sometimes, and sometimes a woman is stolen. By what mode soever a man procures a bride, it is very seldom an occasion of rejoicing for the female”. B. The wife is entirely in the power of her husband, and treated accordingly. a. Sometimes such general expressions are found, as the wile being her husband’s “property” or “slave”. So on Moreton Bay: wives are slaves; on Herbert River: wives are slaves; in N. S. Wales: “the woman is the absolute property of her husband”; in S. W. Australia: “the state of slavery in which they [the women] are all held, is really deplorable”; in Central Australia: the wife is desired by the husband only for a slave; in Tasmania: the women are slaves and do all the menial work. We may add Curr’s statement about the Australians in general:

Definition And Distinction From Kindred Phenomena

21

The wife “is not the relative, but the property of her husband”. “The husband is the absolute owner of his wife (or wives)”. Brough Smyth too remarks that the husband is called the owner of the wife. b. He treats her with contempt. Among some tribes of Queensland and South Australia women and children are forbidden to eat some kinds of food, especially such things as the men are very fond of. In S. Australia women are despised. In the Moore River District of W. Australia the husband gives his wife only the offal of the chase. Central Australian men “eat alone, and throw what they can’t eat to the women”. In N. S. Wales “as her husband walks along, she follows him at a respectful distance.... If they sit down to a meal, she still keeps behind and gets her share flung to her without ceremony”. c. He may sometimes ill-use and even kill her. On Moreton Bay the wife is often beaten by her husband, especially when he is drunk with rum. The Queenslander of Herbert River “treats his wife with but little consideration, is often very cruel; he may take her life if he desires”. In N. S. Wales the husband “may do with her whatever he likes, even to the extent of putting her to death, without any challenge from social or tribal law”. The Cammarray beats his wife violently for a trifling fault even a few hours before her confinement. Dawson speaks of the “apparent hard usage to which the women [of W. Victoria] are subjected”. In S. W. Australia “the method he [the husband] adopts for correcting her is so barbarous, that it often occurs that for a single look he pierces her leg with the ghici, breaks her head with the dauac, and treats her to other similar caresses”. The natives of King George Sound treat their wives very badly. In the Moore River District most of the women die a violent death before they have reached an advanced age. If, after an unsuccessful chase, the husband finds that his wife has not enough yams, she is glad to get off with only a flogging. In cases of famine the women are eaten. In Central Australia women are very badly and roughly treated. Nobody aids an ill-used woman. Tasmanian wives were often cruelly beaten by their jealous husbands. According to Curr, the Australian husband may “treat her well, or brutally illuse her, at his pleasure”. The wives “are, occasionally, cruelly beaten, or speared, for even a trifling offence”. And Brough Smyth states that “if she shows favour towards another and be discovered, she may suffer heavy punishment, be put to death even”. d. The husband exchanges and lends his wife. At Powell’s Creek wives are sometimes exchanged. In Queensland and S. Australia “it frequently occurs, that a woman is exchanged, and passes to a number of husbands in a few years”. The Moreton Bay aborigines lend their wives to each other and offer them to Europeans. In N. S. Wales “when visitors come to the camp they are accommodated with wives while they remain; and a brave chief, who has done much for

22

Nieboer

their tribe by his prowess, gets the wives of other men sent to him by them as a mark of respect and friendship. Two men may even agree to exchange wives for a time”. At Port Lincoln the men frequently exchange wives. The S. Australian husband offers his wife to friends and strangers. In Central Australia the husband lends his wife to his friends. When he goes abroad a husband is given her for the time. A guest is also provided with a wife. Men and parents prostitute their wives and daughters. Another writer informs us that “they often bring them [their wives] up to white men and beg of them to take them”. The natives of Port Darwin “exchange wives occasionally”. Tasmanian women were offered to whites for payment. A describer of the Australians in general states that the husband may “keep her to himself, prostitute her, exchange her for another, or give her away to any male of the same class as himself. According to another writer a young man who has no wife sometimes gets one from an old man, who is weary of her. e. After his death she becomes the property of his brother. Among the Dieyerie “the elder brother claims her as she is the wife of his brother”. On Herbert River the widow belongs to the deceased man’s brother. In N. S. Wales “when a man dies, his widow is the property of his next brother”. Among the Kurnai the same custom prevails. In N. Australia “a widow belongs to her late husband’s brother”. We may add Curr’s general statement that “when a man dies, his widows devolve on his eldest surviving brother”. C. The husband makes his wife work for him. As regards the Dieyerie we are told that “the more wives a man has, the more indolent he becomes; as they do not till the soil, each wife has to go daily in search of food, gather seeds, roots, and other vegetable products according to the seasons; the men with a plurality of wives stay at home making weapons, ornaments and fishing nets from rushes grown on the banks of the lakes”. At Powell’s Creek “polygamy is common, more so amongst the old men, who find a plurality of wives useful in hunting for them, and as carriers when shifting camp, etc.”. On Herbert River the women procure the food, and for this often make long journeys; they do all the hard work. The husband makes the frame of the hut; she covers it. When travelling she carries all that is to be carried. The husband often keeps the animal food to himself; his hunting has rather the character of a sport; the procuring of food is entirely incumbent on the wife. According to Fraser the fate of the native wife in K. S. Wales is very pitiable. “Married at an early age, she has not only to bear and rear the children, but she does all the heavy work of the family; in camp, it is her duty to put up the rude windsheiter of sticks and foliage which serves them as a home, to make a fire and keep it burning, and to cook the food; on the march, she carries in a bag, resting on her back and slung from her neck, all their portable property, and seated on this bag is her young-

Definition And Distinction From Kindred Phenomena

23

est child,... in this bag, in addition to the few utensils she requires for domestic labours, she has a yam-stick with which to dig up the numerous native roots which are used as food, a supply of these and other articles of food required for a meal, a quantity of native string and hooks for catching fish.... For the ready kindling of a fire, whenever it is required, she has to carry with her a smouldering piece of firewood; if she allows this to go out, and thus put her lord and master to the labour of getting fire by friction, or if she in any other way gives him displeasure, he will beat her severely, even till her body is covered with bruises and her hair is matted with blood”. At Victoria River Downs Station an old man generally has many wives, “probably to work and get food for him, for in their wild state the man is too proud to do anything except carry a woomera and spear.” In Western Victoria “after marriage the women are compelled to do all the hard work of erecting habitations, collecting fuel and water, carrying burdens, procuring roots and delicacies of various kinds, making baskets for cooking roots and other purposes, preparing food, and attending to the children. The only work the men do, in time of peace, is to hunt for oppossums and large animals of various kinds, and to make rugs and weapons.” In S. W. Australia “when, wandering through the woods, the savage observes that the sky threatens rain, he enjoins his wife to erect a hut at the place which he thinks most fit, and where he intends to pass the night”. At King George Sound the women look very miserable; they do all the work. In the Moore River District the wife who has not yams enough for her husband is severely beaten (as quoted above). The Central Australian wife is the drudge of her husband. About the natives of Port Darwin we get this information: “The only reason I know of for the practice of polygamy is that, as the wives have to provide food for their lords and carry all their family possessions when travelling, the husband can lead a perfect life of indolence”. Tasmanian women had to procure all sorts of food, except the kangaroo. Ling Roth quotes a description of a Tasmanian repast: “Hitherto we had had but a faint idea of the pains the women take to prepare the food requisite for the subsistence of their families. They quitted the water only to bring their husbands the fruits of their labour, and frequently returned almost directly to their diving, till they had procured a sufficient meal for their families”. Curr, surveying the mode of life of the Australians in general, remarks: “Wives have to undergo all the drudgery of the camp and the march, have the poorest food and the hardest work”. Brough Smyth enumerates as duties of the wife “building a new camp, getting firewood etc. and on journeys acting as a carrier for all the worldly goods of her husband. They are packed on her back, all excepting his war implements, which he himself deigns to carry”. This picture, surely, is very black. But, unlike Letourneau, we must not view the dark side only. We may remark, first that, as it appears from the foregoing

24

Nieboer

survey, there are with regard to each of the Australian tribes but a part of the enumerated phenomena on record; the black picture is produced by blending the dark sides of each into a whole. And, secondly, the same writers relate some particulars, which prove that the life of the Australian wife is not all darkness. These too we shall arrange in the order observed above. A. In some cases we are told, that the girl’s wishes are to some extent taken into consideration as to the choice of her husband. On Herbert Biver the woman sometimes gets the man she loves; she is then very happy; sometimes she runs away with the beloved man. She is fonder of a handsome face than of a good figure; she prefers a wild, open face, especially as to the eyes. In N. W. Queensland, when a young man and a girl are in love with each other, and the camp-council is not opposed to it, they elope, live as husband and wife for some two months, and then return to the camp. In N. S. Wales a girl, to escape from the betrothed man (oftentimes an old one), may elope with her young lover; she is then brought back and beaten by her family, “but it may be that she elopes again and again, and, if at last they see that she is determined on it, they let her have her own way”. In Tasmania the woman was stolen from her tribe, but not against her will. Most often the girl succeeded in getting from her father the man she wanted; otherwise she had to run away with him. Curr remarks: “In no instance, unless Mr. Howitt’s account of the Kurnai be correct, which I doubt, has the female any voice in the selection of her husband.” This may be true, if we take “voice” in the sense of a legally recognized right; virtually, however, she sometimes has a “voice”, as appears from the instances given here. Howitt’s account which Curr alludes to we have not been fortunate enough to meet with. According to Brought Smyth “a young man who has engaged the affections of a girl of a neighboring tribe, agrees with her to run away at the first opportunity that offers”. They are then persecuted by the members of her tribe, as custom and law require, but not energetically. After a few days the yonng man and his wife return to his tribe. Except at first some scolding and muttering his new state provokes little comment. “His young wife is treated well, and is soon familiar with all the women of the tribe, to which she has become attached”. B. a, b. Sometimes the ethnographers tell of much affection existing between husband and wife. At Moreton Bay there is often a great affection. On Herbert River “as a rule man and wife apparently get on very well”. Fraser remarks about the aborigines of K. S. Wales: “the kuri or black man is usually kind and affectionate to his jiu, wife”; “in spite of the hardness of their mode of life, married couples often live happily and affectionately together to a considerable age”. Dawson, after describing the work imposed on women in W. Victoria (as quoted above), adds: “But notwithstanding this drudgery and

Definition And Distinction From Kindred Phenomena

25

the apparent hard usage to which the women are subjected, there is no want of affection amongst the members of a family”. Even Salvado, who so pities the S. W. Australian wife, remarks: “Sometimes I heard a betrothed man say: I love her and she loves me”. Of the Tasmanians we are told that they “treat their women kindly”. Brough Smyth makes this general statement: “It is hard to believe that even in a lower state the male would not have had the same feeling of affection for his mate and an equal jealousy of love as we see among the aborigines now”. We may add, that the Tasmanian women, though overburdened with work, are described as a merry and laughter-loving kind of people. And Curr remarks about Australian women in general: “In every way the female’s looks to us a hard lot; and yet, notwithstanding, I do not hesitate to say that they are, on the whole, fairly happy, merry and contented.” c. The husband does not always enjoy such an entire freedom of action towards his wife. Sometimes, for punishing and divorcing her, he must have the consent of the tribe. So in N. S. Wales, in case of adultery “he may complain to the elders of the tribe, and they, on cause shown, decree a divorce; but not if she has children.” In W. Victoria “a man can divorce his wife for serious misconduct, and can even put her to death; but in every case the charge against her must be laid before the chiefs of his own and his wife’s tribes, and their consent to her punishment obtained. If the wife has children, however, she cannot be divorced”. Here we find also some slight traces of protection of the wife by her relatives: “A man is allowed to marry his brother’s widow, or his own deceased wife’s sister, or a woman of her tribe; but he is not permitted to do so, if he has divorced or killed his wife”. In N. W. Central Queensland the wife is avenged by her family. “In the case of a man killing his own gin, he has to deliver up one of his own sisters for his late wife’s friends to put to death, he personally escaping punishment.... A wife has always her “brothers” to look after her interests”. At the initiationfeasts “each woman can exercise the right of punishing any man who may have illtreated, abused or “hammered” her.... the delinquent not being allowed to retaliate in any way whatsoever”. If these women are slaves, they at least have their saturnalia.21 We even find cases of the wife putting a check upon her husband, especially in a sexual respect. On Herbert River the wife is furious if her husband is unfaithful to her. In N. S. Wales “a wife may similarly complain to them [the elders of 21  Curr asserts that, if the husband killed his wife, “her death would be avenged by her brothers”. But the information we get about the several tribes makes it probable, that this is not true regarding the majority of Australian tribes.

26

Nieboer

the tribe] of the conduct of her husband, and they may order both the man and her paramour to be punished”. In W. Victoria “if a husband is unfaithful, his wife cannot divorce him. She may make a complaint to the chief, who can punish the man by sending him away from his tribe for two or three moons; and the guilty woman is very severely punished by her relatives”. “A chief who has been married under the law of betrothal, is not permitted to marry another woman for a long time; and should he do so without obtaining the consent of his wife, there would be constant quarreling”. At Port Lincoln an old, former wife sometimes forces her husband to desist from taking a young, new one. Finally we meet with instances of the wife having a real ascendency over her husband. On Herbert River the husband is sometimes led by his wife, and even beaten by her. A curious piece of information we get about W. Victoria. When a wife treats her husband with such persistent disrespect or unkindness as to make him wish to get rid of her, he goes away to some neighburing tribe and tries to bring about her death by sorcery. The wife, being informed of this, repairs thither and entreats him to return, and so a reconciliation is effected. In Tasmania the husband could divorce his wife; but she could also force him to do so. d. Exchange of wives does not seem always to take place against their will. In W. Victoria wives may be exchanged only after the death of their parents and with the consent of the chiefs, but not if one of them has children. After the exchange both couples live amicably together in one hut, each in a separate compartment. If a man knows that his wife is in love with another, and he is not opposed to it, she can be amicably transferred to the other man with the consent of the chief. At Port Lincoln the men frequently exchange wives; brothers and near relatives have their wives nearly in common. The wife calls the brothers of her husband by the name of husbands. This thus seems rather a kind of group-marriage than a bartering of wives as of commodities. These two instances point to the possibility that in other cases too exchange of wives may be not so arbitrary an action as at first sight it seems. As to the lending of wives, in some cases it appears that these offer themselves to strangers. In If. S. Wales the husband “is quite ready to bargain with a white man, and with her consent too; for a black woman considers it an honor to be thus courted by a man of a superior race”. The Cammarray women prostitute themselves to Europeans for almost nothing, and among themselves without any shame. In Central Australia marriage does not impose any obligation of chastity; the wives always prostitute themselves. On Moore River the wives often have connections with young men; the husbands do not seem to take much notice of it.

Definition And Distinction From Kindred Phenomena

27

e. The levirate law sometimes appears in the character of a duty rather than a right of the deceased man’s brother. Fraser (describing If. S. Wales) calls it a “refuge” to the widow. Dawson, speaking of the aborigines of W. Victoria, states: “When a married man dies, his brother is bound to marry the widow if she has a family, as it is his duty to protect her and rear his brother’s children”. Salvado speaks of the philanthropy of the S. W. Australian, who takes upon himself the care of the wife of an absent friend or parent, or of a brother’s widow. C. Among the Kurnai the man must hunt for the sustenance of his wife and children, and fight for their protection. The last-cited cases of levirate law, too, show that the subsistence of the family does not depend on the wife only Even the instances quoted above under C provide us with evidence that the men perform some kind of work as hunting the larger animals, making weapons and fishing-nets, getting fire by friction, etc. And what is said here about the Kurnai certainly applies to all these tribes: the husband fights to protect his wife. This being his great and indispensable function, we must not wonder at his not liking to do other work that women can perform as well.22 The division of labour between the sexes is not always so unreasonable as at first sight it seems. Hore, speaking of the African Wajiji, very justly remarks: “Much has been said about the unfair division of labour in such circumstances, but when it is considered that a wild man finds scarcely anything to his hand, but must himself cut the wood and the grass to build his house, manufacture his spear and cooking vessels, take his part in tribal duties, and is frequently compelled to seek food in long and laborious hunting expeditions, it will be seen that he often gets his fair share of work”.23 A similar division of labour is admirably described by Pinart, as existing among the Indians of Panama: 22  Literature referred to in surveying the state of the Australian wife. On the Dieyerie: Gason in Frazer’s Notes, p. 170; Powell’s Creek: The Stationmaster, eod. pp. 177,178; Victoria River Downs Station; Cranford, eod. p. 181; Queensland and S. Australia: Matthews, eod. pp. 187,188; Moreton Bay: Lang, pp. 337,338; Herbert River: Lumholtz, pp. 100, 160–164, 213; N. S. Wales: Fraser, pp. 2,26–28, 35, 35; Cammarray: Collins pp. 559–562; Kurnai: Fison and Howith, pp. 204, 206; N. W. Central Queensland: Roth, pp. 141, 176, 181; W. Victoria: Dawson, pp. 27, 28, 33–37; Port Lincoln: Woods, p. 223; S. Australia: Angas, I pp. 82, 93; S. W. Australia: Salvado pp. 313, 314, 349; King George Sound: Browne, pp. 450, 451; Moore River District: Oldfield, pp. 248–251; Central Australia: Eyre II p. 322; Willshire in Frazer’s Notes, pp. 183, 184; Port Darwin, etc.: Foelsche eod. pp. 194; Tasmania: Bonwick, pp. 56, 62, 66, 68, 73, Ling Roth, Tasmania, pp. 125, 46; Australia in general: Curr, I pp. 106–110, Brough Smyth, I pp. 76,79–82, 85, 86. 23  Hore, p. 11.

28

Nieboer

“I may be allowed to make here a short digression on woman’s place in the Indian family. It is commonly said by those who have not lived intimately with the Indians, that they consider woman as a beast of burden, that to her share falls a life full of troublesome and fatiguing work, and to the man’s an easy and idle existence. It may, indeed, seem strange to the superficial observer to see the woman charged with heavy burdens and the man walking before her carrying nothing but his weapons. But if the observer will only reflect a little, he will understand that, whereas the man carries his weapons only, the responsibility and the safety of his wife and children are incumbent on him. The Indian’s life is indeed surrounded with dangers; when traversing a savannah, or forest, a hostile Indian may appear at any moment; a tiger, a snake etc. may throw himself upon the travellers. Therefore it is the man’s task to be continually on the alert, to have his hands and his movements free, in order to be able immediately to take his arms and defend those who are dear to him. How often have not I seen the Indian, when about to traverse a river, making his family stand still, entering into the water and reconnoitring whether it was not too deep or the stream too rapid; then inspecting the opposite bank to see whether all was right there; then crossing the river again, helping his wife and children to pass through, often even carrying the burdens, and several times re-crossing the river to transport on his back his wife and children. The river being crossed, the man again takes the lead with his arms, the wife and family resume their burdens, and the little caravan continues its way in the same order”.24 Another fact, proving that the Australian women are not in every respect considered as slaves, is the great influence they often have in intertribal matters. “The peace-making influence of the women is very great, and has often been observed among many tribes”. “The peace-making function of the women is also very characteristically shown by their being employed as international ambassadresses.25 Darwin justly remarks: “We thus see that with savages the women are not in quite so abject a state in relation to marriage, as has often been supposed”.26 The question to be settled now is this: Are these Australian wives, and accordingly all the wives that live in an abject state, to be called slaves? Remembering the conclusion we arrived at in the third paragraph, we may put the question thus: are they objects of possession? Under B a we have quoted several statements of ethnographers calling them the slaves, or the property, of their husbands. We must not, however, forget Ingram’s warning against taking 24  Pinart, pp. 44, 45. 25  Steinmetz, Strafe, II p. 45. 26  Darwin, Descent of Man, p, 593.

Definition And Distinction From Kindred Phenomena

29

a rhetorical use of the word “slave” too literally. The facts recorded under Bc, Bd, and C are of more interest to us. The husband may do with his wife as he likes: ill-use and kill her, overtax her with work, exchange and lend her. It is but seldom that her family protect her; in but very few cases the man’s power is interfered with by the chief or elders of the tribe. Therefore we cannot but admit that she is the property of her husband. Yet there is a reason, why we are not to bring these wives under the denomination of slaves. We may refer here to the point of view we have taken in determining the nature of slavery. Slavery is an organ in the social body, that in a peculiar manner brings about a division of labour. The Australian wives share the character of this organ as an object of possession. Yet they are not the same organ; for besides being forced labourers they are wives; hence it follows that their relation towards their husbands is wholly bound up with the sexual and family life: it is their character as women, not as labourers, that prevails. We may remember here the mutual affection observed in so many cases by our ethnographers. As the mother of his children, too, the husband is likely to value his wife. We have seen (under B c) that in a few cases she cannot be divorced or exchanged if she has children. Besides, it is frequently stated, that the Australian aborigines are very fond of their children.27 The Australian woman discharges the duties of a wife and a mother, and besides, to some extent, the work that among other peoples falls to the share of the slave; therefore she is not a slave. If she were, her place, in a slave-keeping society, would be entirely occupied by the slave; but no one will doubt whether in any such society there are wives. In an evolutionary sense the slave and the Australian wife differ in this: the Australian wife is a not-yet-differentiated organ, performing two functions, which at a later stage of development will be incumbent on two quite distinct organs: the peculiar function of a wife, and the labour of a slave. This reasoning is not an assertion a priori, by a biological parallelism, of a development that must actually have taken place; it is only intended to show the fundamental difference existing between wives however abject their condition, and slaves.28 We may even go farther and say: Slavery proper does not exist, when there are none but female slaves. For when females only are enslaved, the reason probably is, that they are valued as women, not only as labourers; otherwise males 27  See Ploss, II pp. 333, 334, and Steinmetz, Das Verhältniss zwischen Eltern und Kindern, p. 613. 28  Lippert (II p. 535) distinguishes the wife, as mistress of the household, from the slave, who has no share in the authority wielded by the master. This may be true, but it is only a small portion of the truth.

30

Nieboer

would be enslaved too. And even where such women are not, all of them, actually treated as wives or concubines, but only kept as labourers, there is no slavery in the true sense of the word. In such cases, the husband keeps his wife or wives subjected; this leads to the keeping of numerous subjected females, who are scarcely to be called wives. But it is always women, as the weaker sex, who are subjected to the men; subjection of labourers, only in their quality of labourers, does not exist. The labourers have the name, if not the state, of wives; this proves that the subjection of labourers as such, i.e. slavery, is not yet developed. We have dwelt at considerable length on this distinction between slaves and subjected wives. There are some more distinctions to be made between slavery and kindred phenomena; but these will not occupy so much space and time. 5

Distinction of Slavery from Kindred Phenomena

5.1 Children Subjected to the Head of the Family There was a time, the time of the good old patriarchal theory, when the condition of children in the early stages of social life was thought to be one of complete subjection to the head of the family, the pater familias, who had over them an unlimited power, extending to the power of life and death. Carey, among others, holds this view, and very plainly expresses it. “By nothing is the progress of mankind in population and wealth made more manifest than by the change in the relation of parent and child. In the infancy of cultivation the one is a tyrant and the other a slave”.29 The adherents of the matriarchal theory have assigned to the Roman-like agnatic family its place as a later product of history; but to the question as to how children were treated in an ante-patriarchal state of culture they have not given much attention. It is to Dr. S. R. Steinmetz that we are indebted for the first exact inquiry into primitive education. His conclusions, based upon a large amount of ethnographical materials, are these: With most savages rational education is out of the question, the children soon growing independent, and when young being either neglected or much petted and spoiled;30 a lesser number of savage tribes show some 29  Carey, p. 275. See also Wuttke and Maine as quoted by Steinmetz, Strafe II pp. 180–181. 30  Chamberlain (p. 116) justly remarks: “Much too little has been made of the bright side of child-life among the lower races.”

Definition And Distinction From Kindred Phenomena

31

slight beginnings of education without or nearly without bodily castigation; in a few cases the children are under strict discipline. In this last set of cases there is to some extent a subjection of the children. “With the power over the mother the father gradually acquired the power over the children.” “The patriarch became master of his children and, whenever circumstances required and allowed it, introduced a strict discipline over them”.31 We may therefore suppose, that there will be instances of children being treated in a somewhat slave-like manner. We shall presently see that there are a few such cases on record in Steinmetz’s book. Among the Apaches the father holds unlimited sway over his children up to the age of puberty.32 Tlinkit boys must render unbounded obedience to their parents and especially to their maternal uncle, to whom, according to the law of inheritance, they are almost more nearly related than to their own father. They have to perform the labour imposed upon them, without any claim to compensation.33 Of the Botocudos we are told, that the father, being stronger than his children, compels them to work for him.34 Among the Aeneze Bedouins the young girls work hard; they drive the cattle to the pasture-ground; if one out of the herd is lost, they are severely beaten by their father.35 Among the Assja Samojedes the father has a patriarchal power, and punishes at his discretion and according to custom.36 In these few cases only is it clearly stated that the head of the family has an arbitrary power. The value of Zu Wied’s statement about the Botocudos is much lessened by the same ethnographer telling us that the children enjoy much freedom.37 Considering now the state of the children in the cases referred to here, are we justified in calling it slavery? 31  Steinmetz, Strafe II pp. 179–253, see especially p. 252. Dr. Steinmetz has recently enlarged upon the same subject in an article entitled: Das Verhältniss zwischen Eltern und Kindern bei den Naturvölkern. 32  Steinmetz, Strafe II p. 190 (after Bancroft). 33  Ibid., p. 194 (after Krause). 34  Ibid., p, 196 (after Zu Wied). 35  Ibid., p. 199 (after Burckhardt). 36  Ibid., p. 201 (after Von Middendorf). 37  Ibid., p. 196.

32

Nieboer

The head of the family has power over the children; and so far as it appears from the particulars given by the ethnographers, this is a legally unlimited power, that may be called right of property, and is likely to lead to compulsory labour, as among the Tlinkits and Aeneze Bedouins it certainly does. The condition of these children may therefore be expressed by the word “possession”, our criterion of slavery. We may even go farther. The condition of slaves is not always very bad; but however kindly treated, they are slaves, are the property of their masters. So with children too. They may not be, as in the cases mentioned above, under strict discipline; yet the father’s, or in a few cases the maternal uncle’s, power, however moderate a use he makes of it, may be legally unbounded, not restricted by social rules, not interfered with by the community. In such a case the head of the family may be called owner of the child, and is really called so in Roman law, so clearly distinct from Roman practice. “The patria potestas of ancient civil law means the full power of the father over the persons subjected to him (the child, the grand-child by the son, the wife in manu), the right of death and life (ius vitae ac necis) and the right to sell into slavery”.38 “This potestas originally was equal to that over the slaves”.39 We see that the term “possession” may well be used here. Yet there is a reason that induces us not to call these children slaves, a reason resembling that for which we have excluded the subjected wives. These children may be called the property of their fathers; but this is not the whole, nor even the main part of their condition. The relation between father and child, if it includes subjection, includes much more. There is mutual sympathy and in many respects a coincidence of interest; there is respect on the side of the child; there is on the side of the father a desire to promote the welfare of the child, however much bound up with egotistical motives. There is also physical and mental superiority on the side of the father and inferiority on the side of the child;40 and this in some cases may bring about a somewhat slave-like condition of 38  Sohm, p. 363. 39  Puchta, II p. 384. As this is not the place to enter into a systematic description of the treatment of children among savages, we have confined ourselves to mentioning the results of Steinmetz’s investigations. Yet we will quote here one ethnographical record, that clearly shows the high degree of development of the patria potestas possible among savages. “In Flores the sons even of rich families as long as their father lives, at public feasts are dressed like slaves, and also at his funeral; this being apparently the external sign of a strict patria potestas, which remains in force till the funeral; until then the son is the father’s slave”. Von Martens, p. 117. 40  Viz. as long as the child is really a child. Savage children are generally much sooner fullgrown than those of civilized nations, see Steinmetz, Strafe, II pp. 215–217.

Definition And Distinction From Kindred Phenomena

33

the latter; but this condition is not an essentiel part of the relation between father and child; a fortiori it is not coextensive with the relation, as in the case of the slave. Biologically expressed: the child is quite another organ, with quite another function, but in some cases, performing in some degree the function of a slave; therefore it is not a slave. We may add, that the child is only temporarily subjected; one day he will be a master himself.41 This also bears upon the treatment of the child: the slave is brought up to servility, the child to authority. Children, too, as such can never form a subjected class. As for adopted children, it is not always easy to distinguish them from slaves. Sometimes they are rather severely treated, especially those captured in war or kidnapped. Tanner was thus adopted by an Indian of the Shahnee tribe. The youngest son of this Indian had lately died, and his wife had told her husband, she could not live if he did not restore her the child. The husband accordingly went off, and came back with Tanner whom he had kidnapped. Tanner was adopted on the grave of the deceased boy, and given an Indian name. But the adoptive father treated him not at all as a son. He had to do the hardest work, got but little food, and was often severely beaten. If the mother tried to protect him, she was beaten too. Finally the father, regardless of the mother’s wishes, sold him to an old Indian woman, who now became his adoptive mother. She treated him kindly, yet made him cut wood, carry water and meat, and perform other kinds of labour, which generally were not imposed upon children of his age. However, he was not a slave. When full-grown, he was considered by the Indians one of their tribe, and married an Indian girl.42 In the second Chapter we shall meet with more instances of captives being adopted either into the tribe or into one of the families within the tribe. As long as such persons are children, it is often not easy to see whether they are slaves or adopted children, it is not always stated, as in Tanner’s case, that they are formally adopted. We must ask then, what becomes of them when full-grown. If they have still a master to serve, it is clear that they are slaves; for if they are adopted members of the community, they will be free when adult, excepting the (most often slight) moral obligations of full-grown children towards their parents. Other facts proving that the captives are slaves, are their not taking share in government affairs, when the tribe is democratically organized, and their being excluded from marriage with native-born women. With the aid of these criteria we shall try, in every particular case, to decide whether the captives are slaves or adopted members of the community. What has been said here of captives, equally applies to purchased persons. 41  See Lippert, II p. 535. 42  Tanner, pp. 8–17, 114; see also p. 315.

34

Nieboer

The last two paragraphs show that there is still something wanting in our definition. Not every state of possession is slavery; those arising from family relations are to be excluded. Thus only can we come to a true understanding of the signification of slavery. For wives and children may accidentally be forced labourers and the like; the slave only is ex definitione a subjected person, a forced labourer, an object of possession. Wives and children there would be, and there are in many cases, without subjection; slaves there are not where there is not subjection and compulsory labour. A society that begins to keep slaves, develops a new organ with a special, well marked function; and it is the evolution of this organ we are to trace in the following Chapters. Our definition therefore wants an addition. We may now put it so: Slavery is the fact, that one man is the property or possession of another beyond the limits of the family proper. 6

Distinction of Slavery from Kindred Phenomena

6.1 Members of a Society in their Relation to the Head of the Community Bastian, after remarking that in a social community nobody is literally free, gives a great number of quotations, describing widely different kinds of subjection, and among these some few, where the subjects of a despot are called his slaves or his property. “The Siamese are all (even marked) slaves of their king.” “The subjects of the king of Djagga are slaves, who may not marry without his consent.” “In Usumbara all are slaves of the king.” “The absolute rulers dispose of all their subjects as their property (even without having acquired a right by the subjects having transgressed the law), and even mark with their badge the different working-guilds, as is done by the king of Siam.” “The princes and princesses on the Congo have the right to sell any one who is not a prince like themselves”.43 What Bastian means by heaping up these various quotations, without any order or attempt at an explanation, is not clear. We, however, must not follow his example, but inquire whether the word “slave” is rightly used here, whether the subjects of a despot may be called slaves. A few moments of consideration will show that they may not. For however great the power of the chief, the king, the despot, in a word the head of the community, over his subjects, they are not his property. “Property” supposes a power of the master, pervading the

43  Bastian, Rechtsverhältnisse, pp. 15, 15 note 2, 487, 187 note 2. Post (Ethn. Jur. I p. 358) also speaks of subjects being the slaves of the king.

Definition And Distinction From Kindred Phenomena

35

whole life, personal, domestic and social, of the slave; so great a power over his subjects a chief never has. The following reasoning will make this clear. Slavery would not be capable of much development, if it depended upon the master’s personal superiority only; for slavery to become a social system, the master’s power over the slave must be recognized by the society. The slave lives in a society that regards him as a slave; slavery cannot exist where there is not a society of freemen. Therefore the despot, however great his power, is not as such a master of slaves. The slave-owner has the community on his side; the chief has subjects who themselves compose the community. Looked at from the practical side, the chief’s power contains much more of voluntary submission than the slave-owner’s. A chief never has the whole person of the citizens subjected in his own behalf; he may exact some performances for his personal benefit, but the restrictions put upon the subjects, encroaching on their freedom in private life, will generally be measures taken in the (real or supposed) interest of the community, and approved of by the community. These restrictions are mutual, and arise from the social connection itself; this is not, as in slavery, using one person as a means to the purposes of another definite person. This yet more distinctly appears, where not a single man imposes these rules, but the council of citizens. In a communistic society there would be an entire absence of personal freedom; yet there would be no slaves, as there would be no freemen whom they could serve. It need hardly be said, that a chief may keep slaves like any other freeman. The public power as such, the state, also sometimes keeps slaves (e. g. the servi publici in Rome). But these slaves are quite distinct from the main body of citizens. Sometimes it is stated, that the chief, or the public power has slaves, whereas no mention is made of any other slaves. In such cases the slaves generally become such as a punishment for some offence. Where such a state of things exists, we may not speak of a slave-keeping people. For here the power of the government is so great, that it can avail itself of the labour of the citizens; whether this is done by imposing an equal amount of labour on all of them, or by selecting a few persons for this purpose and keeping them in a slavelike state, does not matter much. Besides, slavery here cannot have the same influence on social life it generally has; for every freeman has to work for himself. This kind of slavery may be compared with the tread-mill and other kinds of penal servitude existing in more civilized societies. And we may not speak of a slave-keeping people, where the only slaves are criminals, who become the slaves of him who represents the public power, any more than we can say that slavery exists in those civilized countries, where penal servitude is still practised.

36

Nieboer

One more remark has to be added here. Hitherto we have used the terms “possession” and “property” synonymously as indicating the nature of slavery. In this paragraph it has been shown, that an essential feature of slavery is its being recognized by the community. Therefore we prefer the term “property” that, better than the other term, conveys the notion, not only of a virtual subjection, but of a subjection considered legal in those communities where it exists. 7

Distinction of Slavery from Kindred Phenomena

7.1 Subjected Tribes; Tributary Provinces; Lower Classes; Free Labourers We shall meet with instances of tribes, the members of which are bound to perform some kind of labour for other tribes or for the members of the latter. This is not slavery; for slavery is subjection of one individual to another, and a subjection that absorbs the whole personality of the subjected; and under such circumstances it is not possible that the subjected lead a tribal life. Therefore, where the subjected are described as forming a separate tribe, we may be sure that they are not slaves. Ingram justly remarks that “the lowest caste may be a degraded and despised one, but its members are not in a state of slavery; they are in collective, not individual, subjection to the members of the higher classes”.44 What Ingram says here of the lowest caste, often applies to subjected tribes. That conquered districts, bound to pay a tribute in kind or money, do not consist of slaves, is clear. The foregoing remarks would be almost superfluous, were it not that some ethnographers in such cases spoke of “slave tribes” and “slave districts”. Lower classes can be of different kinds. Where they are only considered inferior to the upper classes, or excluded from governmental functions, it is easy to see that they are not slaves. Greater difficulties are presented by some other cases. Sometimes a lower class consists of free labourers. Now theoretically free labourers are easy to distinguish from slaves: the slave is compelled to work, the free labourer voluntarily submits to it. But the accounts of our ethnographers do not always make it clear, which of these two kinds of labourers we have to deal with in any particular case. When a labourer lives in the house of his master and is wholly dependent on him, it may be rather difficult at first sight to decide whether he is free or a slave. Sometimes the details 44  Ingram, p. 3.

Definition And Distinction From Kindred Phenomena

37

given are sufficient to settle the question; if not, we shall have to leave it undecided. A lower class can also consist of serfs. What they are, and what is the difference between them and slaves, will be shown in the next paragraph. 8

Distinction of Slavery from Kindred Phenomena

8.1 Serfs What we have said of free labourers applies also to serfs: to draw the theoretical line of demarcation between them and slaves is not so very difficult; but practically it is not always easy to decide, whether a subjected class we get some information about consists of slaves or of serfs; sometimes even, because of the unstable terminology and the scanty information, it is quite impossible. But there are several unequivocal cases of serfdom, too, on record in history. Mentioning a few of these will suffice to give the reader a clear idea of its nature as distinct from slavery. In Germany Leibeigene was, in the earliest times, synonymous with slave. The law placed the Leibeigenen on a level with the domestic animals. The master had the ius vitae ac necis, an unlimited right to sell them, the right to exact from them all possible services, to marry and divorce them. The owner of the Leibeigene was also owner of his goods and chattels. The lord was responsible for any damages caused by his servant, as for those caused by his horse, and might claim indemnity if any one injured his man. But gradually this slavery was mitigated into a state of subserviency. First the claim to unlimited services was waived, and on the Leibeigene were imposed definite Roboten (labour dues) and tributes. He had to work on fixed days, to perform fixed services, to pay fixed sums. His earnings legally still belonged to the lord, and the latter succeeded to his goods; but from the 13th century the lord’s right of inheritance dwindled into a present (mortuarium). From the 14th century the serfs acquired a usufruct of the soil they tilled, and so their obligations assumed more or less the character of a quit-rent. Sometimes they were even allowed to choose another lord. In the Frankish empire the lords were already forbidden to sell them abroad; from the 13th century they lost the right to kill them, and afterwards also the right to whip them. The church took away from the lord the right to divorce his serfs, if the marriage had been contracted with his consent. The ius primae noctis remained longer. Moreover, the relations of the serfs towards others were gradually recognized by law, at first only as to unjust acts, later on as to contracts. And so, when at last serfdom was abolished, the only

38

Nieboer

changes effected by this were: allowing of the right of emigrating, abolition of the marriage-consent and of the court-services and personal tributes. Thus Siegel describes the development of serfdom in Germany.45 Other writers come to nearly the same conclusions. According to Brunner, there was among the Western Germans a class between freemen and slaves, called Liten or Aldien, a hereditary class, whose position was secured by law. They had the right of acquiring property and making contracts; they could by emancipation become fully free, or ransom themselves by means of their fortune. To marry they wanted the consent of their lord. They had the right of feud (Fehderecht), and when they were killed a wergild was paid, that fell partly to the lord. Their right of inheritance was originally not recognized.46 Schröder remarks, that the difference between freemen and subservients (Hörigen) consisted in this, that the landed possessions of the latter were smaller and liable to tribute. Moreover, they had no connubium with freemen, nor any political rights; the wergild paid for them was one half of that paid for a freeman.47 In medieval Prance a similar state of things prevailed. There were no longer slaves, but serfs. “Serfdom is a transitory stage between slavery and entire liberty. The serf of the middle ages is not, like the ancient slave, indissolubly riveted to his condition, deprived of rights by his very birth, placed on a level with the beasts of burden of his lord’s estate. Public opinion is favourable to him.” “The facts agree with the doctrine. The serf has some means of acquiring property; he may marry and have legitimate descendants, who will succeed to his goods; he may give evidence in the courts; he may purchase his liberty by means of his peculium. By getting some profits he is interested in the cultivation of the soil. Giving his labour to the land, he may expect to enjoy the fruits of it, by paying fixed tributes. By marrying his children to free women he secures the liberty of his offspring. By paying an indemnity he acquires the succession to his father’s inheritance, and the right of property over his savings.... He may dispute the tributes (tailles et cens) which the lord levies on the tenement he cultivates, invoke an enquiry of experts who attest his means, contract to pay a fixed annuity and so know beforehand what profit he may depend upon”.48 With these serfs may be compared the Roman coloni. “The colonatus consists in this, that men are inseparably attached to a landed property for the 45  Siegel, pp. 291–293. 46  Brunner, I pp. 101, 102. 47  Schröder, p. 41. He states that these Hörigen were also called lati or aldio (1.e., p. 40); so they are the same class as those described by Brunner. 48  Gasquet, II, pp. 281, 282.

Definition And Distinction From Kindred Phenomena

39

purpose of cultivating it.... This connection with a determinate estate, from which the colonus might be severed only in some cases fixed by law, brought about an approximation of the colonus to the slave (as servus terrae), but also a difference between them, a security for the colonus, which protects him from the lord’s arbitrary power. Hence the colonus stands with regard to the lord on the free footing of one bound only to comply with the yearly canon, annua functio, a tribute fixed by contract or custom, which he has to pay to the lord, generally in products of the land”.49 The foregoing statements once more prove the sufficiency of our definition of slavery. As soon as the forced labourer is no longer entirely at the disposal of the lord, the latter being entitled to fixed services and tributes only, such a state of things is called serfdom, or colonatus, or subserviency, but not slavery. This agrees with our definition of slavery. The slave, as we have remarked above, is the property of his master, whose power is in principle unlimited, not restricted to fixed performances. Therefore, even if the writers referred to here called such institutions as serfdom and colonatus slavery, we are not to do so; but we may regard it as a corroboration of the conclusion we had arrived at before, that such writers, most of whom have not made any special research into the nature of slavery, when they meet with such an institution as serfdom, feel that they are not to call it slavery. Now let us look what our theorists have to say on the subject. Ingram remarks: “The transition to serfdom took place in civic communities, when the master parted with or was deprived of his property in the person of the slave, and became entitled only to his services, or a determinate portion of them. In rural life, where the march of development was slower, the corresponding stage was reached when, in accordance with the fundamental principles of feudalism, the relation between the lord and serf, from being personal, became territorial”.50 The first words here perfectly express the truth: when the master loses “his property in the person of the slave”, he is no longer a slave-owner. What follows, that the master “became entitled only to his services”, is less correct; for he who is entitled to all the services of another is his owner; just the limiting of the master’s right to “a determinate portion of them” is the change from slavery to something else. If I may require all the services a man can perform, I am his owner; if I am restricted to a determinate portion of them, I am not.

49  Puchta, II p. 97. 50  Ingram, p. 262.

40

Nieboer

Spencer says: “As the distinctions between different forms of slavery are indefinite, so must there be an indefinite distinction between slavery and serfdom, and between the several forms of serfdom. Much confusion has arisen in describing these respective institutions, and for the sufficient reason that the institutions themselves are confused”.51 This consideration, however true, will not prevent us from drawing a theoretical line of demarcation. Not a single social institution is practically strictly separated from kindred institutions; yet we cannot understand such institutions, unless we make a distinction, and not an “indefinite” one. Letourneau, after describing the state of the colonus, adds: “In a word, he was not an object of possession, a slave, but only a proletarian attached to the soil.” In another passage he remarks that slavery always undergoes same mitigation in the course of civilization: “Less and less is the person of the slave himself oppressed; one is contented with exploiting him, bereaving him in a larger or smaller degree of the fruits of his labour, in a word the slave becomes a serf”.52 These quotations may suffice to show that our view of the matter is held by theorists as well as by historians. The serf, therefore, is not a slave, because he is not the property of his master, and the particulars of serfdom related by our historians provide us with means of more clearly understanding the practical meaning of this notion “property”. It means a power that, however leniently exercised in many cases, is in principle unlimited. Among many peoples the master may ill-use and even kill his slave, without the law taking any notice of it. And even where his power is restricted by social regulations, he may have a right of property, viz. if his authority be in principle unbounded, and any limitation put upon it suppose a special legal provision. Our social reformers of to-day express the desire to make a great change in the regulation of the right of property. Till now, in Roman and Romanized modern law, the proprietor has had a right to do with his property whatever he is not by special rules forbidden to do;53 they will place the right of property on a level with the other iura in re, so that the proprietor may do nothing but what he is by law allowed to do. Now such a reform as these people intend to bring about as to property in general, has taken place wherever slavery has passed into serfdom. The slave-owner may do with his slave whatever he is not by special laws forbidden to do; the master of a serf may require from his man such services and tributes only, as the law 51  Spencer, Ind. Inst., p. 472 52  Letourneau, pp. 423, 355–356. 53  See Dargun, p. 3.

Definition And Distinction From Kindred Phenomena

41

allows him to require. The slave-owner has a right of property; the master of a serf has, so to speak, a ius in re aliena. 9

Pawns or Debtor-Slaves

In the course of our investigation it will be shown, that among some peoples a debtor, unable to pay a debt he has contracted, becomes the slave of his creditor. Sometimes such persons are ordinary slaves; but pawns or debtor-slaves in the restricted sense (who are of frequent occurrence in the Malay Archipelago, Dutch pandelingen) are a class whose slave-state is conditional; they become free as soon as the debt is paid by or for them; the creditor cannot refuse to accept the money. Because of this great difference between pawns and ordinary slaves (who generally have not a right to be ransomed), most of our ethnographers do not call the former slaves, but give separate descriptions of slavery and pawning. The question arises, and has to be settled here, whether we for our purpose have to call these pawns slaves. We shall quote here one description of pawning. Among the Tshi-speaking peoples of the Gold Coast of West Africa “a pawn is a person placed in temporary bondage to another by the head of the family.... either to pay a debt, or to obtain a loan.... When a person is pawned on account of a debt, the services of the pawn, even should they extend over a considerable number of years, count for nothing towards the liquidation of the debt; and a pawn has to serve his master, until the amount of the original debt with 50 per cent, interest, is paid by the person who pawned him”.54 Here the debtor pawns one of the members of his family; among some other peoples (e. g. in the Malay Archipelago) he pawns himself; this is not essential. The main fact is that the pawn is in “bondage”, however temporarily, that he “has to serve his master.” Therefore, as long as the debt remains unpaid, the pawn is in the same condition as a slave. He has not to perform a fixed amount of labour, he must serve his master without any limitation; the master has over him a power that is, in principle, unlimited. Now we have to inquire: Is this pawn a slave, i.e. is he the property of his master? In a legal sense the creditor has not a right of property over his pawn; his right agrees with a kind of pignus which the Romans called antichresis, i.e. something yielding profit was handed over to the creditor, who utilized it instead of receiving the usual interest.55 54  Ellis, Tshi-speaking peoples, p. 294. 55  Puchta, II p. 250; see also Wilken, Pandrecht, pp. 42–44.

42

Nieboer

Yet the right of the holder of the pawn bore much resemblance to that of the owner: he had a utilis in rem actio, a vindicatio pignoris.56 We, for our purpose, may classify the pawns among the slaves, if we can prove that sociologically a system of pawning performs the same function as a slave-system. And this certainly is the case. The same system of compulsory labor, the same subjection of the entire person exists, whether the subjected are perpetually slaves or temporarily pawns, viz. in those cases where, as among the Tshi-speaking peoples, the master’s power is in principle unlimited. Where pawns have a fixed amount of work to do, they are temporary serfs; but where (as is most often the case) no limit is put to the amount of work the master may exact from them, they are temporary slaves, and as long as they are slaves, take the same place as other slaves in the social system. 56  Puchta, II p. 264.

The Causes of Slavery or Serfdom: A Hypothesis Evsey D. Domar

I

The purpose of this paper is to present, or more correctly, to revive, a hypothesis regarding the causes of agricultural serfdom or slavery (used here interchangeably). The hypothesis was suggested by Kliuchevsky’s description of the Russian experience in the sixteenth and seventeenth centuries, but it aims at a wider applicability.1 According to Kliuchevsky, from about the second half of the fifteenth century Russia was engaged in long hard wars against her western and southern neighbors. The wars required large forces that the state found impossible to support from tax revenue alone. Hence the government began to assign lands (pomest’ia) to the servitors, who were expected to use peasant labor (directly and/or via payments in kind and/or money) for their maintenance and weapons. In exchange, the servitor gave the peasants a loan and permitted them, free men as yet, to work all or part of his land on their own. The system worked rather badly, however, because of shortage of labor. Severe competition among landowners developed, the servitors being bested by lay and clerical magnates. Things became particularly difficult for the servitors after the middle of the S ource: Domar, Evsey D., “The Causes of Slavery or Serfdom: A Hypothesis,” Journal of Economic History 30 (1970): 18–32. © Cambridge University Press, reproduced with permission. * For many helpful comments on an earlier draft, I am grateful to the following persons: Abraham Becker, Oleg Hoeffding, Clayton La Force, Edward Mitchell, William Parker, George Rosen, Matthew Edel, Peter Temin, Helen Turin and Charles Wolf, Jr. Alexander Gerschenkron’s earlier suggestions were also very helpful. Thanks are also due Ann Peet for her excellent research assistance.  I am also grateful to the RAND Corporation for its support of an earlier version of this study (20 October 1966), and to the National Science Foundation for its assistance (Grant No. NSF-GS-2627) in revising and extending the first draft. Neither these two organizations, nor the persons listed above, are responsible for the views expressed here. 1 V. Kliuchevsky, Kurs russkoǐ istorii (Moscow: Gosudarstvennoe sotsial’no-ekonomicheskoe izdatel’stvo, 1937). The original work was published in 1906. All my references apply to the 1937 edition. An English translation by C. J. Hogarth, A History of Russia, was published in New York by Russell and Russell in 1960. For specific references, see Part II.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_004

44

Domar

sixteenth century when the central areas of the state became depopulated because of peasant migration into the newly conquered areas in the east and southeast. Under the pressure of the serving class and for certain other reasons, the government gradually restricted the freedom of peasants, already hopelessly in debt to their landlords, to move. They became enserfed by the middle of the seventeenth century, though the process itself continued for many decades to come. This is a very rough summary of Kliuchevsky’s story which hardly does him justice but which will serve my purposes until Part II. Like many a historian, he assembled and described the relevant facts (and in beautiful Russian at that) and stopped just short of an analytical explanation. The economist would recast Kliuchevsky’s account as follows: The servitors tried to live off rents (in one form or another) to be collected from their estates. But the estates could not yield a significant amount of rent for the simple reason that land in Russia was not sufficiently scarce relative to labor, and ironically, was made even less scarce by Russian conquests. The scarce factor of production was not land but labor. Hence it was the ownership of peasants and not of land that could yield an income to the servitors or to any non-working landowning class. A simple economic model may sharpen the argument (if any sharpening is needed) and help to develop it further. Assume that labor and land are the only factors of production (no capital or management), and that land of uniform quality and location is ubiquitous. No diminishing returns in the application of labor to land appear; both the average and the marginal productivities of labor are constant and equal, and if competition among employers raises wages to that level (as would be expected), no rent from land can arise, as Ricardo demonstrated some time past. In the absence of specific governmental action to the contrary (see below), the country will consist of family-size farms because hired labor, in any form, will be either unavailable or unprofitable: the wage of a hired man or the income of a tenant will have to be at least equal to what he can make on his own farm; if he receives that much, no surplus (rent) will be left for his employer. A non-working class of servitors or others could be supported by the government out of taxes levied (directly or indirectly) on the peasants, but it could not support itself from land rents. As a step toward reality, let us relax the assumption of the ubiquity of uniform land, and let capital (clearing costs, food, seeds, livestock, structures and implements) and management be included among the factors of production. Owners of capital, of superior skill and of better-than-average land will now be able to pay a hired man his due (or to use a tenant) and still obtain a surplus. But so long as agricultural skills can be easily acquired, the amount of

The Causes of Slavery or Serfdom

45

capital for starting a farm is small, and the per capita income is relatively high (because of the ample supply of land), a good worker should be able to save or borrow and start on his own in time. Most of the farms will still be more or less family-size, with an estate using hired labor (or tenants) here and there in areas of unusually good (in fertility and/or in location) land, or specializing in activities requiring higher-than-average capital intensity, or skillful management. But until land becomes rather scarce, and/or the amount of capital required to start a farm relatively large, it is unlikely that a large class of landowners, such as required by the Muscovite government, could be supported by economic forces alone. The American North in the Colonial period and in the nineteenth century would be a good example of an agricultural structure of this type. So far the institutional structure has been shaped by economic forces alone without direct interference by the government.2 Suppose now that the government decides to create, or at least to facilitate the creation of, a non-working class of agricultural owners. As a first step, it gives the members of this class the sole right of ownership of land. The peasants will now have to work for the landowners, but so long as the workers are free to move, competition among the employers will drive the wage up to the value of the. marginal product of labor, and since the latter is still fairly close to the value of the average product (because of the abundance of land) little surplus will remain. The Russian situation prior to the peasants’ enserfment corresponds to this case. The next and final step to be taken by the government still pursuing its objective is the abolition of the peasants’ right to move. With labor tied to land or to the owner, competition among employers ceases. Now the employer can derive a rent, not from his land, but from his peasants by appropriating all or most of their income above some subsistence level.3 That Russian serfs could stay alive, and even to multiply, while working for themselves half-time and less suggests that the productivity of their labor (with poor technique, little capital, but abundant land) must have been quite high. To recapitulate, the strong version of this hypothesis (without capital, management, etc.) asserts that of the three elements of an agricultural structure relevant here—free land, free peasants, and non-working landowners— any two elements but never all three can exist simultaneously. The combination 2  I mean by the “government” any organization capable of maintaining some measure of law and order and particularly of using non-economic compulsion. It can be a king, an assembly of landowners, a magnate, etc. 3  He may be restrained by custom and by the fear that his serfs can run away—a common occurence in Russia.

46

Domar

to be found in reality will depend on the behavior of political factors— governmental measures—treated here as an exogenous variable. The presence of this exogenous political variable seriously weakens the effectiveness of my model: it makes the presence of free land by itself neither a necessary nor a sufficient condition for the existence of serfdom. It is not a necessary condition because so long as marginal productivity of labor is high, serfdom may continue to exist even if free land is no longer present; it may even be imposed at this stage, as it was in the Russian Ukraine in the eighteenth century. Free land is not a sufficient condition because, as I stated above, without proper governmental action free land will give rise to free farmers rather than to serfs. For the same reasons the model cannot predict the net effect of a change in the land/labor ratio on the position of the peasants. Suppose that with constant land, technology, and per capita stock of capital, population increases. The economic position of the peasants will worsen (even serfs can be exploited more), but the landowners will be less inclined to interfere with the peasants’ freedom. Let population decline instead. The peasants will be better off provided they do not become less free. Thus a change in the land/labor ratio can set in motion economic and political forces acting in opposite directions. The strength and usefulness of the model could be increased by making the political variable endogenous. But this I cannot do without help from historians and political scientists. These difficulties notwithstanding, I would still expect to find a positive statistical correlation between free land and serfdom (or slavery). Such a correlation was indeed found by H. J. Nieboer of whom you’ll hear more in Part III. What about the end of serfdom (or slavery)? Traditionally it was assumed that it would or did disappear because of the inherent superiority of free labor. This superiority, arising from the higher motivation of the free man, was supposed to increase with greater use of capital and with technological progress. Let us disregard the possibly greater reliability of the slave and the longer hours he may be forced to work (particularly in traditional societies where leisure is highly valued), and let us assume that the economy has reached the position where the net average productivity of the free worker (Pf) is considerably larger than that of a slave (P8). The abolition of slavery is clearly in the national interest (unless the immediate military considerations, such as of the Muscovite government, overwhelm the economic ones), but not necessarily in the interest of an individual slave owner motivated by his profit and not by patriotic sentiment. He will calculate the difference between the wage of a free worker (Wf) and the cost of subsistence of a slave (W8) and will refuse to free his slaves

The Causes of Slavery or Serfdom

47

unless Pf —P8 > Wf —W8, all this on the assumption that either kind of labor can be used in a given field.4 As the economy continues to develop, the difference Pf – P8 can be expected to widen. Unfortunately, the same forces—technological progress and capital accumulation—responsible for this effect are apt to increase Wf as well, while W8 need not change. We cannot tell on a priori grounds whether Pf —P8 will increase more or less than Wf —W8. Therefore we cannot be sure that technological progress and greater use of capital necessarily reduce the profitability of slave as compared with free labor. Much will depend on the nature of technological progress. Thus Eli Whitney’s gin greatly increased the profitability of slavery, while a transition from raising crops to breeding sheep in medieval England might have acted in the opposite direction by creating a surplus of workers. (See Part II.) American planters must have used better agricultural techniques and more capital than their Latin-American and particularly Russian colleagues, but the Americans defended slavery with much greater zeal.

4  Actually, it is not easy to compare the relative profitability of free and slave labor. Since the free worker is paid more or less concurrently with his work, while a slave must be either reared or purchased, and may have children, etc., the streams of receipts and expenditures from the two kinds of labor must be properly discounted. It is assumed in the text that all indirect costs of using slaves, such as medical expense, extra supervision, etc., are included in W8.  In a well-organized slave market, the price of a slave will approximate the present value of his discounted net lifetime marginal product. A buyer who pays this price will discover that he will earn not much more than the going rate of interest; he will complain about the high cost of slaves and express doubt regarding the profitability of slavery in general, because at the margin he will be fairly indifferent between employing free or slave labor. But so long as the supply of food and of similar items for the maintenance of slaves is elastic (which it is likely to be), the slave-breeder should do very well. He benefits from the chronic perpetual disequilibrium in the slave market created by the abundance of land and by the limited human capacity to procreate (assuming no importation of slaves). But if the slave-breeder computes his rate of return on the current value of his slaves and land, he may not record much more than the market rate of interest either. In other words, the market mechanism transforms the profit from slaves into capital gains.  On this see Lewis Cecil Gray, History of Agriculture in the Southern United States to 1860, published in 1933 and reproduced in part in Harold D. Woodman, Slavery and the Southern Economy: Sources and Readings (New York: Harcourt, Brace & World, Inc., 1966), pp. 106– 09, and Alfred H. Conrad and John R. Meyer, The Economics of Slavery and Other Studies in Econometric History (Chicago: Aldine Publishing Company, 1964), pp. 43–92.

48

Domar

In a traditional society without technological progress and capital accumulation, the end of slavery is, paradoxically, more certain. As population continues to increase and the society eventually becomes Malthusian, the marginal product of labor descends to the subsistence level. Now the free man costs little more to employ than the slave, while, hopefully, being less bothersome and more productive. The ownership of human beings becomes pointless because of the great multiplication of slaves, and they become free provided they stay poor.5 It is land that becomes valuable, and rents collected from estates worked by free laborers or tenants without any non-economic compulsion are sufficient to support an army of servitors or idlers. If the Muscovite government could have only waited a few hundred years!

II

Where I come from, an economic model without empirical testing is equated with a detective story without an end. My attempts to test the present model, however, merely taught me that the job is not for the amateur. I shall report to you the results of my skin deep investigation in the hope that my mistakes will stimulate the specialists. I concentrate on the Russian case, with short excursions into the histories of Poland-Lithuania, Western Europe and the United States. 1. Russia. The phenomenon to be explained here is not only the development of serfdom but its particular timing: before 1550 Russian peasants were free men; a hundred years later they were serfs. The relevant variables are: (1) the number of servitors required by the military needs of the Moscow state, and (2) the population density. According to Kliuchevsky, prior to the middle of the fifteenth century, Moscow, still a Tatar vassal surrounded by other Russian lands, fought very few foreign wars; its population became dense because Moscow was the safest spot in the area with few outlets for emigration.6 We may conclude that there was no need 5  It is possible that even in a Malthusian society slavery (or serfdom) may linger on. Slaves may be kept for reasons of social prestige (a relic from the times when slavery was profitable), or simply because a slave is more reliable than a hired man. On the other hand, the use of a tenant (with a limited lease) or of a hired man allows the landowner to choose the best among several applicants with much greater ease than among slaves or serfs protected by custom. 6  Kliuchevsky, Vol. I, p. 379; Vol. III, pp. 9–10, 121. Blum, however, talks about depopulation already in the fourteenth and fifteenth centuries. See Jerome Blum, Lord and Peasant in

The Causes of Slavery or Serfdom

49

as yet for a large class of servitors, and that the landowners could derive rents from their estates (patrimonies, to be exact) without enserfing the peasants. It is true that Russia, from the Kievan times onward, always had a substantial number of slaves. At the time, these were mostly household servants and retainers rather than peasants.7 From the middle of the fifteenth century the situation changes drastically. Having become independent from the Tatars (officially in 1480, actually earlier), and having gathered a number of Russian lands, Moscow was confronted with powerful enemies: with Poland-Lithuania and Sweden in the west and northwest, and with the Crimean Tatars in the south. The struggle with the latter went on continuously, while 50 out of the 103 years from 1492 to 1595 were spent in wars against Poland-Lithuania and Sweden, as were the following 30 out of 70 years from 1613 to 1682, not to mention the Time of Troubles, 1598–1613, filled with both civil and foreign wars.8 The military proficiency of the Muscovite armies being poor, refuge was sought in large numbers. More than 300,000 men were reported to have been under arms during Ivan the Terrible’s Livonian War. There must have been a great increase in the number of servitors. With trade and industry making no significant progress, the government had to assign land to them. This process began on a large scale in the second half of the fifteenth century and was accelerated throughout the sixteenth century.9 In the meantime, the central areas of the country became depopulated. The conquest of the whole expanse of the Volga river (begun in 1552) opened up large areas of better soil and attracted large masses of peasants fleeing from high taxes, Ivan the Terrible’s oppression (the famous oprichnina) and Crimean invasions. And then came the Time of. Troubles which devastated the country once more. Already in the sixteenth century there was fierce competition for peasant hands among the landowners. It must have intensified after 1613.10 Thus both ingredients for the development of serfdom—a high land/labor ratio and the government’s determination to create a large class of servitors— were present. In addition, there were several other forces working in the same Russia from the Ninth to the Nineteenth Century (Princeton: Princeton University Press, 1961), pp. 60–61. It is possible that Kliuchevsky describes the relative position of Moscow among other Russian lands, while Blum refers to the whole country. 7   Kliuchevsky, Vol. I, pp. 282–83; Vol. II, pp. 182–83. 8   Ibid., Vol. II, pp. 121, 125, 221–22; Vol. III, p. 135. 9   Ibid., Vol. II, pp. 221, 229–42, 248; Vol. III, pp. 63–64, 230–31, 257, 283. Blum, pp. 93, 157. 10  Kliuchevsky, Vol. II, pp. 254–57, 339–44; Vol. III, pp. 182, 244. Blum, pp. 147, 152–54, 157, 160, 252. B. D. Grekov, Krest’iane na Rusi s drevneǐshikh vremen do XVII veka (MoscowLeningrad: Izdatel’stvo Akademii Nauk SSSR, 1946), pp. 794–96, 849.

50

Domar

direction. The first was the decline in the power of the great magnates, both at the hands of Ivan the Terrible and during the Time of Troubles. By offering the peasants privileges and protection, these magnates had been quite successful in bidding the peasants away from the servitors; for this reason the magnates favored the free movement of peasants, while the servitors, quite naturally, opposed it. Now the peasants lost the support of their “friends.”11 The second reason lay in the fiscal interest of the state: peasant migrations, particularly from the center to the periphery of the state, disorganized tax collections.12 And finally, the peasant communities objected to the emigration of their members because the community carried a collective responsibility for the tax liabilities of its members (until in later years this responsibility was taken over by the masters); the departure of several members would leave the rest overburdened until the next census.13 Space does not allow me to give additional details of the process which gradually enserfed the peasants, or to discuss the disagreement between Kliuchevsky, who emphasized the hopeless indebtedness of the peasants to their landlords as the main obstacle to their movement, and Grekov and Blum who put greater stress on legislative enactments (particularly on the so-called “Forbidden Years,” zapovednye gody).14 Let me mention instead two further reflections of the scarcity of labor in Russia: the first manifested itself in the replacement of the basic land tax by a household tax in the seventeenth century, and by a poll tax under Peter the Great.15 The second is an interesting cultural trait which remained long after its cause had probably disappeared: as late as in the first half of the nineteenth century, the social position of a Russian landowner, as described in contemporary literature, depended less on the size of 11  Kliuchevsky, Vol. II, pp. 259, 307. Blum, pp. 253–54. Grekov, pp. 870–71, 903, 909. Grekov, Glatmeǐshie etapy v istorii krepostnogo prava v Rossii (Moscow-Leningrad: Gosudarstvennoe sotsial’no-ekonomicheskoe izdatel’stvo, 1940), p. 46.  It is interesting to note that when the leaders of the gentry militia were negotiating a treaty with the Polish king Sigismund regarding the accession of his son to the Moscow throne in 1610 and in 1611, they demanded the inclusion of a provision forbidding the movement of peasants. Kliuchevsky, Vol. II, p. 349. 12  Kliuchevsky, Vol. III, p. 188. 13  Ibid., Vol. II, pp. 317–18, 336–37, 340. Blum, pp. 96, 234. 14  Kliuchevsky, Vol. II, pp. 321–23, 331–50; Vol. III, pp. 181–88. Blum, pp. 254–55. Grekov, Krest’iane, pp. 826, 850. Grekov, Glavneǐshie, pp. 64–65.  If the peasants’ debts tied them to their lords as strongly and as hopelessly as Kliuchevsky asserts, it is puzzling that the government had first to limit and then to forbid their movement by law. 15  Kliuchevsky, Vol. III, pp. 243–46; Vol. IV, pp. 142–48. Grekov, Glavneǐshie, pp. 71–72.

The Causes of Slavery or Serfdom

51

his land holdings (which are seldom mentioned) than on the number of souls (registered male peasants) that he owned.16 2. Poland-Lithuania. On the theory that the length of a report should be proportional to the intensity of research done, this section will be very short. The relevant facts are as follows: (1) In the fourteenth century vast open and very sparsely populated territories in the Ukraine were conquered by the Lithuanians.17 (2) In the fifteenth and sixteenth centuries, Ukraine was repopulated by immigrants from the more central areas of the state. The migration depopulated the central areas to such an extent as to constitute, according to Grekov, a threat to the Polish state.18 (3) By the end of the sixteenth century, the peasants were enserfed.19 What is not clear to me is the time sequence of events (2) and (3). In Vol. III (p. 110), Kliuchevsky dates the repopulation of the Ukraine in the sixteenth century; in Vol. I (p. 293), in the fifteenth century. But in both places he attributes the migration of peasants to the intensification of serfdom in PolandLithuania. Polish serfdom, according to him, had been established already in the fourteenth century, and Lithuanian, in the fifteenth century.20 On the other hand, Grekov asserts that according to the Polish constitution of 1493, each peasant could still leave the land, having settled accounts with his landlord. But he also reports that in 1444 the Galician gentry demanded that the government prevent other landlords from interfering with the peasant movements.21 Evidently, such interference was taking place even then. 16  Here are a few examples: In Pushkin’s Dubrovsky, the old Dubrovsky is identified as the owner of seventy souls, and Prince Vereisky, of three thousand; in The Captain’s Daughter, the commandant’s wife is impressed by Grinev’s father’s ownership of three hundred souls; in Gogol’s The Dead Souls, Pliushkin owns more than a thousand souls; in Goncharov’s Oblomov, the principal hero owns three hundred and fifty; in his A Common Story, a certain Anton Ivanich has twelve, mortgaged over and over again…. 17  Kliuchevsky, Vol. I, p. 293. 18  Ibid., Vol. I, pp. 293–94. Grekov, Krest’iane, p. 387. 19  Jerome Blum, “The Rise of Serfdom in Eastern Europe,” American Historical Review, LXII (1957), pp. 807–36. See particularly pp. 821–22. 20  Kliuchevsky, Vol. III, pp. 101–02. 21  Grekov, Krest’iane, pp. 381–83. There seems to be considerable disagreement among the authorities he cites. He mentions a number of legislative enactments passed at the end of the fifteenth century and in 1510, 1519, 1520, 1532 limiting the freedom of peasants to move (p. 387).

52

Domar

In Poland-Lithuania great gaps between legal enactments and the actual state of affairs were quite possible. There were probably considerable regional variations, both in law and in practice as well. I would be happier if it could be established that migration to the Ukraine preceded the development of serfdom, but I am certainly not in a position to settle the matter. It is quite possible that migration and serfdom were reinforcing each other. Since I have not studied the development of serfdom in other East European countries, I can make only two brief comments on Blum’s well-known and very interesting article on “The Rise of Serfdom in Eastern Europe.” His stress on the increasing power of the nobility and on the general depopulation of the area “from the Elbe all the way across to the Volga …” is heartily welcome.22 But his use of alternating periods of prosperity and depression as important causes of the rise and decline of serfdom cannot be evaluated until he presents an analytical explanation of the causation involved. 3. Western Europe. We shall deal here very briefly with four events: (1) (2) (3) (4)

The emergence of serfdom in the late Roman Empire The decline of serfdom by 1300 Its non-recurrence after the Black Death The relationship between sheep breeding and serfdom.

The depopulation of the late Roman Empire is, of course, well known. Referring to Byzantium, Georg Ostrogorsky states: “And so ever-increasing masses of the rural population were tied to the soil. This is a particular instance of the widespread compulsory fastening of the population to their occupation which scarcity of labour forced the later Roman Empire to pursue systematically.”23 This is the clearest statement on the relation between scarcity of labor and the development of serfdom that I have come across in my reading of European economic history. Similarly, the great increase in population in Western Europe by the end of the thirteenth century when serfdom was declining is also well known. Thus Ganshof and Verhulst talk about “… a considerable and growing reserve of surplus labor …” in France, and Postan discusses signs of overpopulation 22  Blum, “The Rise of Serfdom,” p. 819. 23  Georg Ostrogorsky, “Agrarian Conditions in the Byzantine Empire in the Middle Ages,” The Cambridge Economic History of Europe, Second Edition (Cambridge: Cambridge University Press, 1966), I, 206. See also pages 11, 27–28, 33, 66 and 257 of the same volume. Also, W. R. Brownlow, Lectures on Slavery and Serfdom in Europe (London and New York: Burns and Oates, Ltd., 1892), pp. 49–50.

The Causes of Slavery or Serfdom

53

in England: a growing number of wholly landless men, sub-holdings of many tenants, shortage of pasture, etc.24 The same information for Western Europe in general is supplied by Smith, who adds that: “The problem therefore for western landowners, at any rate before the demographic collapse of the midfourteenth century, was not to keep tenants, but how to get the most out of them.25 Since these facts fit my hypothesis so nicely, let me stop here while I am still winning. But when we come to the depopulation caused by the Black Death after 1348 (though, according to Postan, English population stopped growing even earlier),26 my hypothesis is of little value in explaining the subsequent course of events. (See Part I.) Why did serfdom fail to come back after such a sharp increase in the land/labor ratio? I address myself only to England. Except for one rather queer economic explanation to be discussed presently, I have none to offer and have to fall back on political factors. Serfdom could not be restored unless the landowners were reasonably united in their pressure on the government, and unless the latter was willing and able to do their bidding. But it is most unlikely that every estate lost the same fraction of its peasants. Hence, those landowners who had suffered most would welcome the freedom of peasant movement, at least for a while, while those who had suffered least would oppose it. If so, the landowners could not be united. Postan also suggests the probability that the main pressure behind Richard II’s legislation came not from feudal landowners, but from smaller men;27 English magnates, like their Russian colleagues (see above), could evidently take care of their own interests. Though I cannot judge the “spirit” of medieval legislation, it seems to me that the measures undertaken by Richard’s government were somewhat halfhearted.28 In any case, they were ineffective. So economic forces could reassert themselves and help the peasants. The queer economic explanation which I have just mentioned would delight an economist if only it squared with facts. It is the expansion of sheep 24  François Louis Ganshof and Adriaan Verhulst, “Medieval Agrarian Society in its Prime: France, The Low Countries, and Western Germany,” Cambridge Economic History, I, 294; M. M. Postan in his essay on “England,” same volume, pp. 552–56, 563–64, 624; Blum, “The Rise of Serfdom,” pp. 810–11. 25  R. E. F. Smith, The Enserfment of the Russian Peasantry (Cambridge: Cambridge University Press, 1968), p. 4. 26  Postan, essay on “England,” Cambridge Economic History, I, 566–70. 27  Ibid., p. 609. 28  Brownlow, Lectures on Slavery, pp. 157–83. Smith, Enserfment, pp. 4–5.

54

Domar

breeding, an activity which is land-using and labor saving.29 Unfortunately such data as I could find do not support the contention that there was an expansion of sheep breeding in the hundred years following the Black Death. The legal exports of English wool, in raw and in cloth, fell from 12 million pounds in 1350 to 8.7 million in 1400—a drop of 27 percent. Another fall of 12 percent (of the 8.7 million) took place by 1450.30 My authorities do not state the proportions of wool consumed at home and smuggled out of the country.31 Perhaps these were affected by the Hundred Years’ War. But as things stand, I certainly cannot claim that an expansion of English sheep breeding took place after 1350 and that it helped to save the peasants from the return of serfdom.32 Judging by Thomas More’s famous passage about sheep devouring men, by Bishop Latimer’s “Sermon of the Plough” (1549), and by other more direct evidence, there must have been considerable expansion of sheep breeding at the expense of crops and of people in the sixteenth century.33 By that time, however, English peasants hardly needed the help from the sheep in staying free. But is it possible that the early expansion of sheep breeding which must have taken place sometime prior to 1350 had helped the English serfs to gain their original freedom after all? 4. The United States. The American South fits my hypothesis with such embarrassing simplicity as to question the need for it. The presence of vast expanses of empty fertile land in a warm climate, land capable of producing valuable products if only labor could be found seems to me quite sufficient to explain the importation of slaves. What is not clear to me is the failure of the North to use them in large numbers. Besides social and political objections, there must have been economic reasons why Negro slaves had a comparative 29  The idea that sheep-breeding may have had something to do with serfdom was suggested by Nieboer in his book (pp. 371–75) discussed in Part III. 30  K. G. Ponting, The Wool Trade Past and Present (Manchester and London: Columbine Press, 1961), p. 30. The figures are based on a chart facing p. xviii of Medieval Merchant Ventures by E. Carus Wilson. 31  According to Postan, p. 568, domestic consumption of cloth is not known. Peter J. Bowden arbitrarily assumed it to be 50 percent. See his The Wool Trade in Tudor and Stuart England (London: Macmillan & Co., Ltd., 1962), p. 37. 32  Data on the size of the sheep population, or more correctly on increments in it, would not be sufficient for our problem. We would have to know how many cropraising peasants were replaced, say, by 1,000 extra sheep. 33  See E. Lipson, The History of the Woollen and Worsted Industries (London?: Frank Cass & Co., Ltd., 1965), p. 19; E. Nasse, On the Agricultural Community of the Middle Ages, and Inclosures of the Sixteenth Century in England (London: Macmillan & Co., 1871), pp. 77–78; Brownlow, Lectures on Slavery, p. 184; Bowden, Wool Trade, p. xvi.

The Causes of Slavery or Serfdom

55

advantage in the South as contrasted with the North. Perhaps it had something to do with the superior adaptability of the Negro to a hot climate, and/ or with his usefulness in the South almost throughout the year rather than for the few months in the North.34 I have a hard time believing that slaves could not be used in the mixed farming of the North; much food was produced on southern farms as well, most of the slave owners had very few slaves, and many slaves were skilled in crafts.35 A study of the possible profitability of slavery in the North, along Conrad and Meyer’s lines, which could show whether the North could have afforded paying the market price for slaves, would be most welcome. I have not come across any good evidence that slavery was dying out in the United States on the eve of the Civil War, and I side here with Conrad and Meyer, though, in truth, I am not sure that such a thorough investigation was required to prove the profitability of slavery in the South.36

III

In conclusion, let me say a few words about the origin of my hypothesis and about its place in economic history. Although I had discussed it in my classes for a good dozen years, I did not write it up until 1966 because I had been told on good authority that the idea was old and well known. My source was indeed correct because a brief search in the library revealed quite a few predecessors. The most important of them was the Dutch scholar Herman J. Nieboer whose magnum opus of 465 pages under the title of Slavery as an Industrial System: Ethnological Researches was published in 1900.37 The hypothesis which I have immodestly called “mine” was stated by him time and again, and tested against

34  Woodman, Slavery and the Southern Economy, p. 7. 35  Conrad and Meyer, Economics of Slavery, p. 80; James Benson Sellers, Slavery in Alabama (University, Alabama: University of Alabama Press, 1950), pp. 71, 120, 162–63; Rosser Howard Taylor, Slaveholding in North Carolina: An Economic View (Chapel Hill: University of North Carolina Press, 1926), p. 72; Harrison Anthony Trexler, Slavery in Missouri (Baltimore: The Johns Hopkins Press, 1914), pp. 13, 19; Woodman, Slavery and the Southern Economy, pp. 14–15. 36  As the authors practically admit on p. 78. On the profitability debate see Stanley L. Engerman, “The Effects of Slavery Upon the Southern Economy: A Review of the Recent Debate,” Explorations in Entrepreneurial History, Second Series, IV (1967), pp. 71–97. 37  It was published in The Hague by Martinus Nijhoff. A republication is scheduled in 1970 by Burt Franklin, Publisher, New York.

56

Domar

a mass of anthropological and historical data. As you might expect, he was satisfied with his results. But the hypothesis was not really original with Nieboer. He in turn referred to A. Loria’s Les Bases Economiques de la Constitution Sociale of 1893, and to E. G. Wakefield’s A View of the Art of Colonization published in 1834. Some glimpses can be found even in Adam Smith’s The Wealth of Nations.38 I have two disagreements with Nieboer. First, his definition of free land has too much legal and not enough economic content to my taste, though he seems to have been unclear rather than wrong. Second, he exaggerated the importance of the hypothesis by claiming, though not in so many words, that free land or other free resources are both necessary and sufficient for the existence of slavery or serfdom: “… Only among people with open resources can slavery and serfdom exist, whereas free labourers dependent on wages are only found among people with close resources.”39 He protected himself with a note on the same page by excluding simple societies of hunters, fishers, and hunting agriculturists, hardly a fit company for the farmers of the American North, He disregarded the possibility that serfdom, once established, could exist for a long time after its initial cause—free land—had disappeared, or that serfdom may be even introduced in the absence of free land. He ignored the role of government. These, however, are minor defects in an important major contribution. On the other hand, my source may have been a bit wrong. If historians have always known about the relation between the land/labor ratio and serfdom (or slavery), they must have tried hard not to scatter too many good, clear statements in places where I could find them, though the students of the American South have been much kinder to me than others.40 Nieboer could also lodge some complaints. His name can be found neither in the bibliography nor in the index of the 1966 edition of the first volume of The Cambridge Economic History of Europe. And it is absent from Blum’s classic study of Russian serfdom. I did find Nieboer’s name in Genovese’s The Political Economy of Slavery in connection with some insignificant point, but with a further notation that 38  Adam Smith, The Wealth of Nations (London: Cannan’s edition, 1922), II, 66-68. There is another book by Wakefield on the-same subject: England and America: A Comparison of the Social and Political State of Both Nations (London: Richard Bentley, 1833), Vol. II. Other sources: J. E. Cairnes, The Slave Power (London: Parker, Son, and Bourn, 1862); J. S. Mill, Principles of Political Economy, 1848 (New York: D. Appleton and Co., 1920), I, 316. 39  Nieboer, Slavery as an Industrial System, pp. 312, 389. 40  A clear statement by Ostrogorsky was quoted in Part II. For the American views, see Woodman’s collection.

The Causes of Slavery or Serfdom

57

“Phillips read and referred to this book.” Phillips had read it, and confirmed that “hired labor was not to be had so long as land was free.”41 Perhaps in history this hypothesis occupies a place similar to that enjoyed by economic growth in economic theory not long ago. That place was once described as “always seen around but seldom invited in.” If so, why not invite it? After all, the land/labor ratio is readily quantifiable.

41  Eugene D. Genovese, The Political Economy of Slavery (New York: Vintage Books, 1967), p. 84. Ulrich B. Phillips, “The Economic Cost of Slaveholding in the Cotton Belt,” Pol. Sci. Q., XX (June 1905), partially reproduced in Woodman, Slavery and the Southern Economy, p. 36.

The Emergence of a Slave Society Moses I. Finley Fustel de Coulanges began his remarkable study of the Roman colonate in this way: ‘The colonate is one of the most obscure institutions of the Roman Empire…. Slavery is more easily explained.’ He was wrong about the latter. No doubt, as he continued, slavery ‘was a primordial fact, contemporary with the origin of society; it had its roots in an age of the human species when all inequalities had their raison d’être.’1 But the Greeks and Romans, independently so far as we can see, transformed this ‘primordial fact’ into something new and wholly original in world history (and something rare throughout history), namely, an institutionalized system of large-scale employment of slave labour in both the countryside and the cities; in Marxist language, ‘the slave mode of production was the decisive invention of the Graeco-Roman world’.2 That invention is not ‘easily explained’. We must begin with a crude, but fundamental and indeed commonplace, distinction between labour for oneself and labour for others. ‘Oneself’ is to be understood not in a narrow individualistic sense but as embracing the family, nuclear or extended as the case may be in any particular society. That is to say, the work of the women and children within the family, no matter how authoritarian and patriarchal its structure, is not subsumed under this category of labour for others (though I am aware that I face objections from several directions when I say that). Nor is interfamily cooperative activity, as during harvest periods. ‘Labour for others’ implies not only that ‘others’ take some of the fruits but also that they customarily control, in direct ways, the work that is done and the manner of its doing, whether in person or through agents and managers. Neither the independent peasant nor the tenant-farmer normally satisfies the second of those conditions, though he pays taxes or rents or both, and may be restricted in various ways by public law. That is again a commonplace, but I must insist on it at the outset because failure to draw so elementary a distinction has plagued much of the debate about ancient slavery in the past century. S ource: Finley, Moses I., “The Emergence of a Slave Society,” in Moses I. Finley, Brent D. Shaw (ed.), Ancient Slavery and Modern Ideology, Princeton, NJ: Markus Wiener Publishers, 1998, 135–160. 1  Fustel de Coulanges (1885) 3. 2  Anderson (1974a) 21. © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_005

The Emergence Of A Slave Society

59

The need to mobilize labour-power for tasks that are beyond the capacity of the individual or the family reaches back into prehistory. Such a need was present whenever a society attained a stage of sufficient accumulation of resources and power in some hands (whether king, temple, ruling tribe or aristocracy). And the requisite labour force was obtained by compulsion—by force of arms or by force of law and custom, usually by both together—for all purposes (or interests) not amenable to straightforward cooperation: in agriculture or mining or public works or arms manufacture. Compulsory labour takes a considerable variety of forms, today as in the past3—debt bondage, clientship, peonage, helotage, serfdom, chattel slavery, and so on. But whatever the form, the compulsion is fundamentally different from that lying behind hired labour, which implies the conceptual abstraction of a man’s labour power from the man himself. The wage-labourer also surrenders some of his independence when he accepts employment, but such loss cannot be classed with that suffered by slaves or serfs. In early societies, free hired labour (though widely documented) was spasmodic, casual, marginal. Significantly, in neither Greek nor Latin was there a word with which to express the general notion of ‘labour’ or the concept of labour ‘as a general social function’.4 Only with the development of capitalism did wage labour emerge as the characteristic form of labour for others. Labour power then became one of the main commodities in the market-place. With slavery, in contrast, the labourer himself is the commodity. The slave is in that respect unique among types of labour despite overlapping with, for example, the most oppressive kinds of serfdom or with convict labour.5 The slave and the free wage-labourer thus stand at the extreme poles of labour for others, but historically the important contrast is rather between slaves and other types of compulsory labour. As institutionalized systems of organizing labour, other kinds of involuntary labour preceded chattel slavery, and both preceded (and then coexisted with) free hired labour. In order to understand ancient slavery, therefore, some preliminary consideration is required of the labour systems within which it arose and which it largely displaced in key areas of the classical world, though by no means in all.

3  On the modern varieties, see the thirteen case-studies in W. Kloosterboer, Involuntary Labour after the Abolition of Slavery (Leiden 1960). 4  See J-P. Vernant, Mythe et pensée chez les Grecs (Paris 1965) pt. 4. Cf. Y. Garlan, in Garnsey (1980), on the relatively late emergence of both hired labour and the free peasant-owner in Greek history. 5  For the present I treat the slave as an ideal type. We shall see later in the chapter how much differentiation there was in reality within the slave population.

60

Finley

It must be acknowledged at the outset that in current historical and sociological writing the classification of types of labour is in a bad state. Behind faulty classification there is of course faulty theory, or at least inadequate conceptualization. Only a few years ago Meillassoux complained, ‘In the present state of research, there is in fact no general theory which permits us to identify slavery or the objective bases of its coming into existence (existence eventuelle)…. No formal criterion has been brought to light that permits a categorical distinction between slaves and all other components.’6 At one extreme, marginal distinctions are insisted on so fiercely that all institutions are reduced to an infinity of discrete instances, stultifying any possibility of analysis or understanding. Thus Lauffer assured us at the Stockholm Historical Congress that we must not translate the Greek doulos or the Latin servus as ‘slave’ because that word is too reminiscent of modern Negro slavery, whereas ‘the ancient “slave” is a wholly different social type’ (though he never told us why).7 At the other extreme there is the tendency to create a ‘suprahistorical mélange’ that ‘defies all scientific principles’.8 One variant, common among Anglo-American anthropologists, goes like this (and it is by no means restricted to slavery). First, a host of, let us say, African statuses and statusterms are translated as ‘slaves’; second, it is observed that at essential points these so-called slaves are extremely unlike the slaves of classical antiquity or of the Americas; third, instead of reconsidering their appellation ‘slaves’ to their own subjects, these anthropologists angrily protest the ‘ethno-centrism’ of ‘western’ historians and sociologists and demand that the latter redefine and reclassify slaves in order to provide a place for their own pseudo-slaves.9 An even worse situation prevails with regard to the other forms of compulsory labour that have appeared throughout history. It is a sign of our grave difficulty in comprehension that we are unable even to translate the labels into modern western languages: ‘helot’ is not a translation but an adoption; ‘debtbondsman’ is an artificial coinage; pelatai, laoi, clientes, coloni have not even been adopted but merely transliterated.

6  Meillassoux (1975) 20. 7  Lauffer (1960) 81. 8  Anderson (1974b) 486. 9  A complete example is provided by S. Miers and I. Kopytoff, in the introduction to Slavery in Africa: Historical and Anthropological Perspectives (Madison 1977), esp. pp. 5–6, 11, 76–8. For a correct approach, see J. Bazin, ‘Guerre et servitude à Ségou’, in Meillassoux (1975), not shared by all the contributors to the volume; P. Hill, ‘From Slavery to Freedom: The Case of FarmSlavery in Nigerian Hausaland’, Comparative Studies in Society and History 18 (1976) 395–426.

The Emergence Of A Slave Society

61

When I say that we cannot translate these terms I do not imply that historians do not translate them all the time, writing about the Graeco-Roman world and even more about the ancient Near East. The magnetism of the traditional tripartite division of labour into slave-serf-free appears to be irresistible. Procedures vary. The most common is to label everyone who is not obviously a slave or a free man a serf—the helots of Sparta, the penestai of Thessaly, the laoi of Hellenistic and Roman Asia Minor, and the varying dependents who made up the bulk of the population of Mesopotamia. The exotic feudalisms that so irritated Marc Bloch10 have proliferated considerably since his day, only to reach what has been called ‘the impasse of a quasi-universal feudalism’.11 One current Marxist school has taken another tack. In a recent article with the revealing title, ‘Slaves, Helots and Serfs in Early Antiquity’, the distinguished Leningrad Assyriologist Diakonoff asserts that ‘there is no difference between these two types of workers [slaves and helots/serfs] in regard to the main points, viz. both were exploited by extra-economic coercion, and both were entirely devoid of property in the means of production’. Underlying that conclusion, factually incorrect in part, there is a desperate rear-guard action designed to ‘save the phenomena’ of the Engelsian unilinear scheme. ‘All ancient societies’, says Diakonoff, ‘in Europe as well as on all other continents’, belong to ‘a typologically identical socio-economic formation’, sharing ‘a common system of production typical of antiquity’.12 Not surprisingly, attempts at classification, good or bad, turn out to depend on underlying theoretical or ideological considerations. Whereas Lauffer defends the humanist valuation of classical society by insisting on the uniqueness of the ancient slave as a social type, Diakonoff and his school defend their version of Marxism by creating an ‘easy supra-historical mélange defying all scientific principles of classification’. In effect the latter fall back on a tautology: the slave is an instrument in the slave mode of production. However, to quote Meillassoux once more, ‘It is really not obvious that slavery is only a 10  M. Bloch, The Historians Craft, trans. P. Putnam (Manchester 1954), pp. 175–6. 11  Anderson (1974b) 484. 12  Diakonoff (1974) 63, 64, 78, respectively; cf. A. Jähne, ‘Zwei Tendenzen gesellschaftlicher Entwicklung im Hellenismus’, Klio 60 (1978) 137–50, at p. 140, and the reply by H. Kreissig, pp. 217–19. Diakonoff, whose pages on Greece and Rome reveal inadequate knowledge, surprisingly continues the equation Marxist=Soviet. His use of the term ‘social formation’ is significantly different from that of Anderson (1974a) 22n6, and his fundamental conception had been effectively demolished in advance by e.g. Hahn (1971), and earlier by K. Zelin, ‘Principes de classification morphologique des formes de dépendance’, in Annequin (1978) 45–77, originally published in Russian in Vestnik Drevnei Istorii (1967) no. 2, pp. 7–30.

62

Finley

relationship of production.’13 Whether it is or not is a matter for demonstration, not an axiomatic premise. One fact is at least indisputable, that chattel slavery existed as a major institution in such different social formations as the Roman Empire and nineteenth-century America. That all forms of involuntary labour can formally be classed in a single category is self-evident. But is that a useful classification? Do extra-economic compulsion and lack of ownership of the means of production exhaust the ‘main points’, as Diakonoff asserts? Are the self-evident differences among the various kinds of compulsory labour unimportant? A few examples will, in my view, suggest the inadequacy of Diakonoff’s simple scheme as a tool of historical analysis.14 It can scarcely be disputed, first, that helots were ‘collective bondsmen’, that is to say, they were a whole population (or populations) subjected to bondage, whereas debt-bondsmen and slaves fell into bondage individually, one by one. That distinction holds even for the hundreds of thousands of captives sold off by Julius Caesar or for the shiploads of African captives transported to the Americas: their fate was an individual one, not a collective one. It is equally certain, secondly, that all categories of compulsory labour other than chattel slaves possessed, in different degrees, restricted rights of property and usually much wider rights in the sphere of marriage and family law. These rights were de iure in at least some societies, well documented in the ‘law code’ of Gortyn in Crete, perhaps only de facto in other communities, though that suggestion may be grounded on nothing more than the paucity of evidence. Either way there were significant consequences: helots and clientes and the rest were selfreproducing, unlike the chattel slave populations, requiring no reinforcement from outside in order to maintain the requisite numbers; and they were looked upon and feared by their masters, with good reason, as potentially rebellious as a group, I might almost say as a subject community.15 The debt-bondsmen of archaic Athens and Rome offer the extreme example (and there may have been similar dependent classes in other early communities about which we 13  Meillassoux (1975) 20. 14  See Finley (1960) (1964) (1965a) for a detailed account of what follows immediately, with the necessary documentation. 15  Ducat (1978) has denied all this in a lengthy and perverse article, replete with misstatements (particularly of the views of others) and omissions. Thus he writes (p. 22) that the chattel slaves and the helots differed ‘only in the fact that instead of belonging to an individual the latter belonged to a collectivity’. To reach that conclusion, he fails to mention the self-reproduction of the helots or to consider the implications of their right to a formally defined share of the produce.

The Emergence Of A Slave Society

63

are uninformed). They succeeded in freeing themselves en bloc and thereby automatically reestablishing themselves as full members of their respective communities. That was civil strife, a conflict within the community, not a slave revolt: the aims of the latter were to emancipate themselves as individuals, not to incorporate themselves into the masters’ community or to transform the social structure. In this context, it is also noteworthy that when the Messenian helots were freed (again en bloc) by the Thebans after they had defeated Sparta at Leuctra in 371 bc, the Messenians were at once accepted by the Greeks generally as a proper Greek community. Objection has been raised in the past to my stress on such distinctions among types of compulsory labour: the whole scheme is only a ‘juristic abstraction’, or a mere description of institutions ‘without inquiring into their function’.16 True, no classification or taxonomy, no matter how detailed, is a sufficient account of the nature of a given society and its transformations. It can only be deemed to be more or less useful than competing classifications as an analytical tool in a particular inquiry. The question in the present context is whether or not such distinctions, the existence of which cannot be seriously denied, contribute substantially to an understanding of the emergence of a slave society, in other words, to the replacement of other forms of compulsory labour by chattel slavery where that occurred. In Diakonoff’s vague terminology, are they or are they not ‘main points’? I shall not try to answer on the level of grand theory, in a more or less metaphysical debate about the dynamic or dialectic of history.17 Instead, I shall proceed immediately to a more detailed examination of the slave and slavery. As a commodity, the slave is property. At least since Westermarck writing at the beginning of the present century, some sociologists and historians have persistently tried to deny the significance of that simple fact, on the ground that the slave is also a human being or that the owner’s rights over a slave are often restricted by the law.18 All this seems to me to be futile: the fact that a slave is a human being has no relevance to the question whether or not he is also property; it merely reveals that he is a peculiar property, Aristotle’s ‘property with a soul’ (Politics 1253b32). Reciprocally, the old Latin word erus also 16  Ducat (1978) 23 and Annequin (1975) 9, respectively. 17  Cf. ‘those explanations which try to reduce Marxist social theory to three factors in all (individual, community, means of production), in my opinion, cannot be anything else than fatal simplifications almost resembling cabbalistic number mysticism’: G. Komoróczy, ‘Landed Property in Ancient Mesopotamia and the Theory of the So-called Asiatic Mode of Production’, Oikumene 2 (1978) 9–26, at 10n3. 18  See Patterson (1977b) 431–2.

64

Finley

implies the peculiarity of a slave-property. Defined in the Oxford Latin Dictionary as ‘a man in relation to his servants, master’, erus was regularly and frequently used by slaves in the comedies of Plautus and even of Terence, in preference to dominus. Later poets continued to employ the word as an occasional archaism, but they extended it to mean also the owner of an animal or other property, thus sacrificing the original implication that the slave-master relationship was odd, indeed unique, among property relations.19 Legal restrictions on the rights of a slaveowner are also a side-issue: in modern sociological and juridical theories of any school, all property is understood to be a matrix of rights, rarely if ever unlimited. The precise rights that constitute the matrix vary with kinds of property and kinds of society. Property, in other words, is an historical category, and that is one more commonplace that I must regretfully enunciate in order to remove the confusions that still prevail on the subject. When Roman lawyers defined a slave as someone who was in the dominium of another, they used the quintessential property-term dominium.20 They were not dissuaded by the slave’s human quality (not even when they used the word homo to refer to a slave, as they did frequently).21 Nor were the millions of slaveowners who bought and sold slaves, overworked them, beat and tortured them, and sometimes put them to death, precisely as millions of horse-owners have done throughout history. Other millions did hot exercise their rights to the full in this way: that is interesting, even important, but it does not undermine the slave-property link conceptually. The failure of any individual slaveowner to exercise all his rights over his slave-property was always a unilateral act on his part, never binding, always revocable. That is a critical fact. So is its reverse, the equally unilateral, always revocable grant by a slaveowner of a specific privilege or benevolence. As for promises, one of Plautus’ slaves succinctly exposed their standing: no master can be brought to court over a promise to a slave (Persa 193–4). Even the act of manumission could be, and frequently was, qualified in a multitude of ways. Failure to appreciate the fundamental significance of this unilaterality— as when Eduard Meyer compared the chances of ancient slaves and modern

19  See Capogrossi (1978). 20   Digest 1.5.4.1.; for other texts, see Buckland (1908) ch. 2. Cf. the definition in Article 1 of the Slavery Convention of the League of Nations (1926): ‘Slavery is a status or condition of a person over whom any or all the powers attaching to the rights of ownership are exercised’ (quoted from C. W. W. Greenidge, Slavery, London 1958, p. 224). 21  Capogrossi (1978) 725–6. Cf. the inclusion of enslaved captives in the ‘tithe’ of war-spoils dedicated to a god: Börner (1957–63) III 252–5.

The Emergence Of A Slave Society

65

wage-earners to achieve wealth and social position22—destroys the possibility of fixing and comprehending the nature and history of slavery within any given society. Paradoxically, it was precisely this quality of the slave, as property, that offered the owning class flexibility (to which I shall return shortly) not available with other forms of compulsory labour. That is one reason why I stress what is a juristic category and therefore of itself not a sufficient ‘definition’ of the slave. How individual owners chose to treat their peculiar property was normally not a matter of mere whim or of differences in personality. Owners frequently offered slaves the incentive of eventual manumission through various arrangements which automatically brought into being a chain of behaviour and expectations that affected the master, too. Although in law and in fact he could always revoke the offer, the material gains to be derived from slavery would have been sharply reduced if such arrangements were not as a rule honoured. The slaveowner’s rights over his slave-property were total in more senses than one. The slave, by being a slave, suffered not only ‘total loss of control over his labour’23 but total loss of control over his person and his personality: the uniqueness of slavery, I repeat, lay in the fact that the labourer himself was a commodity, not merely his labour or labour-power. His loss of control, furthermore, extended to the infinity of time, to his children and his children’s children—unless, again, the owner by a unilateral act broke the chain through unconditional manumission. And even then, only children born subsequently were beneficiaries, not those already in existence at the moment of manumission. There is indeed ample evidence that manumission was not rarely withheld until a slave had progeny who could replace him (or her) in servitude, though there is no way of determining how frequent a practice this was, or whether it was largely restricted to certain categories of slaves, such as those in the Roman imperial employ. This totality of the slaveowner’s rights was facilitated by the fact that the slave was always a deracinated outsider—an outsider first in the sense that he originated from outside the society into which he was introduced as a slave, second in the sense that he was denied the most elementary of social bonds, kinship. ‘Quem patrem, qui servos est?’ (Plautus, Captivi 574). ‘What father, when he is a slave?’ The contrast with Spartan helots, Thessalian penestai or the clientes of early Rome provides the clearest delineation, as has already been 22  Meyer (1898) 211. 23  Patterson (1977b) 431. He offers as a ‘working definition’ of slavery ‘that condition in which there is an institutionalized alienation from the rights of labor and kinship’.

66

Finley

suggested: never did the master-class need or attempt to replenish the supply of dependent labour of this kind from outside. In stressing the kinlessness of slaves, I am concerned less with the juridical position, primarily the exclusion of slaves from legally recognized marriage (hence the Roman insistence on the word contubernium) than with the de facto position. There were slave unions and slave families, beyond a doubt, but they counted among the privileges that could be granted unilaterally by a slaveowner, and withdrawn unilaterally. The very possibility could be totally withdrawn by castration. As the poet Statius wrote in praise of Domitian’s (ineffectual) prohibition of the practice (Silvae 3.4.76–7), ‘nor by a harsh law are slave mothers fearful of the burden of sons’. Complete and brutal withdrawal of the kinship privilege also took the form of dispersing a slave family through sale. In ad 325 Constantine ordered an official to stop the break-up of slave families on imperial estates in Sardinia which he had transferred to private ownership (Theodosian Code 2.25.1), and that was almost certainly the earliest governmental interference with the practice.24 A little earlier the jurists were ruling that in an intestate succession or in certain other situations in which the contrary intention was not made explicit, slave families should not be broken up.25 That reflects a humane tendency, no doubt, but the more significant point is the stress on intent: the slaveowner’s freedom in this respect was not being challenged as late as the third century. We shall never know how common the practice was in antiquity to break up slave ‘families’. On the one hand, the presence in substantial numbers of slave children may be thought to argue for a relative permanence of slave unions. On the other hand, one body of evidence is highly suggestive in the opposite direction. We now possess sixty-odd original documents (nearly all from Egypt) recording the private sale of slaves, and in not one case is an adult male slave sold together with a wife or children. Furthermore, a study of the twenty-nine papyri that record the sex and age of female slaves sold in Egypt under the Roman Empire shows only two instances of the sale of a mother with child or children.26 Eleven of the girls in the catalogue were under the age of seventeen, seven under the age of thirteen. One fourteen-year old had already been sold three times previously. Unless one is prepared to believe that all the girls were orphans, all the mature women either unmarried or 24  I follow in essence the analysis of this text, and of the related Justinian Code 3.38.11, by Puglisi (1977). 25  See Buckland (1908) 77–8. 26  K. R. Bradley, ‘The Age at Time of Sale of Female Slaves’, Arethusa 11 (1978) 243–52; cf. Hopkins (1978) 164–5 on the Delphic manumissions.

The Emergence Of A Slave Society

67

widowed and childless, even such limited documentation has serious implications. As a pointer, I quote from a recent American study in which it was calculated that even on the low estimate that a mere 1.92% of the slave population of the southern states was sold in any given year, the statistical consequence was that any given slave had a virtually 50% ‘chance of being sold at least once in the course of a 35-year lifetime’ and on average ‘would witness 11.4 sales of members of his family of origin and of his own immediate family’. After all allowance is made for personal, regional and temporal variations, the authors rightly conclude that ‘the threat of sale was sufficiently large to affect the life of every slave’.27 A priori these three components of slavery—the slave’s property status, the totality of the power over him, and his kinlessness—provided powerful advantages to the slaveowner as against other forms of involuntary labour: he had greater control and flexibility in the employment of his labour force and far more freedom to dispose of unwanted labour.28 In consequence, a hierarchy arose within the slave population. One need only think of the following contemporaries: the slaves in the Spanish gold and silver mines or in the chain-gangs on Italian estates, the slaves in the imperial civil service, the slave overseers and stewards on the land, the urban slaves conducting their own commercial and manufacturing establishments in Rome and the other cities of Italy through the device of the peculium (to which we shall return). The slaves, in other words, constituted a type within the larger class of involuntary labour, but they were at the same time significantly divisible into sub-types. Stated differently, the slaves were a logical class and a juridical class but not, in any usual sense of that term, a social class.29 Yet, for all the advantages (or apparent advantages), slavery was a late and relatively infrequent form of involuntary labour, in world history generally and in ancient history in particular. Advantages and disadvantages are not essences but historical attributes that come and go under changing social and economic conditions. The critical question in the development and decline of ancient slavery can therefore be examined only by an inquiry into the necessary and sufficient conditions. What, in other words, brought about the transformation from the ‘primordial fact’ of individual slaves to the existence of slave societies, and what subsequently brought about a reversal of that process? The process was not only complex, but uneven and in a sense incomplete. Free labour was never eliminated, not only incidental hired labour but the 27  David (1976) 110–11. 28  On flexibility in American slavery, see Degler (1976) 8–10. 29  Vidal-Naquet (1968).

68

Finley

central labour of the independent peasants and artisans. The coexistence of free and slave labour, furthermore, was more than a coincidence in time and place; it was often a symbiosis, as in Italian agriculture where an adequate supply of free seasonal labour was a necessary condition for both the proper operation of slave latifundia and the economic survival of the free peasantry.30 That is straightforward enough, unlike the survival of non-slave forms of involuntary labour. The latter opens up the very large question of the unity we implicitly posit when we speak of the Graeco-Roman world, a question I cannot discuss except in the one aspect immediately relevant to the matter of involuntary labour. So long as the individual communities remained relatively small and selfenclosed, embracing urban centre and rural hinterland within a single structure, the polis in Greek, the rule seems to have been that slavery and other forms of compulsory labour did not coexist: in Athens, the rather mysterious groups known as pelatai and hektemoroi were replaced after Solon’s reforms at the beginning of the sixth century bc by slaves (or by free peasants); whereas, in contrast, the survival of helots rendered slaves unnecessary for the Spartans. That was the pattern, however, only in the old Greek world—the Greek mainland and the Aegean and Ionian islands. Elsewhere, Greek dispersion from early in the eighth century bc into territories occupied by peoples whose social structure was less advanced often brought about a mixed system: in the countryside widespread non-slave involuntary labour was introduced, as on the shores of the Black Sea, in parts of Asia Minor, and even for perhaps two centuries in Syracuse; whereas in the cities proper (Greek innovations in these areas) genuine slavery developed, though I do not believe we can say much more about it than those few, admittedly vague words. The subject is one about which we have hardly any documentation, but, in broad outline, recent study has substantiated the picture I have sketched so briefly.31 Later the establishment of territorial states through conquest had similar consequences with respect to exploitation. The Graeco-Macedonian rulers in the eastern lands conquered by Alexander left the rural labour regime basically unchanged, and that means perhaps 80 or even 90% of the total labour force. Why should they have done otherwise? But the ruling class in the newly created Greek poleis, in their determined effort to continue their old way of life, expected slave labour and obtained it.32 30  This has been well stressed by P. Garnsey and J. E. Skydsgaard, in Garnsey (1980). 31  Pippidi (1973) is now fundamental. 32  See Kreissig (1978) pt. II, with bibliography; E. S. Golubcova, in Blavatskaya (1972). The contrary view has been maintained by I. S. Svencickaya, without meeting the arguments

The Emergence Of A Slave Society

69

The Romans behaved in the same way as their Hellenistic predecessors. In Sicily, where the Syracusan Kyllyrioi had apparently disappeared by the early fourth century bc at the latest, slave labour became the sole form of involuntary labour and remained so after the Roman conquest. The same was largely true in Italy, though I take the view that in the areas of latest conquest on the peninsula debt-bondsmen were a significant element, and that throughout Italy tenants who fell into debt were in effect held on the land by compulsion.33 Everywhere else, apart perhaps from the northern provinces, the common pattern until the late Empire seems to have been one of urban slavery and rural dependent, rather than slave, labour, though more study is still necessary, for Spain and Gaul in particular.34 Once more, I had better repeat that free independent peasants and artisans survived all the political changes, in large numbers, and that nothing I have said implies that there were no slaves in the provincial countryside. A pattern does not require unanimity. In sum, slave societies, as distinct from societies in which there were slaves, were not to be found in all parts of what eventually became the Roman empire. What we accept as a political unit, and in a sense as a cultural unit, was not ipso facto an economic or social unity. In Wallerstein’s conceptual scheme, it was a ‘world-empire’, not a ‘world-system’; a structure in which different labourregimes and modes of production co-existed and were tied together politically rather than economically.35 An account of the development of Graeco-Roman slavery must therefore restrict itself, at least initially, to the central areas of Greece, Italy and Sicily, and that is what I shall do. It is conventional to begin the analysis by what I have repeatedly called the ‘numbers game’. I shall not join in, both because it has long been clear that on the other side; most recently, ‘Some Problems of Agrarian Relations in the Province of Asia’, Eirene 15 (1977) 27–54. Although I should not lay heavy stress on the point, it may be important that Isocrates, in a letter to Philip II (Epistles 3.5), chose to say ‘helots’, not ‘slaves’, in anticipating the fate of the people to be conquered in Asia. 33  By debt-bondsmen I mean the obaerati (or obaeraii) of Varro, De re rustica 1.17.2, and the men tied by nexum who are mentioned by Columella (1.3.12); on tenants in debt, see Finley (1976) 112–17. 34  Contrast C. R. Whittaker, ‘Rural Labour in Three Provinces of Rome’, in Garnsey (1980), with A. Daubigney and F. Favory, ‘L’esclavage en Narbonnaise et Lyonnaise’, in Colloque (1972) 315–88. On Visigothic Spain, see below at the beginning of ch. 4. 35  I. Wallerstein, ‘A World-System Perspective on the Social Sciences’, British Journal of Sociology 27 (1976) 343–52. Cf. Anderson (1974a) 22: The Ancient World as a whole was never continuously or ubiquitously marked by the predominance of slave-labour. But its great classical epochs … were those in which slavery was massive and general, amidst other labour systems.’

70

Finley

the evidence does not permit genuine quantification and because most of the players start from the false assumption that they must either produce astronomical figures in order to justify the label ‘slave society’ or, in opposition, that they somehow eliminate a slave society by demolishing the extreme figures. In 1860 the slaves made up 33% of the population in the southern states of the United States, a slightly lower percentage in Cuba and Brazil.36 On conservative estimates—60,000 slaves in Athens at the end of the fifth century bc, 2,000,000 in Italy at the end of the Republic—the comparable percentages are in precisely the same range, about 30 and 35%, respectively. That is more than sufficient, especially since the signs all say that slaveowners in antiquity were found considerably lower in the social and economic scale than in the New World,37 and since this proportion of slaves was retained in antiquity over a long period of time: the entire history of slavery in the United States lasted no more than the period from Augustus to Septimius Severus. It is equally impossible to quantify the distribution of slaves among the free men. However, a few indisputable figures will give an idea of the concentration at the upper end. Early in the fourth century bc, the father of Demosthenes had two groups of slaves, making furniture, swords and cutlery, 52 or 53 in total; in the previous generation, the orator Lysias and his brother jointly inherited about twice that number, engaged in producing shields, the largest single manufacturing establishment on record in the whole of antiquity; the slave workforce in the Athenian silver mines during the classical period often ran into five figures; a prefect of the city of Rome in the reign of Nero, Lucius Pedanius Secundus, had 400 slaves in his town house alone; at about the same time, the administration of the Roman aqueducts maintained a permanent staff of 700 slaves, including the ‘architects’.38 I have cited these few, admittedly outsize figures as a prelude to the general point that an assessment of the place of slaves in a society is not a matter of their totals, given a reasonably large number, but of their location, in two senses—first, who their owners were; secondly, what role they played, in the economy but not only in the economy. There were no slave employments, apart from mining as a general rule and domestic service, the latter defined as service in households other than those of one’s immediate family. Equally there were no free employments, other than law and politics (as distinct from administration) and normally the army (but not the navy, and excluding the servants of individual soldiers). In practice, however moralists such as Aristotle or Cicero 36  See the table in Hopkins (1978) 101; cf. Degler (1959). 37  See Jameson (1977–8). 38  Demosthenes 27.9–11, Lysias 12.19, Tacitus, Annals 14.43, Frontinus 96–118, respectively.

The Emergence Of A Slave Society

71

may have evaluated the work, all other occupations were shared by slaves and free men, often working side by side on identical tasks. Xenophon’s remark that ‘those who can do so buy slaves so that they may have fellow workers’ (Memorabilia 2.3.3) is not mere sententiousness.39 The ratios between slaves and free men in any occupation varied greatly in time and place: one need only contrast the normally free, respected physicians of Greece and their often unfree, always low-status counterparts in Rome and Italy. Such distinctions are interesting but marginal. The location of slavery cannot be determined by occupational tests. One fundamental nuance in Xenophon’s remark is that the fellow-workers were a slaveowner together with his slave or slaves, not slaves alongside hired free labour. In all Greek or Roman establishments larger than the family unit, whether on the land or in the city, the permanent work force was composed of slaves (or of other kinds of involuntary labour where that regime survived). I stress the word ‘permanent’, for, as I have already indicated, free casual and temporary hired labour was common enough, and indeed indispensable, in agriculture and in such abnormal activities as temple building. Not many generalizations about the ancient world can be substantiated with such certainty, with so few exceptions in the documentation.40 Farm tenancy, which is often adduced as an alternative to agricultural slavery, was no exception to this rule. Tenants were not employees: they either took on family-size farms which they worked without additional labour or they leased larger holdings and themselves employed slaves. Either way they did not breach the normal structure of labour on the land. Nor was there a slave ‘level’ of work: in the larger establishments, urban and rural, slaves performed all tasks from the most menial to the professional and managerial. We may therefore locate slavery neatly and simply. With one exception which I shall clarify in a moment, free men dominated small-scale farming, much of it subsistence farming, as well as petty commodity production and small-scale trading in the cities; slaves dominated, and virtually monopolized, large-scale production in both the countryside and the urban sector. It follows that slaves provided the bulk of the immediate income from property (that is, income from other than political sources, such as the vast sums pocketed by Roman Republican commanders and provincial officials or tax-farmers, and other than the secondary income obtained by the rich from moneylending) of

39  Cf. Philochorus 328F97, ap. Macrobius, Saturnalia 1.10.22. 40  The possibility that the situation changed in the later Roman Empire will be considered in ch. 4.

72

Finley

the élites, economic, social and political.41 The exception I mentioned is primarily a procedural one: the practice, probably much more common in Italy than in Greece, whereby slaves enriched their owners by working as ‘independent’ craftsmen, shopkeepers and ‘businessmen’—through what the Romans called peculium—was merely a variant procedure for the benefit of the élite; in one sense it led to considerable slave participation in petty commodity production, with important social consequences, but economically it did not disturb the location of slavery as the basic source of élite income. Manumission, finally, was often but a further extension of the peculium-idea. I have of course been referring only to those ‘central’ areas in which other forms of involuntary labour were displaced by slavery. They were the slave societies of Graeco-Roman antiquity, and they were that precisely because of the location of slavery within them. And so the time has come to consider how and why that rare phenomenon came into being. The conventional starting-point has almost always been the ‘natural’ state of war that has supposedly existed in early times and in simple societies among different tribes and peoples. Already in antiquity, and in modern times ever since the international lawyers of the sixteenth and seventeenth centuries, it has been repeated like a litany that slavery was in the first instance a mitigation of barbarian modes of warfare. That was Fustel’s ‘primordial fact’. Historians of antiquity have then gone a step further, and insisted on war and conquest as the necessary condition for the creation of a slave society. I must demur at some length. The fallacy arises from a view of Roman history that is so dazzled by the vastness of the conquests and enslavements of the last two centuries before Christ that it is blind to the unmistakable, considerably earlier growth of slavery in Rome.42 No one will deny the enormous leap that occurred after the second Punic War. A similar leap occurred in the course of American history, for different reasons, but that does not invalidate the view that the southern states were already a slave society in the first half of the eighteenth century. So was Rome not later than the third century bc. All vigorous new institutions develop and expand, but that process follows their introduction and cannot be confused with the latter. 41  I make no assessment of the relative income from these various sources; see Finley (1973) ch. 2. 42  This is equally true, despite profound differences in other respects, of Meyer (1898) and Hopkins (1978) 8–15, 102–6; contra Westermann (1955) 70; P. Ducrey, Le traitement des prisonniers de la guerre dans la Grèce antique … (Paris 1968), pp. 74–5; and in a different context, F. De Martino, ‘Interno all’ origine della schiavitù a Roma’, Laheo 20 (1974) 163–93, at pp. 179–93.

The Emergence Of A Slave Society

73

Admittedly the two centuries before 200 bc are a wasteland for the student of Roman social and economic history, but the signs all point in the same direction. A first tax on manumissions is reported to have been levied as early as 357 bc (Livy 7.16.7). By the last quarter of that century controversy arose over the votes of freedmen in the Roman assemblies (Livy 9.46.10–14). In 296, during the third Samnite War, freedmen were called up for military service (Livy 10.21.4). In that same war, Livy records the enslavement of nearly 40,000 captives, a figure which may not be accurate but is also not complete.43 In 262 came the first of a long series of mass enslavements during the Punic wars, 25,000 after the capture of Agrigento.44 No one will seriously suggest that all these men, women and children were sold off to Carthage or to the Greek east rather than in Italy. Nor can it be maintained that, faced with an unprecedented and ‘unexpected’ phenomenon of many tens of thousands of slaves, the Romans quickly, whether consciously or unconsciously, slid into the habit of employing slave labour on a considerable scale for the first time. Another approach is available. The tradition of social conflict in Rome from the foundation of the Republic stresses the persistent efforts to restrict the landholdings of the oligarchy, and particularly the amount of ager publicus, conquered land, occupied by individuals for both farming and grazing. It is enough to mention the Licinian-Sextian laws of 367 bc and the three occasions in the 290s when men were prosecuted for exceeding either the maximum holding or the permitted number of grazing animals.45 No single detail need be correct—the maximum supposedly fixed by the Licinian-Sextian laws is a highly improbable duplicate of Tiberius Gracchus’ 500 iugera in 133 bc— but it would require remarkable scepticism to ignore the tradition totally. No one will deny that there were wealthier Roman landowners in the fourth and third centuries, and that they were not working farmers in the style of the legendary Cincinnatus. Who worked their land? The choice lies between clients, men bound by nexum (abolished, at least formally, by the lex Poetelia Papiria of 323 bc), and hired labour on the one hand, and slaves on the other. The choice seems to me an easy one to make. In the sources, Gelzer, noted long ago, ‘special emphasis is laid … on the way in which only the wealthy gained by

43  The evidence is presented by Volkmann (1961) 227–8. 44  Diodorus 23.9.1, Polybius 1.19.15. 45  The references are Livy 7.16.7; 10.13.14; 10.23.13; 10.47.4. See G. Tibiletti, ‘Il possesso dell’ ager publicus e le norme de modo agrorum sino ai Gracchi’, Athenaeum, n.s. 26 (1948) 173–235; 27 (1949) 2–41.

74

Finley

the system of occupatio, because the use of slave labour gave them the advantage over the free peasant-farmers.’46 Were the authors whose work lay behind Livy and other later writers all creating a fiction? Surely not, at least for the third century. Much has been made by some historians of the fact that the word vilicus is not in evidence before about 200 bc and that archaeologists have not found what they loosely call villas before the second century.47 I am unimpressed by this argument from silence. During the Hannibalic War, one in every two citizens of military age was mobilized in the armies and navies, and that is inexplicable, indeed unthinkable, without the presence in the labour force of large number of slaves and of a well established system of slave labour.48 There were even enough slaves by then to permit their enlistment in substantial numbers.49 None of this is to deny the peculiar importance of conquest in the history of Roman slavery. Its essential role, however, was in creating the basis for large estates, with all the consequences that followed for Roman society and therefore for the ‘structure’ of Roman slavery. The ‘conquest theory’ thus helps to explain the specific character of the Roman slave society, not its emergence. Comparative evidence reveals that a necessary condition for an adequate supply of slaves is not conquest but the existence, outside the society under consideration, of a ‘reservoir’ of potential slave labour on which the society can draw systematically and, as it has been neatly phrased, ‘on institutionally agreeable (legal and cultural) terms’.50 Need I recall that neither the Americans nor their slave traders from Portugal and later England made war in Africa (with the possible exception of the Portuguese in Angola) in order to draw on that reservoir? Nor did the ancient Greeks systematically make war on their main sources, the ‘barbarians’ of the east and northeast. Neither Chios nor Corinth, two of the city-states singled out by Greek writers as slave centres, was a conquering or imperial state. That the ‘barbarians’ fought each other and sold their own captives to ancient Greek (and modern Portuguese) slave-traders is irrelevant. So is the question whether or not the ‘barbarians’ themselves employed slaves.51 And so is the fact that, whenever the captives of wars within the Graeco46  M. Gelzer, The Roman Nobility, trans. R. Seager (Oxford 1969), p. 21. Gelzer’s pages 18–22 (originally published in 1912) lay out the evidence neatly. 47  E.g. E. Maróti, ‘The Vilicus and the Villa-System in Ancient Italy’, Oikumene 1 (1976) 109–24. 48  P. A. Brunt, Social Conflicts in the Roman Republic (London 1971), pp. 18–19. 49  The evidence is collected by J. M. Libourel, ‘Galley Slaves in the Second Punic War’, Classical Philology 68 (1973) 116–19. 50  Hahn (1971) 35 and Mintz (1977) 257, respectively. 51  See Pippidi (1973) 65.

The Emergence Of A Slave Society

75

Roman world were sold off, some of them were slaves to begin with, especially when a city had been captured: such actions merely redistributed the existing slaves, they did not add to the total stock. One illustration is worth pursuing because it has broad implications. The police force of Athens, from possibly c. 477 to c. 378 bc, consisted of a corps of state-owned Scythian slaves, originally numbering three hundred.52 That is a remarkable institution. How did the Athenians come to think of it, and how did they maintain it? Scythians were famous archers and some had been employed as mercenaries under the Pisistratid tyranny. But hired mercenaries and slave policemen are qualitatively very different. One precondition for the introduction of the latter was the prior existence of an established slave system; the second was the existence of an organized slave trade. Warfare and piracy, if frequent enough, might conceivably maintain a general supply of slaves; they could not possibly have guaranteed the necessary supply of such specialists as Scythian archers. Even war, furthermore, produces captives, not slaves; captives are transformed into slaves by the consumers, who obtain them through the agency of slave traders.53 In sum, war and conquest were no doubt important contributing factors to the establishment and preservation of a slave society, they were not a necessary condition (at least not in a direct way) and certainly not a sufficient condition. Stated differently, my argument is that, logically, the demand for slaves precedes the supply. The Romans captured many tens of thousands of men, women and children during the Italian and Punic wars because the demand for slaves already existed, not the other way round. Existence of a sufficient demand requires at least three necessary conditions.54 The first, in a world which was overwhelmingly agrarian, is private ownership of land, with sufficient concentration in some hands to need extra-familial labour for the permanent work-force. The second is a sufficient development of commodity production and markets (for the present discussion it does not matter whether the market is a distant one, an export-market in the popular sense, or a nearby urban centre). Hypothetically, helots and other forms of dependent labour can be 52  The fullest account is still A. Plassart, ‘Les archers d’Athènes’, Revue des études grecques 26 (1913) 151–213. Cf. O. Jacob, Les esclaves publics à Athènes (Bibl. de la Fac. de Philosophie et Lettres à l’Univ. de Liège 35, 1928), ch. 2. 53  See M. I. Finley, Aspects of Antiquity (2nd ed., Harmondsworth and New York 1977), ch. 12; ‘The Black Sea and Danubian Regions and the Slave Trade in Antiquity’, Klio 40 (1962) 51–9. The primacy of trade over conquest was argued briefly, but neatly, by Burckhardt (1898–1902) I 142. 54  For what follows, I am heavily indebted to Hahn (1971).

76

Finley

employed in non-commodity-producing societies, but not slaves, who must be imported regularly, in quantity, and therefore paid for. The third is a negative condition, the unavailability of an internal labour supply, compelling the employers of labour to turn to outsiders. All three conditions must exist at the same time, as they did in Athens and other Greek communities in the sixth century bc and in Rome by the third century bc at the latest. As documentation, the Roman evidence is unsatisfactory, as I have already indicated, though I myself have no doubt about the chronology and the broad outlines of the development. In contrast, the Athenian evidence, both archaeological and literary, is good enough, and I shall concentrate on Athens.55 Before 600 bc, the population of Attica had grown spectacularly from the ‘Dark Age’ nadir, a ruling class of ‘Eupatrid’ families held much of the land, there had been a measure of urbanization, and some commodity production had developed in both rural and urban sectors. Several of the necessary conditions for slavery were in existence, therefore, little as we may understand the processes by which they arose. Solon then provided the vital negative condition: however one imagines the status of the pre-Solonic hektemoroi and pelatai (Aristotle, Constitution of Athens 2.2), there can be no denial that after Solon debtbondage and other non-slave forms of involuntary labour effectively ceased to exist in Attica (though not, it is unfortunately necessary to repeat over and over, in many areas of the Greek world). The Eupatrids and presumably some non-aristocratic wealthy families now required a labour-force to replace those they had lost through the Solonic reforms; they were unable to find it internally and they turned to outsiders, which means to slaves. Why? That is the critical, and most difficult, question. I do not pretend that I am able to answer it satisfactorily, but we can at least get rid of some persistent, bad arguments. One such argument dismisses the whole problem by stressing the paltry scale of Athenian wealth and landholding as compared with the Roman and Italian (a comparable argument to dismissing Italian slavery before the second century bc). No doubt a Pompey or an Ahenobarbus would have laughed at the wealth pretensions of a sixth-century pentakosiomedimnos, a member of the highest census-class in the Solonic constitution. But the latter had to produce his 500 medimnoi annually to retain his status, and what matters is not how that figure compares with wealth-standards in another society 55  On what follows I restrict references to two points: on population growth, see the preliminary analysis in A. M. Snodgrass, Archaeology and the Rise of the Greek State (Cambridge Inaug. 1977), pp. 10–16; on private property in land, M. I. Finley, ‘The Alienability of Land in Ancient Greece: A Point of View’, Eirene 7 (1968) 25–32), reprinted in French, Annales, e.s.c. 25 (1970) 1271–7.

The Emergence Of A Slave Society

77

but how it ranks within its own society: two-thirds higher than the qualification for a knight, two and a half times the qualification for military (hoplite) service. What matters even more is that some, at least, of the 500 medimnoi were commodities. Even if one believes that the whole of Plutarch’s life of Solon is a fiction, including his stress on olive production and export, the fact remains that every scrap of metal employed in Attica for civilian and military purposes had to be imported; that silver was being mined at Laureion in the Bronze Age, and in the archaic period from the ninth century bc at the latest;56 that fine Athenian pottery was being exported from well before Solon’s time and achieved virtually monopoly status by the middle of the sixth century; that the urban population, relatively small though it may have been, required pots and pans and work-tools as well as food and drink.57 It is no objection that, on present evidence, Athens did not begin to coin money until some decades after Solon: Near Eastern trade had managed for millennia without coins, as had the earlier Greeks, and the Phoenicians and Cathaginians continued to do so long after the Greeks adopted coinage. Without a sufficient cash-income, the Athenian élites could not have acquired the necessities for even their relatively low life-style, for their indispensable weaponry, or for the taxes which paid for public works, public festivals and public cults.58 None of this requires a revival of the Beloch-Meyer ‘modernism’; it requires merely an acceptance of some commodity production, in particular by the élites, and of the notion that members of the élites did not personally perform all the requisite labour, aided solely by members of their individual families. And so we are back to the critical question. Why was it necessary to seek the labour outside? Slavery as such did not have to be invented—it was a 56  See J. Servais, in Thorikos 1965 (Brussels 1967), pp. 22–24, and J. Bingen in Thorikos 1964 (Brussels 1967), pp. 29–30, respectively. 57  As evidence of commodity production in other Greek poleis I note (1) the carriage of wine, olive and pottery to Naucratis in the Egyptian Delta by traders from a number of cities in Asia Minor and the Aegean islands, beginning before 600 bc.: M. M. Austin, Greece and Egypt in the Archaic Age (Proceedings of the Cambridge Philological Society, Supp. 2, 1970), pp. 22–7, 36–40; (2) several Thasian regulations controlling the sale of local grapes and wine and the import of foreign wine; the earliest clear text, Inscriptiones Graecae XII Supp., no. 347, originally published with a useful commentary by G. Daux, in Bulletin de correspondance hellénique 50 (1926) 214–26, dates from the final decades of the fifth century bc, but there can be no doubt that regulation began much earlier, as shown by the fragmentary early fifth-century inscription, Supplementum Epigraphicum Graecum XVIII 347. 58  In this context, taxes and liturgies need not be distinguished.

78

Finley

‘primordial fact’, as familiar to the Greeks as to anyone else. But slavery as the form of labour for others was a radically new idea. I assume that the decision was imposed not by those who needed to employ labour but by those Athenians whom they sought to employ. It is the unavailability of the latter, en masse not as miscellaneous individuals, that requires explanation. In the complete absence of documentation, we have no choice but to speculate, and that speculation is not helped by the experience in the New World. The latter was a settlement area, with vast quantities of land becoming available as time went on; without a resident labour force (other than the native Indians who proved insufficient, when not useless); with the possibility, realized at an early date, of a small number of export-crops, cotton, tobacco, sugar, in great demand in a world-market. Nothing could have been more unlike Attica, with little arable land, none of it ‘free’, and with a sufficient, and often excess, population, in a pre-industrial, pre-capitalist world. In Attica, slavery was never accompanied by the massive growth of concentration in land-holding that occurred in Italy through the dispossession of peasant smallholders (what may be called the peculiar Roman substitute for the land ‘freely’ available in the New World).59 Yet even in Attica slavery was an agrarian as well as an urban institution.60 There is a way, however, through which land ownership offers an initial approach to our speculation about the origins of the slave society. Fundamental to the polis, Greek or Roman, from the time when it emerged in its archaic form out of the pre-polis stage, was the deep conviction that membership in the 59  Hopkins (1978) 102 correctly stresses the point: ‘We have to explain not only the importation of slaves, but the extrusion of citizens.’ However, as I have already indicated, I link that not with the establishment of the slave society but with its expansion. On Athens, I believe I was the first to challenge the prevailing view of a sharply declining peasantry in fourth-century Athens: Studies in Land and Credit in Ancient Athens 500–200 B.C. (New Brunswick, N. J., 1952), pp. 79–87; ‘Land, Debt, and the Man of Property in Classical Athens’, Political Science Quarterly 68 (1953) 249–68. My arguments have been generally accepted, e.g. by Cl. Mossé, ‘La vie économique d’Athènes au IV e siècle: Grise ou renouveau?’, Praelectiones Pataviniae (1972) 135–44; V. N. Andreyev, ‘Some Aspects of Agrarian Conditions in Attica in the Fifth to Third Centuries B.C.’, Eirene 12 (1974) 5–46, at pp. 18–25; G. Audring, ‘Zur wirtschaftlichen und sozialen Lage der attischen Bauern im ausgehenden 5. und im 4. Jahrhundert v. u. Z.’, Jahrbuch für Wirtschaftsgeschichte (Sonderband 1977) 9–86, at pp. 35–43. The role of ‘free’ land in the rise of New World slavery is still contested: see the negative views of S. L. Engerman, ‘Some Considerations relating to Property Rights in Man’, Journal of Economic History 33 (1973) 43–65, and Patterson (1977a), with an important qualification by Mintz (1977). 60  Jameson (1977–8).

The Emergence Of A Slave Society

79

polis (which we may call citizenship) was inextricably bound up with possession of the land, the obligation of military service, and religion. I do not know of a single exception to the rule that land ownership was restricted to citizens by law, except for those rare individuals to whom the state granted the right as a personal privilege in reward for public service of one sort or another. The standard ‘revolutionary’ slogan was ‘Cancellation of debts and redistribution of the land’, the slogan of a dispossessed peasantry: it was raised in Solonic Athens and it was acknowledged by Roman senators and emperors alike in their colonies and their veteran settlements. The psychology was briefly and pathetically summed up by Thucydides (2.16.2) in reporting the reaction of the country people to the Periclean strategy, which required them, a clear majority of the whole citizenry, to withdraw behind the city walls when the Spartans made their first raid, in 431 bc.: ‘They grieved and could hardly bear to leave their homes and ancestral shrines which had always been theirs from the establishment of the ancient constitution.’ Psychology is all that we have to fall back on—the political and social psychology that prevailed in the period when the élite, having lost their older forms of involuntary labour, turned to slave outsiders. The peasantry had won their personal freedom and their tenure on the land through struggle, in which they also won citizenship, membership in the community, the polis. This in itself was something radically new in the world, and it led in turn to the second remarkable innovation, a slave society. My model, I need hardly add, diverges radically from Eduard Meyer’s, despite certain apparent similarities. His revolutionary proletariat—making up at least half the citizen population, engaged in a great three-way class war with agrarian aristocrats and urban capitalists, jealous of the growing body of slaves, lacking skills themselves, unwilling to work for others, and therefore demanding public support—is one gigantic fiction.61 One need only ask, Which ancient states supported their poor citizens in idleness? Not even Athens, though Athens in the fifth and fourth centuries bc went much further than any other state in providing ways of supplementing income; and surely not the Ionian cities or Aegina, Rhodes, Corinth, which Meyer includes in his fictitious account. He surely knew that, but he apparently allowed his strong political convictions to confuse the picture by linking ancient slavery to democracy, a patently false link. Nothing I have said implies deliberative action, a consideration of choices followed by a decision to select one of them, namely, slavery. Everyone knew 61  Meyer (1898) 193–8.

80

Finley

the impossibility of compelling the peasant- or artisan-citizens to become a hired labour force, the citizens who were themselves required in the army;62 everyone knew that free men would not regularly work for another voluntarily; and everyone knew that there was an institution we call slavery. Hence, in my speculation, there was general acquiescence in the shift to slave labour. There was no jealousy of slaves, no competition with them, either in the early stages or in the peak periods;63 on the contrary, the dream of the man who could not afford a slave was to be able to do so one day (Lysias 24.6). The free man was one who neither lived under the constraint of, nor was employed for the benefit of, another; who lived preferably on his ancestral plot of land, with its shrines and ancestral tombs. The creation of that type of free man in a lowtechnology, pre-industrial world led to the establishment of a slave society. There was no realistic alternative. It will have been noticed that in this speculation I have said nothing about the absolute or relative profitability of slavery as a system of production. At no point has the modern historian’s ideology been a more distorting influence— not one particular ideology, furthermore, but otherwise conflicting ideologies. Christian moral judgments, Adam Smith’s faith in a free economy, and a mechanical application of the Marxist theory of surplus value have joined hands to substitute faith for demonstration. Thus, in a discussion headed ‘Did Slavery Pay?’, Michell concludes that ‘except where it can be used in the simplest and least technical of tasks and in large numbers’ it ‘does not pay’.64 Shtaerman agrees and goes so far as to say, incredibly, that ‘no Roman who was known

62  Mintz (1977) 257 includes ‘police power adequate to the legal-military containment of a free population’ among the conditions adverse to the introduction of modern slavery. 63  The only statement known from the whole of antiquity that has been taken to imply slave-free competition is a brief fragment of the lost work of the third-century bc historian Timaeus (566F11a, ap. Athenaeus 6.264D), reporting that the acquisition of 1000 slaves in backward Locris by ‘Aristotle’s friend’ Mnason stirred great resentment because it deprived ‘the young’ of their customary livelihood earned by ‘serving the elders domestically’. Even if true, and that is by no means certain, the anecdote is irrelevant in an account of slavery as a labour-force. Heitland (1921) 441n4 curtly dismissed it as referring ‘solely to domestic and personal attendance’; cf. the more nuanced but equally dismissive analysis of Vidal-Naquet (1968) 105–6. Recently G. Nenci, while acknowledging the uniqueness of the text, has made an ingenious effort to squeeze greater significance from it but I am not persuaded: ‘Il problema della concorrenza fra mandopera libera e servile nella Grecia classica’, Annali della Scuola Normale di Pisa, Classe di Lettere e Filosofia, 3rd ser., 8 (1978) 1287–1300. 64  H. Michell, The Economics of Ancient Greece (2nd ed., Cambridge 1957), p. 166.

The Emergence Of A Slave Society

81

through his wealth achieved that from agriculture’.65 One could rescue such false statements by introducing various qualifications but that would be a useless exercise. The significant points are, first, that Greeks and Romans for many centuries went on happily deriving income, often large income, from their slaves; secondly, that we have no basis for making comparisons of the efficiency, productivity or profitability of different types of labour in antiquity (nor, I may add, have we adequate data on the relative cheapness or costliness of slaves in general or in my particular period); thirdly, that the ancient themselves lacked the tools with which to make such comparisons, which for them would have been a purely academic exercise anyway. The one certain conclusion that has emerged from the harsh debates among cliometricians is that they have been unable to come to any agreed conclusion on such questions as the relative efficiency or profitability of American slave labour or the economic significance of changing slave prices, despite the wealth of data and the sophisticated mathematical techniques applied to them;66 but virtually everyone is now agreed that ‘slavery was generally profitable to the individual planter’ and that the society itself ‘had come to recognise bondage as a way of making money as well as of making a living’.67 The rare calculations by ancient writers are simply pathetic in their incompetence,68 It is a mistake of modern historians to take seriously the occasional moralizing generalizations—Columella’s assertion (1.7.5) that personal supervision always yields a greater return than supervision by a vilicus or than tenancy; the elder Pliny’s endlessly quoted ‘the latifundia have destroyed Italy’ (Natural History 18.35)—on the assumption that they are based on systematic calculations. Similarly gloomy generalizations are attested from the United States in infinitely greater number; no historian of American slavery any longer accepts them as evidence, other than ideological, and I see no reason to give the ancient parallels higher credence. In short, considerations of efficiency, productivity, profitability, played little part if any in the creation of a slave society in Greece or Rome; comparative considerations none at all. 65  Shtaerman (1969) 13 and 19, respectively. 66  It is enough to cite R. W. Fogel and S. L. Engerman, Time on the Cross (2 vols., Boston 1974), the critique by David (1976), and the review of the latter by D. Macleod in the Times Literary Supplement for June 23, 1978. The older discussions can be conveniently studied in Did Slavery Pay?, ed. H. G. J. Aitken (Boston 1971). 67  Degler (1976) 8. 68  Mickwitz (1937) and (1939) remain fundamental; cf. G. E. M. de Ste. Croix, ‘Greek and Roman Accounting’, in Studies in the History of Accounting, ed. A. C. Littleton and B. S. Yamey (London 1956), pp. 14–74; Duncan-Jones (1974) ch. 2.

82

Finley

The story does not end there, of course. Once established, a slave society had its own dynamic: the conditions that led to its creation were not identical with the conditions that led to its maintenance, expansion or decline. Some of the latter were the consequences of the existence of the slave society, which we shall look at later. Bibliography Note: Works cited only once are not normally repeated here. Abbreviations: Abh. Mainz = Akad. d. Wissenschaften u. d. Literatur, Mainz, Abhandlungen der geistes- u. sozialwiss. Klasse. RE = G. Wissowa, W. Kroll, et al., ed., Paulys Real-Enzyklopädie der klassischen Altertumswissenschaft, 1894–. Stuttgart. ZRG = Zeitschrift der Savigny-Stiftung für Rechtsgeschickte, Romanistische Abteilung. Abignente, G. (1890). La schiavitù nei suoi rapporti colla chiesa e col laicato. Turin. Anderson, P. (1974a). Passages from Antiquity to Feudalism. London. Anderson, P. (1974b). Lineages of the Absolute State. London. Annequin, J., et al., ed. (1975). Formes d’exploitation du travail et rapports sociaux dans l’antiquité classique (Recherches internationales à la lumière du marxisme 84). Paris. Backhaus, W. (1974). Marx, Engels und die Sklaverei. Düsseldorf. Backhaus, W. (1975). ‘John Elliott Cairnes und die Erforschung der antiken Sklaverei’, Historische Zeitschrifi ccxx 543–67. Barrow, R. H. (1925). Slavery in the Roman Empire. London. Bellen, H. (1971). Studien zur Sklavenflucht im römischen Kaiserreich. Wiesbaden. Beloch, J. (1886). Die Bevölkerung der griechisch-römischen Welt. Leipzig. Biezunska-Malowist, I. (1974–7). L’esclavage dans l’Egypte grécoromaine, 2 vols. (Polish Academy of Science). Biot, Ed. (1840). De l’abolition de l’esclavage ancien en Occident. Paris. Blair, W. (1833). An Inquiry into the State of Slavery amongst the Romans. Edinburgh. Blavatskaya, T. V., et al. (1972). Die Sklaverei in hellenistischen Staaten im 3–1. Th. v. Chr., trans. M. Bräuer-Pospelova et al. Wiesbaden. Böckh, A. (1817). Die Staatshaushaltung der Athener, 2 vols. (3rd ed. by Max Fränkel, 1886). Berlin. Bömer, F. (1957–63). Untersuchungen über die Religion der Sklaven in Griechenland und Rom, in Abh. Mainz 1957 no. 7, 1960 no. 1, 1961 no. 4, 1963 no. 10. Brockmeyer, N. (1971). Bibliographie zur antiken Sklaverei, ed J. Vogt. Bochum. Bücher, K. (1904). Die Entstehung der Volkswirtschaft, 5th ed. (1st ed. 1893). Tübingen. Bücher, K. (1922). Beiträge zur Wirtschaftsgeschichte. Tübingen.

The Emergence Of A Slave Society

83

Büchsenschütz, B. (1869). Besitz und Erwerb im griechischen Altenthume. Halle. Buckland, W. W. (1908), The Roman Law of Slavery. Cambridge. Burckhardt, J. (1898–1902). Griechische Kulturgeschichte, 4 vols., cited from reprint (1956), Darmstadt. Calabi, F. (1979). ‘“Despotes” e “technites”. Definizioni essenziali e definizioni funzionali nella “Politica” di Aristotele’, Quaderni di storia ix 109–34. Calderini, A. (1908). La manomissione e la condizione dei liberti in Grecia. Milan. Capogrossi, L. (1978). ‘Il campo semantico della schiavitù nella cultura latina del terzo e del secondo secolo a. C.’, Studi storici xviii 716–33. Carandini, A. (1976). ‘Le forme di produzione dell’ economica e le forme di circolazione dell’ antropologia economica’, Quaderni di Critica marxista 215–34, reprinted in his Archeologia e cultura materiale (2nd ed., Bari 1979) 354–84. Cassirer, E. (1951). The Philosophy of the Enlightenment, trans. F. C. A. Koelln and J. C. Pettegrove. Boston. Ciccotti, E. (1899). Il tramonto della schiavitù nel mondo antico, cited in the 1977 ed. Turin. Ciccotti, E. (1977). Il tramonto della schiavitù nel mondo antico, 2 vols. (first ed. 1899). Bari. Colloque (1971). Actes du Colloque 1971 sur l’esclavage (Annales littéraires de l’Univ. de Besançon 140, 1972). Colloque (1972). Ibid. 163, 1974. Colloque (1973). Ibid. 182, 1976. David, P. A., et al. (1976). Reckoning with Slavery. New York. Davis, D. B. (1966). The Problem of Slavery in Western Culture. Ithaca. Davis, D. B. (1974). ‘Slavery and the Post-World War II. Historians’, Daedalus ciii 1–16. Degler, C. N. (1959). ‘Starr on Slavery’, Journal of Economic History xix 271–7. Degler, C. N. (1970). ‘Slavery in Brazil and the United States. A Comparison’, American Historical Review lxxv 1004–28; reprinted in (and cited from) Weinstein/Gatell (1973) 342–73. Degler, G. N. (1976). “The Irony of American Slavery’, in Perspectives and Irony in American Slavery, ed. H. P. Owens, pp. 3–25. Jackson, Miss. Diakonoff, I. M. (1974). ‘Slaves, Helots and Serfs in Early Antiquity’ Acta Antiqua xxii 45–78. Ducat, J. (1978). ‘Aspects de l’hilotisme’, Ancient Society ix 5–46. Duncan-Jones, R. (1974). The Economy of the Roman Empire. Cambridge. Dureau de la Malle, A. J. C. A. (1840). Economie politique des Romains, 2 vols. Paris. Ehrenberg, V. (1951). The People of Aristophanes, 2nd ed. Oxford. Elkins, S. M. (1959). Slavery. Chicago. Finley, M. I. (1959). ‘Was Greek Civilization Based on Slave Labour?’, Historia viii 145– 64; corrected reprint in Finley (1981).

84

Finley

Finley, M. I. (1960). ‘The Servile Statuses of Ancient Greece’, Revue internationale des droits de l’antiquité, 3rd ser., vii 165–89; corrected reprint in Finley (1981). Finley, M. I. (1964). ‘Between Slavery and Freedom’, Comparative Studies in Society and History vi 233–49, also in French in Annequin (1975); corrected reprint in Finley (1981). Finley, M. I. (1965a). ‘La servitude pour dettes’, Revue historique de droit français et étranger, 4th ser., xliii 159–84; corrected reprint in English in Finley (1981), in German in Kippenberg (1977). Finley, M. I. (1965b). ‘Technical Innovation and Economic Progress in the Ancient World’, Economic History Review, 2nd ser. xviii 29–45; corrected reprint in Finley (1981). Finley, M. I., ed. (1968). Slavery in Classical Antiquity (repr. with supp.). Cambridge and New York. Finley, M. I. (1973). The Ancient Economy. Berkeley and London. Finley, M. I., ed. (1974). Studies in Ancient Society. London and Boston. Finley, M. I. (1975). The Use and Abuse of History. London. Finley, M. I., ed. (1976). Studies in Roman Property. Cambridge. Finley, M. I. (1977a). ‘The Ancient City: from Fustel de Coulanges to Max Weber and beyond’, Comparative Studies in Society and History xix 305–27; corrected reprint in Finley (1981). Finley, M. I. (1977b). ‘“Progress” in Historiography’, Daedalus cvi no. 3, 125–42. Finley, M. I. (1981). Economy and Society of Ancient Greece, ed. R. P. Sailer and B. D. Shaw. London. Francotte, H. (1900–1). L’industrie dans la Grèce antique, 2 vols. (Bibliothèque de la Fac. de Philosophie et Lettres, Univ. de Liège, fasc. 7–8). Francotte, H. (1916). ‘Industrie und Handel’, in re ix 1381–1459. Frank, T., ed. (1933–40). An Economic Survey of Ancient Rome, 6 vols. Baltimore. Frederikson, G. M., and Lasch, C. (1967). ‘Resistance to Slavery’, Civil War History xiii 315–29, reprinted in (and cited from) Weinstein/Gatell (1973) 118–33. Fustel de Coulanges, N. M. (1885). ‘Le colonat romain’, in his Recherches sur quelques problèmes d’histoire, pp. 1–186. Paris. Garlan, Y. (1972). ‘Les esclaves grecs en temps de guerre’, in Actes du Colloque d’histoire social 1970, 29–62. Paris Garnsey, P. (1970). Social Status and Legal Privilege in the Roman Empire. Oxford. Garnsey, P., ed. (1980). Non-slave Labour in Graeco-Roman Antiquity (Proceedings of the Cambridge Philological Society, Supp. 6). Genovese, E. D. (1974), Roll, Jordan, Roll. New York. Gernet, L. (1955). Droit et société dans la Grèce ancienne (reprint). Paris. Hahn, I. (1961). ‘Freie Arbeit und Sklavenarbeit in der spätantiken Stadt’, Annales Univ…. Budapestiensis, Sectio historica iii 23–39.

The Emergence Of A Slave Society

85

Hahn, I. (1971). ‘Die Anfänge der antiken Gesellschaftsformation in Griechenland und das Problem der sogennanten asiatischen Produktionsweise’, Jahrbuch für Wirtschaftsgeschichte ii 29–47; reprinted in Kippenberg (1977) 68–99. Hahn, I. (1976). ‘Sklaven und Sklavenfrage im politischen Denken der Spätantike’, Klio lviii 459–70. Heeren, A. H. L. (1826). Ideen über die Politik, den Verkehr und den Handel der vornehmsten Volk der alten Welt iii 1, 4th ed., in his Historische Werke xv. Göttingen. Heitland, W. E. (1921). Agricola. Cambridge. Highet, G. (1973). ‘Libertini Patre Natus’, American Journal of Philology xciv 268–81. Himmelmann, N. (1971). Archäologisches zum Problem der griechischen Sklaverei, in Akad. Mainz, no. 13. Hobsbawm, E. (1964). Introduction to Marx. London. Hopkins, K. (1978). Conquerors and Slaves. Cambridge. Hume, D. (1752). ‘Of the Populousness of Ancient Nations’, cited from the World’s Classics ed. of his Essays (London 1903), pp. 381–451. Ingram, J. K. (1895). History of Slavery and Serfdom. London. Jameson, M. H. (1977–8). ‘Agriculture and Slavery in Classical Athens’, Classical Journal lxxii 122–45. Jameson, R. P. (1911). Montesquieu et l’esclavage. Paris (reprinted New York, 1941). Joly, R. (1969/70). ‘Esclaves et médecins dans la Grèce antique’, Sudhoffs Archiv liii 1–14. Jones, A. H. M. (1964). The Later Roman Empire 284–602, 3 vols. Oxford. Kajanto, I. (1969). ‘Tacitus on the Slaves. An Interpretation of the Annales. xiv, 42–45’, Arctos n.s. vi 43–60. Kiechle, F. (1969). Sklavenarbeit und technischer Fortschritt im römischen Reich. Wiesbaden. Kippenberg, H. G., ed. (1977). Seminar; Die Entstehung der antiken Klassengesellschaft. Frankfurt. Klees, H. (1975). Herren und Sklaven: Die Sklaverei im oikonomischen und politischen Schrifttum der Griechen in klassischer Zeit. Wiesbaden. Kreissig, H. (1978). Wirtschaft und Gesellschaft im Seleukidenreich. Berlin. Kudlien, F. (1968). Die Sklaven in der griechischen Medizin der klassischen und hellenistischen Zeit. Wiesbaden. Kühne, H. (1962). ‘Zur Teilnahme von Sklaven und Freigelassenen an der Bürgerkriegen der Freien im 1. Jahrhundert v.u.Z. in Rom’, Studii clasice iv 189–209. Lange, R. (1899). Review of Ciccotti (1899), Wochenschrift für klassische Philologie xvi 761–70. Lauffer, S. (1960). ‘Die Sklaverei in der griechisch-römischen Welt’, Rapports (XIth Intl. Historical Congress) ii 71–97. Uppsala. Lencman, Ja. A. (1966). Die Sklaverei in mykenischen und homerischen Griechenland, trans. M. Braüer-Pospelova. Wiesbaden.

86

Finley

Lepore, E. (1970). ‘Economia antica e storiografia moderna (Appunti per un bilancio di generazioni)’, in Ricerche … in memoria di Carrado Barbagallo i 3–33. Naples. Letourneau, C. (1897). L’évolution de l’esclavage dans les diverses races humaines. Paris. Levy, E. (1948). ‘Von römischen Precarium zur germanischen Landleihe’, zrg lxvi 1–30. McCulloch, J. R. (1845). The Literature of Political Economy: a Classified Catalogue. London. Marx, K. (1964). Pre-capitalist Economic Formations, trans. J. Cohen. London. Marx, K. (1973). Grundrisse, trans. M. Nicolaus. Harmondsworth. Mazza, M. (1977). Introduction to Ciccotti, E. (1977), reprinted in Klio lxi 57–83 (1979). Mazzarino, S. (1966). Il pensiero storico classico, 2 vols, in 3. Bari. Meillassoux, C., ed. (1975). L’esclavage en Afrique précoloniale. Paris. Meyer, E. (1895). Die wirtschaftliche Entwicklung des Altertums, reprinted in (and cited from) Meyer (1924) 79–168. Meyer, E. (1898). Die Sklaverei im Altertum, reprinted in (and cited from) Meyer (1924) 169–212. Meyer, E. (1899). Forschungen zur alten Geschichte, vol. 2. Halle. Meyer, E. (1902). Zur Theorie und Methodik der Geschichte, reprinted in (and cited from) Meyer (1924) 1–78. Meyer, E. (1924). Kleine Schriften, 2nd ed., vol. I. Halle. Mickwitz, G. (1937). ‘Economic Rationalism in Graeco-Roman Agriculture’, English Historical Review lii 577–89. Mickwitz, G. (1939). ‘Zum Problem der Betriebsführung in der antiken Wirtschaft’, Vierteljahrschrift f. Sozial- u. Wirtschaftsgesch. xxxii 1–25. Milani, P. (1972). La schiavitù nel pensiero politico: dai Greci al Basso Medio Evo. Milan. Millar, J. (1771). Observations concerning the Distinction of Ranks in Society. Dublin. Mintz, S. W. (1977). ‘The so-called World System: Local Initiative and Local Response’, Dialectical Anthropology ii 253–70. Momigliano, A. (1966). Studies in Historiography. London. Mommsen, T. (1961). Römisches Strafrecht (reprint of ed. of 1899). Darmstadt. Morrow, G. R. (1939). Plato’s Law of Slavery and Its Relation to Greek Law. Urbana, Ill. Nehlsen, H. (1972). Sklavenrecht zwischen Antike und Mittelalter (vol. 1 only so far). Göttingen. Nieboer, H. J. (1900). Slavery as an Industrial System. The Hague. 2nd ed. 1910. Oertel, F. (1925). Appendix to Pöhlmann (1925), reprinted in Oertel (1975) 40–98. Oertel, F. (1975). Kleine Schriften zur Wirtschafts- und Sozialgeschichte, ed. H. Braunert. Bonn. Patlagean, E. (1977). Pauvreté économique et pauvreté sociale à Byzance, 4e–7e siècles, Paris and The Hague.

The Emergence Of A Slave Society

87

Patterson, O. (1970). ‘Slavery and Slave Revolts: A Sociohistorical Analysis of the First Maroon War, 1665–1740’, Social and Economic Studies xix 289–325; reprinted in (and cited from) Price (1973). Patterson, O. (1977a). ‘The Structural Origins of Slavery. A Critique of the NieboerDomar Hypothesis from a Comparative Perspective’, Annals of the N.Y. Acad, of Science ccxcii 12–33. Patterson, O. (1977b). ‘The Study of Slavery’, Annual Review of Sociology iii 407–49. Percival, J. (1969). ‘Seigneurial Aspects of Late Roman Estate Management’, English Historical Review lxxxiv 449–73. Pignoria, L. (1613). De servis, et eorum apud veteres ministeriis commentarius. Augsburg. Pippidi, D. M. (1973). ‘Le problème de la main-d’oeuvre agricole dans les colonies grecques de la mer Noire’, in Problèmes de la terre en Grèce ancienne, ed. M. I. Finley, pp. 63–82. Paris and The Hague. Pöhlmann, R. (1925). Geschichte der sozialen Frage und des Sozialismus in der antiken Welt, 3rd ed. by F. Oertel. Munich. Popma, T. (1608). De operis servis liber, Leiden. Price, R., ed. (1973). Maroon Societies: Rebel Slave Communities in the Americas. Garden City, N.Y. Puglisi, A. (1977). ‘Servi, coloni, veterani e la terra in alcuni testi di Constantino’, Labeo xxiii 305–17. Raskolnikoff, M. (1975). La recherche soviétique et l’histoire économique et sociale du monde hellénistique et romain. Strasbourg. Reitemeier, J. F. (1789). Geschichte und Zustand der Sklaverey und Leibeigenschaft in Griechenland. Berlin. Richter, W. (1958). ‘Seneca und die Sklaven’, Gymnasium lxv 196–218. Rivet, A. L. F., ed. (1969). The Roman Villa in Britain. London. Roscher, W. (1871). Ansichten der Volkswirtschaft aus dem geschichtlichen Standpunkte. Leipzig and Heidelberg. Rostovtzeff, M. (1941). The Social and Economic History of the Hellenistic World, 3 vols. Oxford. Rostovtzeff, M. (1957). The Social and Economic History of the Roman Empire, 2nd ed. by P. M. Fraser, 2 vols. Oxford. Rostowzew, M. (1910). Studien zur Geschichte des römischen Kolonates (Archiv f. Papyrusforschung, Beiheft 1). Salvioli, G. (1906). Le capitalisme dans le monde antique, trans. A. Bonnet. Paris. Sargent, R. L. (1924). The Size of the Slave Population at Athens (Univ. of Illinois Studies in the Social Sciences xii 3). Sartori, F. (1973). ‘Cinna e gli schiavi’, in Colloque (1971) 151–69.

88

Finley

Schiller, L. (1847). Die Lehre des Aristoteles von der Sklaverei (Jahresbericht d. K. Studienanstalt zu Erlangen in Mittelfranken). Erlangen Schwartz, Ed. (1938). Gesammelte Schriften I. Berlin. Seyfarth, W. (1963). Fragen der spätrömischen Kaiserzeit im Spiegel der Theodosianus. Berlin. Shtaerman, E. (1964). Die Krise der Sklavenhalterordnung im Western des römischen Reiches, trans, from the Russian (1957) by W. Seyfarth. Berlin. Shtaerman, E. (1969). Die Blütezeit der Sklavenwirtschaft in der römischen Republik, trans, from the Russian (1964) by M. Bräuer-Pospelova. Wiesbaden. Shtaerman, E. M., and Trofimova, M. K. (1975). La schiavitù nell’ Italia imperiale, I–III secolo, trans, from the Russian (1971). Rome. Spranger, P, P. (1960). Historische Untersuchungen zu den Sklavenfiguren des Plautus und Terenz, in Akad. Mainz, no. 8. Stier, H. E. (1945). Grundlagen und Sinn der griechischen Geschichte. Stuttgart. Stroud, G. M. (1856). A Sketch of Laws relating to Slavery in the Several States of the United States of America, 2nd ed. Philadelphia. Temperley, H. (1977). ‘Capitalism, Slavery and Ideology’, Past & Present lxxv 94–118. Thür, G. (1977). Beweisführung vor den Schwurgerichtshöfen Athens: Die Proklesis zur Basanos. Vienna. Tourmagne, A. (pseud. of A. Villard) (1880). Histoire de l’esclavage ancien et moderne. Paris. Verlinden, C. (1955). L’esclavage dans l’Europe mediévale (only vol. 1 published). Brugge. Vidal-Naquet, P. (1968). ‘Les esclaves grecs étaient-ils une classe?’, Raison Présente, no. 6, 103–12. Vidal-Naquet, P. (1972). ‘Réflexions sur l’historiographie grecque de l’esclavage’, in Colloque (1971) 25–44. Villard, A. see A. Tourmagne (pseudonym). Vittinghoff, F. (1960). ‘Die Theorie des historischen Materialismus über dem antiken “Sklavenhalterstaat”: Probleme der alten Geschichte bei den “Klassikern” des Marxismus und in den modernen sowjetischen Forschung’, Saeculum xi 89–131. Vittinghoff, F. (1961). ‘Die Bedeutung der Sklaven für den Übergang von der Antike ins abendländische Mittelalter’, Historische Zeitschrift cxcii 265–72. Vlastos, G. (1941). ‘Slavery in Plato’s Thought’, Philosophical Review l 289–304, reprinted in Finley (1968) 133–49. Vogt, J. (1974). Ancient Slavery and the Ideal of Man, trans. T. Wiedemann. Oxford. Vogt, J. (1978). ‘Die Sklaverei in antiken Rom’, Antike Welt ix, no. 3, 37–44. Volkmann, H. (1961). Die Massenversklavungen der Einwohner eroberter Städte in der hellenistisch-römischen Zeit, in Akad. Mainz, no. 3.

The Emergence Of A Slave Society

89

Wallace, R. (1753). A Dissertation on the Numbers of Mankind in Ancient and Modern Times. Edinburgh. Wallon, H. (1847). Histoire de l’esclavage dans l’antiquité, 3 vols. Paris. Wallon, H. (1879). 2nd ed. of Wallon (1847). Weber, M. (1891). Die Römische Agrargeschickte. Stuttgart. Weber, M. (1896). ‘Die sozialen Gründe des Untergangs der antiken Kultur’, reprinted in (and cited from) Weber (1924) 289–311. Weber, M. (1909). ‘Die Agrarverhältnisse des Altertums’, in HWB der Staatswiss., 3rd ed., reprinted in (and cited from) Weber (1924) 1–288. Weber, M. (1924). Gesammelte Aufsätze zur Sozial- und Wirtschaftsgeschichte. Tübingen. Weinstein, A., and Gatell, F. O., ed, (1973). American Negro Slavery: A Modern Reader, 2nd ed. New York. Welskopf, E. C. (1957). Die Produktionsverdältnisse im alten Orient und in der griechischrömischen Antike. Berlin. Welwei, K.-W. (1974–77). Unfreie im antiken Kriegsdienst, 2 vols. Wiesbaden. Westermann, W. L. (1935). ‘Sklaverei’, in re Supp. vi 894–1068. Westermann, W. L. (1955). The Slave Systems of Greek and Roman Antiquity. Philadelphia. White, Lynn, Jr. (1964). Medieval Technology and Social Change, paperback ed. Oxford. Wilamowitz, U. von (1910). Staat und Gesellschaft der Griechen und Römer (with B. Niese). (2nd ed. 1923). Berlin and Leipzig. Will, Ed. (1954). ‘Trois quarts de siècle de recherches sur l’économie grecque antique’, Annales, E.S.C. ix 7–22. Yanoski, J. (1860). De l’abolition de l’esclavage ancien au moyen âge et sa transformation en servitude de glèbe. Paris. Yeo, G. A. (1952). ‘The Economics of Roman and American Slavery’, Finanzarchiv, n.F. xiii 445–85.

Authority, Alienation, and Social Death Orlando Patterson All power strives for authority. A. Geoffrey Woodhead, in his study of Thucydides and the nature of power, observes that “there remains a need for spiritual and moral support, a need to say that an action is ‘right’ ... whatever the realpolitik behind the action.”1 In our examination of the property concept in the last chapter we saw that the master-slave relationship cannot be divorced from the distribution of power throughout the wider society in which both master and slave find themselves. Total power or property in the slave means exclusion of the claims and powers of others in him. If the master sought to exclude as far as possible all other claims and powers in his slave, it nevertheless remains true that he needed both the recognition and the support of the nonslave members of his community for his assumption of sovereign power over another person. An isolated master faced grave risks. Plato, who knew what he was talking about on this issue, shrewdly pointed out that a slave owner within his community had nothing to fear from his slaves because the entire state was ready to defend each individual citizen. But if he and his immediate family with more than fifty slaves were transported to the middle of a desert where no freeman could come to his defense, that citizen would be in great fear for his own life and that of members of his family, and he would try to ingratiate himself with the slaves by making promises and offers of freedom.2 Actually, the situation was more complex than this, for the danger the master faced was not merely physical. In all slaveholding societies the slave posed grave moral and spiritual dangers. Most slave populations have been so small that they were rarely considered a serious political menace; their danger lay in their capacity to offend supernaturally. The master’s task, then, had both a negative and a positive aspect. On the negative side, he had to defuse the S ource: Patterson, Orlando, “Authority, Alienation, and Social Death,” in Orlando Patterson, Slavery and Social Death: A Comparative Study, Cambridge, Mass.: Harvard University Press, 1982, 35–76. Copyright © 1982 by the President and Fellows of Harvard College. 1  Thucydides on the Nature of Power (Cambridge, Mass.: Harvard University Press, 1970), pp. 153–154. 2  The Republic, 9: 578 in Benjamin Jowett, ed. and trans., The Dialogues of Plato (New York: Random House, 1937), vol. 1, pp. 836–837.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_006

Authority, Alienation, And Social Death

91

potential physical and spiritual threat posed by his slave’s presence. And on the positive side, he had to secure extracoercive support for his power. Both were achieved by acquiring the thing we call authority.

Authority as Symbolic Control

What was involved in the acquisition of authority? From the community at large, authority came with the institutionalization of the slave relationship. It was achieved by incorporating it into the normative order. As Siegfried Lauffer puts it, the power relationship (Gewaltverhältnis) that formed the basis of the slave relationship had to become a rights relationship (Rechtsverhältnis).3 Those who were not directly involved with the relationship—though indirectly influenced by it—had to come to accept it not just grudgingly, but as the normal order of things (as did the nonslaveholding Greek, Roman, Hausa, or antebellum southern farmer). Nor was it only the nonslaveholding “freeman” whom the master wished to acknowledge his authority. The arrogance of power knows no bounds, for the master desired too that the slave recognize his authority, as well as his right to dominate him. To the extent that he did, to that degree was he able to walk fearlessly into the desert with his slave. And the truth is that many masters succeeded. As the history of the slaveholding peoples of the Sahara shows, many a master accompanied by his slave knew, for long periods of their lives, only each other and the desert. Understanding how this happens is no easy matter. Most social scientists faced with the problem of authority are content to cite a few well-known passages from Weber—the acknowledged authority on the subject—then continue blithely with their analysis. There is too much that is unsatisfactory in Weber’s analysis for us to take this course. He tells us that authority has three sources: law, charisma, and tradition.4 Law, however, cannot be a source of authority, for it is merely that complex of rules which has the coercive power of the state behind it. As the Scandinavian and other modern jurists have pointed

3  Siegfried Lauffer, “Die Sklaverei in der griechisch-römischen Welt,” in Rapports II, Eleventh International Congress of Historical Sciences, Stockholm, August 21–28, 1960 (Uppsala: Almquist and Wiksell, 1960), p. 76. 4  Max Weber, Basic Concepts in Sociology, trans. H. P. Secher (Secaucus, N.J.: Citadel Press, 1972), pp. 81–83. For a detailed analysis of Weber’s use of this concept see Reinhardt Bendix, Max Weber: An Intellectual Portrait (London: Methuen & Co., 1966), pt. 3.

92

Patterson

out, to define law as normative rules is to evade the crucial issue.5 Law itself begs for the thing we call authority; and as every student of jurisprudence knows, one of the major sources of law, and of law’s authority, is tradition. Nor does Weber’s notion of charisma get us very far. By its very nature, this is an exceptional phenomenon. No doubt there was the occasional master who was genuinely charismatic, but in general masters were no more endowed with unusual personal qualities than other persons, including their slaves. We are left only with tradition. Weber was on the right track here, but unfortunately was too vague. What does tradition mean? And why should the traditional automatically convey authority? By tradition Weber was obviously referring to the total complex of norms, values, ideas, and patterned behavior we call culture. I agree that somewhere in this vast universe of received human experience is to be found the source of authority; but where? The answer, I think, has been provided by students of symbolic anthropology, beginning with Meyer Fortes’ critique of Weber.6 Fortes and other British anthropologists, especially Raymond Firth, have argued that symbols, both private and public, constitute a major instrument of power when used directly or indirectly. Herein lies the source of authority. Those who exercise power, if they are able to transform it into a “right,” a norm, a usual part of the order of things, must first control (or at least be in a position to manipulate) appropriate symbolic instruments. They may do so by exploiting already existing symbols, or they may create new ones relevant to their needs. The full mechanics of this process of symbol appropriation is beyond the scope of the present work; what I shall do is examine the nature of symbolic control in the case of the master-slave relationship. Symbolic processes, like so many other areas of human experience, have both an intellectual and a social aspect. On the intellectual level symbolic thought attempts to explain, in the language of symbols, a given area of actual experience. It is essentially mythic, similar in intellectual form to the validating concepts and beliefs of 5  G. B. J. Hughes, Jurisprudence (London: Butterworth & Co., 1955), pp. 161–166. Hans Kelsen states the issue bluntly: “Law is a coercive order.” “The Pure Theory of Law,” in M. P. Golding, ed., The Nature of Law (New York: Random House, 1966), p. 112. 6  After criticizing Weber, Fortes points to the role “played by ceremony and ritual in the confirmation of status.” Drawing on the seminal work of Everett Hughes, he notes that all statuses require a “mandate from society” (Hughes’s term) and adds that “ritual mobilizes incontrovertible authority behind the granting of office and status and this guarantees its legitimacy and imposes accountability for its proper exercise.” “Ritual and Office in Tribal Society,” in Max Gluckman, ed., Essays on the Ritual of Social Relations (Manchester: Manchester University Press, 1962), p. 86.

Authority, Alienation, And Social Death

93

religion. The social aspect of symbolic behavior refers to the ritual processes by means of which symbolic ideas are acted out in terms of real human interactions. Such actions invariably are highly formalized and ceremonial. Where the experience being symbolized extends over a long period of time, there is a tendency for a clearly defined symbolic pattern to develop: critical stages in the developmental process, and especially the transition from one stage to the next, are given special ritual expression. The celebrated work of Arnold Van Gennep examined, for example, the various ritual expressions of the human life cycle among a variety of peoples.7 Similar rites of passage may be found in lasting relationships—and slavery, as we shall see, is one such case. A final theoretical point to note is the contribution of Victor Turner who, in his masterful treatise on the Ndembu and in later theoretical writings, developed the concept of the dominant symbol.8 Mythic and ritual processes by nature are multivocal, ambiguous, diffuse, and sometimes downright incomprehensible. Within a given cultural domain, however, a dominant symbol—a major mythic theme, a key ritual act—stands out as pivotal. By its emergence it makes possible an internal interpretation of the symbolic process on both the intellectual and the social level. Slavery, I intend to show in this chapter, is a highly symbolized domain of human experience. While all aspects of the relationship are symbolized, there is overwhelming concentration on the profound natal alienation of the slave. The reason for this is not hard to discern: it was the slave’s isolation, his strangeness, that made him most valuable to the master; but it was this very strangeness that most threatened the community and that most exercised that “primacy of feeling and willing over thinking” which is at the core of the symbolic mind. On the cognitive or mythic level, one dominant theme emerges, which lends an unusually loaded meaning to the act of natal alienation: this is the social death of the slave. On the ritual level, the enslavement process is expressed in terms of well-defined rites of passage.

7  The Rites of Passage, trans. M. B. Vizdeom and G. L. Caffee (London: Routledge & Kegan Paul, 1960). For a critical assessment of this work see Max Gluckman, “Les rites de passage,” in Gluckman, Essays on the Ritual of Social Relations, pp. 1–52. 8  Victor Turner, The Forest of Symbols (Ithaca, N.Y.: Cornell University Press, 1967), pp. 30–32, esp. chap. 4. For a more general and theoretical statement see his “Symbolic Studies,” Annual Review of Anthropology 4 (1975): 145–161.

94

Patterson

The Two Conceptions of Social Death

If the slave no longer belonged to a community, if he had no social existence outside of his master, then what was he? The initial response in almost all slaveholding societies was to define the slave as a socially dead person. Claude Meillassoux and his associates have most thoroughly explored this aspect of slavery. They reject the simplistic materialist view, which fails to take account of this problem—which indeed does not even recognize the existence of the problem.9 From the structural viewpoint, Meillassoux argues, slavery must be seen as a process involving several transitional phases. The slave is violently uprooted from his milieu. He is desocialized and depersonalized. This process of social negation constitutes the first, essentially external, phase of enslavement. The next phase involves the introduction of the slave into the community of his master, but it involves the paradox of introducing him as a nonbeing. This explains the importance of law, custom, and ideology in the representation of the slave relation. Summarizing his own views and those of his associate Michel Izard, Meillassoux writes: “The captive always appears therefore as marked by an original, indelible defect which weighs endlessly upon his destiny. This is, in Izard’s words, a kind of ‘social death.’ He can never be brought to life again as such since, in spite of some specious examples (themselves most instructive) of fictive rebirth, the slave will remain forever an unborn being (non-né).”10 There is much of value in this analysis, although it exaggerates to make the point. It goes astray, or at any rate is likely to mislead, mainly in its overemphasis of external sources and conquest as the initiating act of enslavement. It is simply not the case that “the condition of slavery never results from an internal process of social differentiation.” Meillassoux is here drawing too narrowly on his field experience in West Africa, in much the same way that an earlier French theorist, Henri Lévy-Bruhl, was led to the same conclusion by generalizing from the single experience of Roman slavery.11 Slavery among the primitive Goajiros of Venezuela, the large-scale slavery in Korea from the Koryŏ period to nearly the end of the Yi dynasty, and Russian slavery during the seventeenth and eighteenth centuries are three cases of slavery operating

9  Claude Meillassoux, L’esclavage en Afrique précoloniale (Paris: François Maspero, 1975), esp. pp. 11–26. 10  Ibid., pp. 20–21. 11  See Henri Lévy-Bruhl, “Théorie de l’esclavage,” in M. I. Finley, ed., Slavery in Classical Antiquity (Cambridge: W. Heffer and Sons, 1960), pp. 151–169.

Authority, Alienation, And Social Death

95

in different contexts and on very different scales as the result of a process of internal differentiation. In almost all premodern slaveholding societies, at least some slaves were locally recruited. The problems these slaves posed were no different from those presented by the more dramatically disrupted captives. What was different, however, was the manner of their social death. I suggest that there were two ways in which social death was represented and culturally “explained,” depending on the dominant early mode of recruiting slaves. Where the earliest and most dominant mode of recruitment was external, the cultural mode of representing social death was what I shall call intrusive and this was likely to continue even where, later, most slaves were internally recruited. The second way in which social death was represented may be called extrusive, and this too was determined by the earliest dominant means of recruiting slaves. It persisted even if, later, there was a shift to external sources. In the intrusive mode of representing social death the slave was ritually incorporated as the permanent enemy on the inside—the “domestic enemy,” as he was known in medieval Tuscany.12 He did not and could not belong because he was the product of a hostile, alien culture. He stood, on the one hand, as a living affront to the local gods, an intruder in the sacred space (the cosmicized circle, as Mireca Eliade would say, that defined the community).13 The views of the Bella Coola Indians of British Columbia and of the Nias of Indonesia are not only nearly identical, but typical of all peoples. The Bella Coola were fond of saying that “no slaves came to earth with the first people,” and Thomas F. McIlwraith comments: “To the Bella Coola, who still consider a man’s power in the land as dependent to a considerable extent on his ancestral myth, a slave’s greatest misfortune lay in the fact that he had no ancestral home, and hence no rights ... A slave was a stranger in a strange land, unsupported by a chain of ancestors reaching back to the beginning of time.”14 Similarly Peter Suzuki reports that among the Nias “the slaves are not mentioned in any ancestral myth, have no place in the world-tree, thus lack religion and consequently, a place in the cosmos. They have no past nor future, living as they do, on the whims and mercy of their masters. They live on the fringes of the cosmos and are viewed as being almost on a par with animals.”15 12  Iris Origo, “‘The Domestic Enemy’: The Eastern Slaves in Tuscany in the Fourteenth and Fifteenth Centuries,” Speculum 30 (1955): 321–366. 13  See Mircea Eliade, The Sacred and the Profane (New York: Harvest Books, 1959), pp. 20–65. 14   Bella Coola Indians (Toronto: University of Toronto Press, 1948), vol. 1, p. 159. 15   The Religious System and Culture of Nias, Indonesia (The Hague: Uitgeverij Excelsior, 1959), p. 45.

96

Patterson

On the other hand, the slave was symbolic of the defeated enemy, the power of the local gods, and the superior honor of the community. Because of the association of the slave with the enemy in this mode of representing social death, we are not surprised to find that slavery was associated with the military, and that the terminology of slavery took on a military flavor in many such societies. Among the Kwakiutl Indians of the northwest coast of America, “a slave is designated by the expression ‘q!aku q!ak’o,’” the basis of which is the root “q!ak,” meaning “to cut off the head.” U. P. Averkieva observes: The custom of cutting off the heads of slain enemies and carrying them away as a sort of trophy, which existed side by side with [the practice] of enslavement, bears witness to the fact that, whereas in the distant past an enemy [taken prisoner] had his head cut off, because, as yet, there was no place in society for a slave, he later began to be inducted into slavery.16 The Ashanti of West Africa, like the peoples of early Mesopotamia, referred to slaves as people of a foreign country. Indeed, adonke, the general term for slave in Ashanti, was the same term for all foreign northerners; and in the Third Dynasty of Ur the word for slave literally meant man or woman of the mountain, the area from which the earliest slaves came.17 The Greek word for slavery, doulos, is still an etymological mystery, but it is significant that in spite of the highly commercial nature of Greek slavery in classical times and the fact that from the sixth century bc on the vast majority of slaves were bought at slave markets rather than captured, the agent of the state responsible for the public regulation of slaves was the war archon.18 The Roman experience was even more revealing. P. R. C. Weaver, in his discussion of the servus vicarius, tells us that the term “is derived, as is much of the 16   Slavery among the Indians of North America (Moscow: U.S.S.R. Academy of Sciences, 1941), p. 80. 17  On Ashanti, see Robert S. Rattray, Ashanti Law and Constitution (Oxford: Clarendon Press, 1929), p. 29. On Ur see Bernard J. Siegel, Slavery during the Third Dynasty of Ur, Memoirs of the American Anthropological Association, no. 66 (1947), pp. 1–54. Siegel, after examining the available data, asserts, “We can thus conclude that the earliest notion of ‘slave’ was incorporated with the idea of ‘foreigner,’” pp. 8–9. This linguistic usage persisted even when the vast majority of slaves were recruited from impoverished families. On the sources of slaves see pp. 9–27. 18  On the words used for “slaves” and their sources see William L. Westermann, The Slave Systems of Greek and Roman Antiquity (Philadelphia: American Philosophical Society, 1955), pp. 5–12. Also M. I. Finley, “Was Greek Civilization Based on Slavery?” in Finley, Slavery in Classical Antiquity, p. 146.

Authority, Alienation, And Social Death

97

domestic terminology of Roman slavery, from military usage and organization” (emphasis added). A common term for slave was “captivus.”19 Roman law fully represented the intrusive conception of the slave. The Roman captured by the enemy lost all claims as a Roman citizen, but if he escaped and found his way back home, the principle of postliminium applied: he was fully restored to his original status, subject to a few restrictions and occasionally to a redeemer’s lien.20 The idea of social death was also given direct legal expression in Roman law. The slave was pro nullo. We learn, too, from the comedies of Plautus and Terence that the slave was one who recognized no father and no fatherland.21 Hebrew slavery in law and practice, in both ancient and medieval times, was highly intrusive. Fellow Jews could be and were enslaved in biblical times, but the slave was conceived of as the quintessential enemy within. In Leviticus we read: And as for thy bondsmen, and thy bondsmaids, which thou shalt have of the nations that are round about you, of them all shall ye buy bondsmen and bondsmaids. Moreover of the children of the strangers that do sojourn among you, of them shall ye buy, and of their families that are with you, which they have begotten in your land; and they shall be your possession. And ye shall make them an inheritance for your children after you, to hold for a possession; of them shall ye take your bondsmen forever.22 The foreign slave according to Maimonides “is like land in regard to the acquisition of title,” and one who was a minor “is like cattle, and one may acquire title to him by the modes whereby title to cattle ... is acquired.”23 Medieval Christendom from its very early days defined all pagans and infidels who resisted conversion as enemies who could justly be enslaved if taken in war. Like the Hebrews, the medieval Christian nations permitted the 19  P. R. C. Weaver, “Vicarius and Vicarianus in the Familia Caesaris,” Journal of Roman Studies 54 (1964): 118. 20  W. W. Buckland, The Roman Law of Slavery (Cambridge: Cambridge University Press, 1908), pp. 291–312. 21  See Peter P. Spranger, Historische Untersuchungen zu den Sklavenfiguren des Plautus und Terenz (Wiesbaden: Akademie Mainz, 1961), p. 65. 22  Lev. 25:44. 23  Maimonides, The Code: Book Twelve, The Book of Acquisition, ed. Isaac Klein (New Haven: Yale University Press, 1951), p. 809, and on the laws concerning heathen slaves, pp. 264–282.

98

Patterson

enslavement of fellow Christians and denied that the conversion of slaves obliged masters to manumit them.24 It is in Islamic religious and social thought that we find the purest expression of the intrusive conception of social death. The outsider was foreigner, enemy, and infidel, fit only for enslavement after the jihad, to be incorporated as the enemy within. Legally the Muslim is not permitted to enslave coreligionists, although, as we shall see, many ways were found to get around this injunction. As a cultural mode of representation, however, the image of the slave as the captured enemy and internalized outsider in a state of social death was firmly fixed in Islamic thought. The most frequent expression for female slaves in the Qoran is “that which your right hand possesses.” The slave is primarily “a person taken captive in war, or carried off by force from a foreign hostile country, and being at the time of capture an unbeliever.”25 Ali Abd Elwahed argued forcefully that in contrast to the basically ethnic conception of the slave’s distinctiveness found in western slave societies, both ancient and modern, the Islamic world’s conception was based on religious differences. He admitted that there were strong traces of racism in both the political and legal thought of the Arabs, but insisted that in their “collective representations” slavery was the result of captivity occasioned by just wars against the infidel.26 Similarly, M. G. Smith has emphasized this difference in the representation of slavery among Islamic and West Indian slavemasters.27 More recently, Paul Lovejoy has called attention to the need to distinguish between ideology and practice in the interpretation of slavery among Islamic peoples.28 Quite apart from the problem of confusing ideology with reality, an overemphasis on the religious content of the Islamic mode of representing the social death of the slave has tended to obscure the more important common element in the Western and Islamic representations: the fact that they are both intrusive. In sharp contrast with the intrusive conception of death was the extrusive representation. Here the dominant image of the slave was that of an insider 24  See Maurice Keen, The Laws of War in the Late Middle Ages (London: Routledge & Kegan Paul, 1965), p. 137; and David Brion Davis, The Problem of Slavery in Western Culture (Ithaca, N.Y.: Cornell University Press, 1966), pp. 48, 100–101. 25  Robert Roberts, The Social Laws of the Qorân (London: Williams & Norgate, 1925), p. 54. 26  Ali Abd Elwahed, Contribution à une théorie sociologique de l’esclavage (Paris: Mechelinck, 1931), pp. 139, 166–167. 27  “Slavery and Emancipation in Two Societies,” in M. G. Smith, ed., The Plural Society in the British West Indies (Berkeley: University of California Press, 1965), pp. 116–161. 28  “Conceptions of Slavery in the Nineteenth Century Sokoto Caliphate,” paper presented at the Conference on the Ideology of Slavery in Africa, York University, Toronto, April 3–4, 1980.

Authority, Alienation, And Social Death

99

who had fallen, one who ceased to belong and had been expelled from normal participation in the community because of a failure to meet certain minimal legal or socioeconomic norms of behavior. The destitute were included in this group, for while they perhaps had committed no overt crime their failure to survive on their own was taken as a sign of innate incompetence and of divine disfavor. Typical of the extrusive representation of social death among primitives were the Goajiro of Venezuela, among whom slavery was essentially “a consequence of the violation of the code of social order.”29 Among advanced archaic civilizations the Aztecs, Egyptians, and Chinese were typical. The Aztecs, while they took many prisoners of war, used them mainly in their religious ceremonies or else resettled them. Slavery was viewed as being of internal origin, and the slave was someone who had fallen as a result of destitution or criminality.30 In pharaonic Egypt the terminology of slavery contrasted strikingly with that of early Mesopotamia in that it did not refer to the slave as a foreigner. Egyptian terminology accurately reflected the internal source of slavery and the fact that it arose primarily from destitution. To the Egyptians this status amounted to social and legal death, as ‘Abd al-Muhsin Bakīr clearly shows.31 And it was into this status that captives who were enslaved were assimilated. Significantly, the Egyptian word for captive, literally translated, meant “living dead.”32 China throughout its long recorded history held firmly to an extrusive conception of slavery. The slave was conceived of as a criminal, and the prisoner of war, if enslaved, was legally and ideologically assimilated to the status of the internal criminal.33 In none of the above-mentioned societies do we find really large-scale slavery, so it may be wondered whether the extrusive mode of representing slavery applies only to social systems in which the institution did not attain marked structural significance. This, however, is not the case. There are two quite dramatic cases of advanced societies highly dependent on slavery in which the institution was intrusively represented: these are Korea during the Koryŏ and 29   Virginia Gutierrez de Pineda, Organización social en la Guajira (Bogota: Instituto Etnologico Nacional, 1950): 172. 30  Carlos Bosch Garcia, La esclavitud prehispánica entre los Aztecas (Mexico City: Colegio de Mexico, Centro de Estudios Históricos, 1944), p. 22. 31   Slavery in Pharaonic Egypt (Cairo: L’Institut français d’archéologie orientale, 1952), chap. 2. 32  Helmut Wiesdorf, Bergleute und Hüttenmänner im Altertum bis zum Ausgang der römischen Republik: Ihre wirtschaftliche, soziale, und juristische Lage (Berlin: AkademieVerlag, 1952), p. 63. 33  E. G. Pulleyblank, “The Origins and Nature of Chattel Slavery in China,” Journal of the Economic and Social History of the Orient 1 (1958): 204–211.

100

Patterson

Yi dynasties, and Russia from the late seventeenth century to near the end of the eighteenth century. Although it is not generally known even among scholars specializing in the study of slavery, both Korea and Russia relied heavily on slaves not only in their economic sectors but for the performance of administrative roles, and Korea at varying periods had slave populations that constituted more than 30 percent of its total population.34 In Korea during the Koryŏ period slavery had a “moral ... connotation”; slaves were persons from whom heaven had withdrawn its favor. In 1300 King Chungnyŏl of Korea responded in alarm to a draft of a plan by a Chinese, Kno-li Chi-su, to reform the system of slavery. The king explained thus: “Our ancestors have instructed us that these servile elements are of a different race and therefore it is not possible for them to become common people—to oppose the instructions of our ancestors is to endanger our social order.”35 Five hundred years later the conception of the slave was much the same. Both Susan Shin and Edward W. Wagner, from their studies of the census data of the late seventeenth century, found that social mobility was “overwhelmingly downward” and that the slave who stood at the bottom of the hierarchy was essentially someone who had fallen there.36 34  On Korea during the Koryŏ period, the major study in English is Ellen S. Unruh, “Slavery in Medieval Korea” (Ph.D. diss., Columbia University, 1978). Two useful general histories that deal with slavery during this period are Han Woo-Keun, History of Korea (Seoul: EulYoo Publishing Co., 1970); and Takashi Hatada, A History of Korea (Santa Barbara, Calif.: ABC-Clio Press, 1969). On the Yi dynasty the major relevant studies in English are Susan S. Shin, “Land Tenure and the Agrarian Economy in Yi Dynasty Korea: 1600–1800” (Ph.D. diss., Harvard University, 1973); John Somerville, “Success and Failure in Eighteenth Century Ulsan: A Study in Social Mobility” (Ph.D. diss., Harvard University, 1974); and Edward W. Wagner, “Social Stratification in Seventeenth-Century Korea: Some Observations from a 1663 Seoul Census Register,” in Occasional Papers on Korea 1 (1974): 36–54. Other works, including several in Korean (translated for the author) are cited in later references.  On Russian slavery the most important work in English is that of Richard Hellie, Slavery in Russia, 1450–1725 (forthcoming); see also his “Recent Soviet Historiography on Medieval and Early Modern Russian Slavery,” Russian Review 35 (1976): 1–32. Of special interest are two other works: George Vernadsky, “Three Notes on the Social History of Kievan Russia,” Slavonic Review 22 (1944): 81–92; and (although it is based entirely on secondary sources) J. Thorsten Sellin, Slavery and the Penal System (New York: Elsevier, 1976). 35   Cited in Herbert Passin, “The Paekchŏng of Korea,” Monumenta Nipponica 12 (1956–1957): 31. 36  In addition to the works by Shin and Wagner cited above see Susan S. Shin, “The Social Structure of Kùmhwa County in the Late Seventeenth Century,” in Occasional Papers on Korea 1 (1974): 9–35.

Authority, Alienation, And Social Death

101

In Russia we find another important slaveholding society with an extrusive conception of slavery. One of the earliest sets of laws dealing with slavery in Russia dates back to the second half of the twelfth century.37 The law listed three ways in which persons became slaves and it is significant that they excluded capture at war. This exclusion has puzzled Russian historians, and all sorts of theories have been advanced to explain the presumed omission, the most widely accepted being that the clauses in question dealt only with cases where a person becomes a slave by his own action. This may be the correct explanation, but equally plausible is another: in the same way that the enslaved prisoner of war in imperial China was assimilated to the status of a person who had become a slave as the result of conviction for a capital offense, so in Kievan Russia the captive may well have been assimilated to the dominant extrusive conception of the slave as an internally fallen person. Even more revealing is the controversy surrounding the connotation of the term “izgoi.” The term referred to aliens and freedmen but, significantly, its primary meaning was “a man who has lost his former status and is in need of special protection.” In this regard it applied as much to orphaned princes and bankrupt merchants as to destitute ex-slaves and aliens.38 Thus we find the alien being assimilated to the status of the fallen insider rather than the other way around. After its virtual disappearance during the early seventeenth century slavery began to expand again during the era of Peter the Great and continued doing so until Russia became one of the most important of the European slaveholding states, including those of the New World. It remained unique among the European slave systems, however, in maintaining a highly extrusive conception of slavery. As in imperial China, slavery was very closely tied to the penal system and the slave was conceived of as someone who had committed a capital offense. Not all criminals became slaves, but the kátorshniki and poseléntsi who were sentenced to a lifetime of hard labor and forced colonization were public slaves in every sense of the term. “Both were by their sentences deprived of all civil rights. In the eyes of the law they were nonpersons; their property was distributed to their heirs; their wives could remarry since all family relations had been annulled by the sentences.”39 The extrusive conception of slavery applied equally to private slaves who served their masters in urban areas. Earlier we saw how the three monotheistic religions reinforced an intrusive conception of slavery. It is therefore 37  Vernadsky, “Three Notes,” pp. 81–82. 38  Ibid., pp. 88–92. 39  Sellin, Slavery and the Penal System, p. 121.

102

Patterson

highly significant that Russia was the only Christian state whose church did not help to define the slave as a converted infidel. The Orthodox church, according to Richard Hellie, “condoned, and in fact, encouraged, the enslavement of Orthodox by Orthodox,” and it did not object to the enslavement of Orthodox Christians by members of other faiths.40 This becomes all the more extraordinary when we realize that in Muscovy national consciousness was expressed mainly in religious terms: “the Orthodox Church played a central role in the rise and consolidation of the Muscovite state.”41 With slavery retaining its highly extrusive character in Russia, the slave was never the enemy within but the internally fallen. Ideological elaboration of the difference between slave and free did not seek the aid of religion but defined the gulf in terms of what Hellie calls “simulated barriers.” One of these barriers is most revealing. Slave owners invented genealogical “claims of foreign origin for their clan.” They claimed to be foreigners of noble dynastic origins “reigning over another people.”42 Almost all of these claims were false, but it is remarkable that the Russian slaveholder, instead of defining his slave as the captured foreigner within his land, chose exactly the opposite course in defining himself as the foreigner of noble ancestry. This of course is consistent with an extrusive conception of slavery, where the slave is the insider who has fallen. We may summarize the two modes of representing the social death that was slavery by saying that in the intrusive mode the slave was conceived of as someone who did not belong because he was an outsider, while in the extrusive mode the slave became an outsider because he did not (or no longer) belonged. In the former the slave was an external exile, an intruder; in the latter he was an internal exile, one who had been deprived of all claims of community. The one fell because he was the enemy, the other became the enemy because he had fallen. At one extreme, even when prisoners of war became the major source of slaves in China during the period of the Northern dynasties, the representation of the slave as the internally fallen, the criminal, persisted; at the other extreme, in dynastic Mesopotamia as late as the Third Dynasty of Ur, when the vast majority of slaves were internally recruited, the intrusive representation of the slave as the defeated enemy, the people “from the mountain,” endured. It is precisely this persistence of one conception of slavery during periods when we might normally expect the other which explains many otherwise puzzling aspects of the study and treatment of the subject of slavery. Let me 40  Hellie, Slavery in Russia, msp. XI-10. 41  Ibid., msp. XI-9. 42  Ibid., pp. XI-10–XI-11.

Authority, Alienation, And Social Death

103

illustrate with one important example. It is almost universally believed by European and American writers and readers of history that slavery was abolished in the northern part of Western Europe by the late Middle Ages. Yet in France, Spain, England, and the Netherlands a severe form of enslavement of Europeans by Europeans was to develop and flourish from the middle of the fifteenth century to well into the nineteenth. This was penal slavery, beginning with galley slavery and continuing with its replacement by the Bagnes, or penal slavery in public works. Both were slavery in every sense of the term. They developed as substitutes for the death penalty at a time when there was not a prison system in Europe to accommodate the huge number of persons found guilty of capital offenses. To be sure, the growing incidence of such offenses was largely a reflection of the increase in the number of acts legally so defined. Indeed, there is growing evidence that the legal redefinition of crime and the resulting increase in penal and public slavery was largely determined by the need to regulate labor.43 It is truly extraordinary that European scholars have either neglected this whole aspect of the subject or defined it as something other than slavery when they have recognized it. When we look for reasons, it is too easy to claim that there has been a conspiracy of silence, or worse, a deliberate attempt to distort the historical facts. My own feeling is that there has been a genuine failure to recognize the institution for what it was owing to the pervasiveness of the intrusive conception of slavery in the Western intellectual consciousness. The same framework may explain the neglect of modern Russian slavery by West European scholars. Galley slavery and slavery in the Bagnes are immediately recognizable to anyone who understands the institutions in extrusive terms. When the King of France issued a royal letter to his judicial authorities requesting them to provide for the galleys “all malefactors ... who have merited the death penalty or corporeal punishment, and also those whom they could

43  I draw heavily on several works in making these assertions, particularly Sellin, Slavery and the Penal System, and Michael R. Weisser, Crime and Punishment in Early Modern Europe (Sussex: Harvester Press, 1979). Sellin’s work is itself largely an elaboration of the thesis that “the demands of the labor market shaped the penal system and determined its transformation over the years, more or less unaffected by theories of punishment in vogue.” This thesis had been developed in two earlier works, those of Georg Rusche and Otto Kirchheimer, Punishment and Social Structure (New York: Columbia University Press, 1939); and Gustav Radbruch, “Der Ursprung des Strafrechts aus dem Stande der Unfreien,” reprinted in Elegantiae juris criminalis (Basel: Verlag für Recht und Gesellschaft, 1950). See also Sidney W. Mintz, “The Dignity of Honest Toil: A Review Article,” Comparative Studies in Society and History 21 (1979): 558–566.

104

Patterson

conscientiously declare to be incorrigible and of evil life and conduct,”44 he was issuing a call for more public slaves in much the same way that an oriental or Russian monarch would have done. The only difference was that the oriental or Russian ruler would have known that he was requesting more slaves and felt no qualms about it, whereas the French king, with his intrusive view of slavery, either believed he was requesting some other category of labor or conveniently persuaded himself that he was.45

Liminal Incorporation

Although the slave might be socially dead, he remained nonetheless an element of society. So the problem arose: how was he to be incorporated? Religion explains how it is possible to relate to the dead who still live. It says little about how ordinary people should relate to the living who are dead. This is the final cultural dilemma posed by the problem of slavery. James H. Vaughan, in his analysis of slavery (mafakur) among the Margi of Nigeria, has addressed this problem with considerable insight.46 He tells us that traditional Margi society was “in theory, a closed system, recognizing birth as the only method of recruitment.” Any outsider was an intruder into this social space and must remain an alien; but, equally, the insider who committed some capital crime offended the gods and his ancestors and in so doing broke society’s invisible boundaries and made himself an alien. The population of slaves among the Margi comprised both types of aliens, although the dominant representation of their social death was intrusive. The rich diversity of groups surrounding the Margi make them particularly aware of their social space. As Vaughan observes: “They are sensitive to a unifying ‘Marginess’—largely consensual—that distinguishes them from the numerous other societies around them,” and slaves are those who have breached “the boundaries of this closed system.” The institution of slavery “bestows a

44  Cited in Sellin, Slavery and the Penal System, p. 47. 45  Not all contemporary penal reformers attempted to disguise the fact that the punishment they were calling for as a replacement for the death penalty was slavery, pure and simple. Thus the Milanese noble and penal reformer Cesare Beccaria in his influential tract, Of Crime and Punishments, stated bluntly that the alternative to the death penalty that he was advocating was “slavery for life.” See Sellin, Slavery and the Penal System, pp. 65–69. 46  “Mafakur: A Limbic Institution of the Margi,” in Suzanne Miers and Igor I. Kopytoff, eds., Slavery in Africa (Madison: University of Wisconsin Press, 1977), pp. 85–102.

Authority, Alienation, And Social Death

105

rational—even utilitarian—place upon the anomaly of the permanent resident alien, by giving him an institutional marginality.” Furthermore: The outstanding general characteristic of mafakur is that all mafa, without regard to political position, private influence, or wealth, hold in common a status that in structural terms is fundamentally and irrevocably intermediate with regard to membership in Margi society. But it is equally apparent that, despite their marginal status, their roles are fully integrated into society.47 Thus slavery involved two contradictory principles, marginality and integration, and Margi society reconciled this contradiction by “formalizing the marginality.” Hence Vaughan calls the institution “limbic” (I prefer the more common anthropological term “liminai”) “for its members exist in the hem of society, in a limbo, neither enfranchised Margi nor true aliens.” But the Margi also enslaved local offenders, and these too were assimilated to the same limbic or liminal status of the institutionalized outsider. The criminal “remained in the society: a part of it, yet apart from it. He was not [physically] expelled, for that would be less humiliating ... Rather, it was the loss of identity and normality that was so objectionable to the proud Margi.” Institutionalized marginality, the liminai state of social death, was the ultimate cultural outcome of the loss of natality as well as honor and power. It was in this too that the master’s authority rested. For it was he who in a godlike manner mediated between the socially dead and the socially alive. Without the master, as the Tuareg insist, the slave does not exist. The slave came to obey him not only out of fear, but out of the basic need to exist as a quasi-person, however marginal and vicarious that existence might be. There were other gains to the master, as well as to the other members of the community, in the slave’s liminality. The marginal person, while a threat to the moral and social order, was often also essential for its survival. In cultural terms the very anomaly of the slave emphasized what was most important and stable, what was least anomalous in the local culture of the non-slave population. This was particularly true of small-scale, highly integrated societies with little class division among the nonslave population. Theda Perdue makes this point in her discussion of the precontact phase of Cherokee slavery. Before they adopted the institution of plantation slavery from the whites, the Cherokees kept slaves; but they contributed nothing to the economic, political, or social life of their warfare-oriented communities. 47  Ibid., p. 100.

106

Patterson

Why then were slaves kept? Perdue’s explanation is that the traditional belief system of the Cherokees rigidly categorized the social and physical universe. As with all systems of categorization, however, there were many anomalies that simply did not fit. The Cherokee way of handling such exceptions was to emphasize them, on the principle that it was precisely what stood at the margins that emphasized the boundaries. The slave acquired the same cultural significance among them as the bear—a four-footed animal, which nonetheless had the human habit of standing on its hind legs and grasping with its two front paws—or the Uktena, the mythical beast, “which had the body of a snake, the antlers of a deer and the wings of a bird.” Similarly the atsi nahsa’i, or slaves, were utterly anomalous; they had the shape of human beings but had no human essence whatever, since humanness was defined in terms of belonging to a clan. The slave, in not belonging, emphasized the significance of belonging; in being clanless, emphasized the clan as the only basis of belonging; in being deviant, “helped establish and strengthen group identity among the Cherokees.”48 We get a fascinating glimpse of the symbolic significance of the liminal slave in a more complex social system fraught with conflict in the Anglo-Saxon epic, Beowulf.49 Beowulf’s world was one riddled with internal feuding and external warfare. It was also a social order with a highly developed class system, in which an aristocratic warrior class lived off the surplus generated by its servants. Slavery and the slave trade were an integral part of this world. Indeed, literally the first event we come across in the prologue is a reference to the Danish hero Shild, who “made slaves of soldiers from every land, crowds of captives he’d beaten into terror.” In addition to social division there was a fundamental cleavage in the moral order of the world of Beowulf, between the old paganism and the newly acquired and not yet fully integrated Christianity. In the poem this is expressed in terms of the conflict between the forces of good and those of evil. This is not a tidy organic world; evil and conflict are ever present and recognized as such: The world, And its long days full of labor, brings good And evil; all who remain here meet both.50 48  Theda Perdue, Slavery and the Evolution of Cherokee Society, 1540–1866 (Knoxville; University of Tennessee Press, 1979), pp. 3–18. 49  Burton Raffel, ed. and trans., Beowulf (New York: New American Library, 1963). All citations are from this edition. 50  Ibid., lines 1060–62.

Authority, Alienation, And Social Death

107

The role of the slave in Beowulf’s last and greatest battle is replete with symbolic meaning. First, it is significant that it was a runaway slave, beaten by his master, who in searching for a place to hide found “the hidden path” to the sleeping dragon, “awoke him from his darkness and dreams and brought terror to his [Beowulf’s] people.”51 There is thus a forceful juxtaposition of the most pronounced social conflict (that between master and brutalized slave), with the most deep-seated moral conflict (that between Christianity and the lurking forces of evil and paganism, symbolized by the hidden “heathen treasure” and the dragon that protects it). So mankind’s enemy, the mighty beast Slept in those stone walls for hundreds of years; a runaway slave roused it, Stole a jeweled cup and bought His master’s forgiveness, begged for mercy And was pardoned when his delighted lord took the present He bore, turned it in his hands and stared At the ancient carvings. The cup brought peace To a slave, pleased his master, but stirred A dragon’s anger.52 Not only is there a symbolic association of social and moral conflict, but in the role of the slave as guide to the dragon’s evil world we find one of the most remarkable statements of the slave’s liminal status. It is significant that the slave was not counted among the twelve men who went to the dragon’s den. And there might be some hint of his anomalous nature in the fact that he was the thirteenth person. It was precisely because he was marginal, neither human nor inhuman, neither man nor beast, neither dead nor alive, the enemy within who was neither member nor true alien, that the slave could lead Beowulf and his men across the deadly margin that separated the social order above from the terror and chaos of the underground, between good and evil, between the sacred world of the Christian and the profane world of the pagan. A consideration of the important role of the slave’s liminality brings us to an important feature of slavery that is often misunderstood. Although the slave is socially a nonperson and exists in a marginal state of social death, he is not an outcaste. The point must be emphasized in view of the easy use often made of 51  Ibid., lines 2210–11. 52  Ibid., lines 2279–87.

108

Patterson

the caste concept in interpreting American slavery and its postemancipation consequences.53 With the noteworthy exception of temple slaves, enslaved persons are never relegated to the status of an outcaste group, nor are they ever stratified as one of several castes in any of the societies that have a complex hierarchy of castes. Before explaining why, let us look at the nature of the relationship between caste and slavery. The Ethnographic Atlas (see Appendix B) classifies the 186 societies of the Murdock World Sample with respect to “caste stratification.” The four groupings are societies where: (1) Caste distinctions of any kind are absent or insignificant. (2) There are one or more despised occupational groupings (whether smiths or leather workers or whatever else), distinguished from the general population, regarded as outcastes, and characterized by strict endogamy. (3) There is ethnic stratification in which a superordinate caste withholds privileges from and refuses to intermarry with a subordinate caste (or castes), which it stigmatizes as ethnically alien (for example, as descended from a conquered and culturally inferior indigenous population, from former slaves, or from foreign immigrants of different race and/or culture). (4) Complex caste stratification exists, in which occupational differentiation emphasizes hereditary ascription and endogamy to the near exclusion of achievable class status. The main advantage of this classification is that it takes account of both the narrow and the wide definitions of caste. Many scholars would hold that the term “caste” strictly applies only to societies in category (4), confined mainly to India and the related societies of Southeast Asia.54 Others, who hold to the more general interpretation of the caste concept, would include societies in category (2) along with (4) as legitimate cases.55 My own position is closer to 53  For the classic statement see John Dollard, Caste and Class in a Southern Town (New Haven: Yale University Press, 1937). 54  For one of the best statements of this view see Edmund R. Leach, Aspects of Castes in South India, Ceylon, and North-West Pakistan (Cambridge: Cambridge University Press, 1960). 55  Typical of this approach is James H. Vaughan who, “following anthropological usage,” defines caste as “a hereditary endogamous group who are socially differentiated by prescribed behavior.” See his “Caste Systems in the Western Sudan,” in Arthur Tuden

109

Authority, Alienation, And Social Death

the second, with the important qualification that for me caste additionally connotes some notion of ritual purity and pollution as a means of maintaining social distance. The existence of “hereditary endogamous groups” that are “socially differentiated by prescribed behavior” is a necessary but not sufficient criterion for defining castes, for this description is true of almost all class systems. In the light of this definition it is clear that I do not accept category (3) of the Ethnographic Atlas grouping as containing genuine caste systems, since ethnic differentiation need not be reinforced by notions of ritual purity and pollution. With these observations in mind we can now examine the relationship between the presence or absence of slavery and types of caste stratification. This is reported in Table 2.1. The table is highly significant (p = 0.002), although there is no strong overall relationship. Most slaveholding societies, we see, do not have castes of any kind. Yet slavery is not incompatible with the existence of castes. There is a weak overall relationship with slavery as the dependent variable: the moderately strong association with occupational castes is balanced by a rather weak relation with complex caste systems. Table 2.1

The relation of slavery to caste stratification, as delineated by the Murdock Ethnographic Atlas Type of society

Presence or absence of slavery

Caste absent

Occupational groupings

Ethnic stratification

Complex caste differentiation

Slavery absent: Number Percent

114  66.3

 6  3.5

1 0.6

4 2.3

Slavery present: Number Percent

 33  19.2

10  5.8

2 1.2

2 1.2

note: Chi square = 14.17, with three degrees of freedom. Significance = 0.0027.

and Leonard Plotnicov, eds., Social Stratification in Africa (New York: Free Press, 1970), pp. 59–92. See also in the same volume Jacques Maquet, “Rwanda Castes,” pp. 93–124.

110

Patterson

More important is what is revealed by the ethnographic data on those societies which have both slavery and caste. In none of them (the rare cases of temple slavery excepted) were slaves either outcastes or segregated as distinct castes. Typical of slaveholding societies with occupational castes were the Margi, the Somali, and the Koreans. The rich historical and anthropological data on these societies indicate that slaves were held to be distinct from the caste groups in question.56 There was never any marriage, or even illicit sexual relations, between the outcaste group and ordinary persons, whereas such relations were common between “free” males and slave women. It is typical of the boundarycrossing capacity of slaves that among the Somali they were the only persons who could have sexual relations and marry with both ordinary “free” Somalis and the outcaste Sab group. Furthermore, the outcaste groups could never lose their caste status, nor did they want to, while in all these societies slaves could be manumitted and become “free” persons. Third, the outcaste groups were usually segregated. The Korean paekchŏng, for example, had a high degree of internal autonomy, living in their own communities as an organized outcaste group. Slaves were never segregated simply because they were slaves. Fourth, the outcaste groups all had a monopoly of certain occupations in which they specialized; slaves were never confined to particular jobs. In all premodern societies they performed virtually the entire range of occupations, and even in modern capitalistic slave systems recent studies have indicated that the range of their occupations was much wider than previously thought.57 Perhaps the most important difference is that while slaves may have been held in contempt, they were never avoided or feared because it was felt that they were polluting. The Sab of Somalia and the paekchŏng of Korea, however, were avoided for this reason.58 It is not difficult to understand why slaves were never assimilated to the status of outcastes. Slavery, we have seen, was primarily a relation of personal 56  On the Margi see Vaughan’s two previously cited papers: “Mafakur: A Limbic Institution of the Margi,” and “Caste Systems in the Western Sudan”; on the Somali see Enrico Cerulli, “Il diritto consuetudinario della Somalia settentrionale (Migiurtini),” Somalia, scritti vari editi ed inediti 2 (1959): 1–74. On Korea see Passin, “The Paekchŏng of Korea.” 57  See Robert S. Starobin, Industrial Slavery in the Old South (New York: Oxford University Press, 1970); Claudia Dale Goldin, “The Economics of Urban Slavery, 1820–1860” (Ph.D. diss., University of Chicago, 1972); and Stanley L. Engerman, “A Reconsideration of Southern Economic Growth, 1770–1860,” in Agricultural History 49 (1975): 343–361. On Jamaica see Barry Higman, Slave Population and Economy in Jamaica, 1807–1834 (Cambridge: Cambridge University Press, 1976), chaps. 2–4, 10. 58  While the vast majority of outcaste groups were despised, a few were not, among them the enkyagu of the Margi.

Authority, Alienation, And Social Death

111

domination. There was an almost perverse intimacy in the bond resulting from the power the master claimed over his slave. The slave’s only life was through and for his master. Clearly, any notion of ritual avoidance and spatial segregation would entail a lessening of this bond. Second, the assimilation of the slave to the status of an occupationally specialized caste would undermine one of his major advantages—the fact that he was a natally alienated person who could be employed in any capacity precisely because he had no claims of birth. Slaves universally were not only sexually exploited in their role as concubines, but -also in their role as mother-surrogates and nursemaids. However great the human capacity for contradiction, it has never been possible for any group of masters to suckle at their slave’s breast as infants, sow their wild oats with her as adolescents, then turn around as adults and claim that she was polluted. Indeed, the comparative data indicate that in societies with highly developed notions of ritual pollution one of the main reasons for keeping slaves was that they were nonpolluting and thus a major means of circumventing pollution norms. Among the Maori, for instance, every free person had tapu, a complex set of prohibitions that were the laws of the gods. Breaking these severely endangered the individual, since he lost his mana (power) and became vulnerable to supernatural forces. According to Elsdon Best: The shadow of tapu lay over the Maori from birth until death, his very bones and their resting place remained tapu for all time. The higher the rank of a person the more tapu was he. It is interesting to note that slaves were held to be free from tapu and yet no explanation is given as to their condition of welfare and their survival, why they did not perish in such a defenceless condition.59 They did not perish, we now know, because as natally alienated persons they were socially dead. “Even though he [the slave] had once been a chief in another tribe,” Raymond Firth tells us, “his capture removed him from the mana of the gods and in things spiritual he ceased to count.” In this liminal state he could cross boundaries prohibited to other persons and could perform the vital task of preparing the master’s food, which if done by a mortal would result in certain spiritual and possible physical death.60 For much the same reasons we find in Nepal that slaves, while “politically the most debased section of Nepalese society,” were nonetheless sometimes 59   The Maori, Memoirs of the Polynesian Society, no. 5 (1924), p. 251. 60  Raymond Firth, Primitive Economics of the New Zealand Maori (Wellington, N.Z.: R. E. Owen, Government Printer, 1959), p. 214.

112

Patterson

selected from the higher castes. “Indeed, in order to perform the various duties imposed on domestic servants, to be permitted to cross the threshold of an owner’s dwelling, it was imperative for the slave to enjoy a degree of ritual purity conferred only by membership in certain castes.” Paradoxically, even Brahmins were enslaved without losing caste.61 A consideration of the relation of slavery to caste leads us back to where we began: the liminality of the slave is not just a powerful agent of authority for the master, but an important route to the usefulness of the slave for both his master and the community at large. The essence of caste relations and notions of ritual pollution is that they demarcate impassable boundaries. The essence of slavery is that the slave, in his social death, lives on the margin between community and chaos, life and death, the sacred and the secular. Already dead, he lives outside the mana of the gods and can cross the boundaries with social and supernatural impunity.

The Rituals and Marks of Enslavement

Symbolic ideas are usually given social expression in ritualized patterns. Let us look now at the ritual aspects of the natal alienation of the slave. For all but the most advanced slave systems the acquisition of a slave is a very special event in the master’s household. Even where slaves number as much as a quarter of the total household, their acquisition may be a once-in-a-lifetime event for the members, especially if the pattern of slaveholding is highly skewed. It was common for people in the premodern world to give ritual expression to special events and when one of those events involved the incorporation of a person defined as socially dead, it is easy to recognize that the event should not proceed without ceremony. The ritual of enslavement incorporated one or more of four basic features: first, the symbolic rejection by the slave of his past and his former kinsmen; second, a change of name; third, the imposition of some visible mark of servitude; and last, the assumption of a new status in the household or economic organization of the master. Many cultures obliged the new slave to make a symbolic gesture of rejecting his natal community, kinsmen, ancestral spirits, and gods—or, where the slave was of local origin, of rejecting his own kin group and ancestral spirits in favor of those of his master. The ceremony was often simple and brief, but it was always deeply humiliating, sometimes even traumatic, for the slave. 61  Lionel Caplan, “Power and Status in South Asian Slavery,” in James L. Watson, ed., Asian and African Systems of Slavery (Oxford: Basil Blackwell, 1980), pp. 177–180.

Authority, Alienation, And Social Death

113

Among the cannibalistic Tupinamba of South America we find slavery in its most primitive form. Most captives were eventually eaten, but in the many years between capture and execution the captives lived as the slaves of their captor and were usually well treated. Before they entered their captor’s village the captives were stripped, dressed as Tupinamba, and decorated with feather ornaments. They were led to the graves of persons who had recently died and forced to “renew” or cleanse the bodies. The captives then performed a vital ritual function. They were given the weapons and other belongings of the deceased to be used for a time, after which they were handed over to the rightful heirs. “The reason for this,” according to Alfred Métraux, “was that touching the belongings of a dead relative was fraught with dangers, unless they were first defiled by a captive.”62 Being socially dead, the captives were able to move between the living and the dead without suffering the supernatural harm inevitably experienced by the socially alive in such boundary crossing. After this ritual the prisoners were taken to the village, where their captivity was celebrated in song and dance, the captives themselves being forced to participate and “to dance in front of the hut where the sacred rattles were kept.”63 Among the more complex Germanic peoples of early medieval Europe, the new slave of local origin placed his head under his master’s arm and a collar or strap was placed around his neck.64 We find a variant of this in late AngloSaxon England, where a man who through poverty was forced to sell himself into slavery had to place his head between his new master’s hands; a billhook or an oxgoad was then given him to symbolize his new condition. This led to a special way of referring to enslavement, as when a Northumbrian mistress spoke of “all those people whose heads she took in return for their food in the evil days.”65 The expression gives a clue to the meaning of the ceremony; a man’s head is associated with his mind and will and it is these, in addition to his labor, that the master takes. If we look instead at traditional Africa, we find some interesting parallels and differences. The objective of the rituals was the same: to give symbolic expression to the slave’s social death and new status. But the emphasis was less on personal and spiritual labor and more on the social use of the slave 62  “The Tupinamba,” in Julian H. Steward, ed., Handbook of the South American Indians (Washington, D.C.: Government Printing Office, 1948), vol. 3, p. 120. 63  Ibid. 64  A. M. Wergeland, Slavery in Germanic Society during the Middle Ages (Chicago: University of Chicago Press, 1916), p. 16. 65  H. R. P. Finberg, The Agrarian History of England and Wales (Cambridge: Cambridge University Press, 1972), p. 507.

114

Patterson

incorporated as a permanent marginal into a network of affiliation after a ritual break with his old kin ties and ancestral protectors. The Imbangala of northwestern Angola are typical.66 All slaves, whether acquired from outside of Kasenje or within its boundaries, were considered alien to local lineages. In a special rite of passage the slave was first “cleansed” of his natal ties, by means of a medicine that denuded him of ancestral protection. Significantly, however, the medicine also eliminated any memory among the master’s lineage of the slave’s ancestry, so the very act of separation paved the way for the possible assimilation of the slave’s descendants. This was followed by a dangerous purgatorical period in which the new slave was spiritually exposed, lacking the protection of both former and prospective ancestral spirits. Finally, the slave was incorporated (though not adopted) into the master’s lineage via a naming ceremony in which he became an “alien dependent,” protected once again, but without the full complement of names that was the birthright of every true member of the lineage. The initiating ritual varied regionally, although its symbolic and practical objectives remained the same. Among the Kwanyama of southwestern Angola the rite was called elyakeko, which literally meant “to tread upon something.” The captive was taken by the parents of the warrior to the whetstone kept in all Kwanyama houses: The father takes the stone and holds it in his hand, while his wife pours water over the whetstone, water which the father forces the prisoner to drink. After this has been done, the prisoner’s master takes the stone and beats the victim on the top of the cranium with it, “to prevent him from having thoughts of escape.” As the stone is motionless by nature, the Kwanyama believe the person so treated comes to possess the same quality.67 Similarly, among the Tiv of central Nigeria, “the purchaser and the man’s agnates split a chicken which was held to sever the slave from his kin, thus making it impossible for him to run away, ‘for he would have no place to go.’”68 Some African groups like the Aboh offered sacrifice to special shrines, and 66  J. C. Miller, “Imbangala Lineage Slavery,” in Miers and Kopytoff, Slavery in Africa, pp. 205–233. 67  Carlos Estermann, The Ethnography of Southwestern Angola (New York: Africana Publishing Co., 1976), pp. 128–129. 68  Laura Bohannan and Paul Bohannan, The Tiv of Central Nigeria (London: Ethnographic Survey of Africa, 1953), pt. 8, pp. 45–46.

Authority, Alienation, And Social Death

115

feasted;69 others, such as the Ila, assigned the new slave “a spirit associated with a deceased member of the patrilineal group” and in a communal ritual the ancestors were informed of the newly affiliated slave and their protection was sought.70 The length of time for full adjustment to their new status varied with the kind of slave: it was usually easier for women, but occasionally the path was smoother for men (as among the Ila). Whatever the variations, in all African traditional societies the newcomer, unless he was a “trade slave” destined for resale, was forced to deny his natal kin ties and acquire certain fictive kin bonds to the master and his family. The exact meaning of his new ties will be examined later. The initiating ceremony served much the same purpose among kin-based societies in other parts of the world. Among the Kachin of highland Burma, for example, the shaved head of the new slave was rubbed with ashes from the master’s hearth prior to his incorporation into the master’s clan.71 Shorn of the memory of his past, the slave received the ashes of his master’s ancestral spirits. As a final example, we may take the Toradja of the central Celebes.72 As soon as he was brought into his master’s house, the slave was given a meal made of the same kind of food his master normally ate, “so that his life spirit will be tranquil.” The meal was usually served on the cover of a pot that was meant to help the slave forget his former attachments. Next, a little basket of rice, eggs, ubi, and coconut was prepared and was turned above the head of the slave seven times to the left and seven times to the right. The basket was then placed on the slave’s head and the master invoked as follows; “You, soand-so, wherever your life spirit may have gone, to your relatives left behind, here is rice which I give to you; eat it so that he may settle down on you and you may have a long life.” The slave ate the contents, after which a priestess usually came and invoked long life for the new slave. The symbolism here is self-evident and needs no commentary. Once again it involved the loss of independent social existence—of the slave’s “life spirit”—the placating of the lost spirits and protection against them, and the incorporation of the slave into the marginal existence of the permanent alien. 69  K. Nwachukwu-Ogedengbe, “Slavery in Nineteenth Century Aboh (Nigeria),” in Miers and Kopytoff, Slavery in Africa, p. 141. 70  Arthur Tuden, “Slavery and Social Stratification among the Ila of Central Africa,” in Tuden and Plotnicov, Social Stratification in Africa, p. 52. 71  Edmund R. Leach, Political Systems of Highland Burma (London: Bell, 1954), p. 304. 72  N. Adriani and Albert C. Kruyt, De Bare’e Sprekende Toradjas van Midden-Celebes [The Bare’e-speaking Toradja of Central Celebes] (Amsterdam: Nood-Hollandsche Uitgevers Maatschappij, 1951), vol. 2, p. 142.

116

Patterson

In large-scale slave systems where the slave became a unit of production outside the household economy we do not, of course, find such elaborate initiating rituals of enslavement. The newcomer was usually handed over to a trusted older slave to be taught the necessary skills to survive in his new environment. This is not to say, however, that ritual did not play a part even here. For we know that even in the brutal capitalistic slave plantations of the modern Caribbean, slaves had a rich ritual life and found their own ways of incorporating the new recruit.73 The same was very possibly true of slaves on the latifundia of ancient Rome, given the rich and intense religious life of the slave population. But if the slave was not incorporated privately by his master, there was still the need to incorporate him publicly, to give ritual expression to his presence as a large and significant, and potentially dangerous, element in the body politic. We shall see later that in such large-scale systems this task was performed by the state religion. The second major feature of the ritual of enslavement involved the changing of the slave’s name. A man’s name is, of course, more than simply a way of calling him. It is the verbal signal of his whole identity, his being-in-the-world as a distinct person. It also establishes and advertises his relation with kinsmen. In a great many societies a person’s name has magical qualities; new names are often received upon initiation into adulthood and into cults and secret societies, and the victim’s name looms large in witchcraft and sorcery practiced against him. As Ernst Cassirer observed: “The notion that name and essence bear a necessary and internal relation to each other, that the name does not merely denote but actually is the essence of its object, that the potency of the real thing is contained in the name—that is one of the fundamental assumptions of the mythmaking consciousness itself.”74 Thus it is understandable that in every slave society one of the first acts of the master has been to change the name of his new slave. One must reject any simplistic explanation that this was simply a result of the master’s need to find a name that was more familiar, for we find the same tendency to change names when slaves come from the identical society or language group as their masters.

73   On Jamaica see Orlando Patterson, The Sociology of Slavery: Jamaica, 1655–1838 (Rutherford, N.J.: Fairleigh Dickinson University Press, 1969), chap. 6. On the U.S. South see Eugene D. Genovese, Roll, Jordan, Roll (New York: Pantheon, 1974), esp. bk. 2. See also the detailed discussion of the slaves’ cultural life in Charles W. Joyner, “Slave Folklife on the Waccaman Neck: Antebellum Black Culture in the South Carolina Low Country” (Ph.D. diss., University of Pennsylvania, 1977), chap. 3. 74   Language and Myth (New York: Dover Publications, 1953), p. 3.

Authority, Alienation, And Social Death

117

There are several reasons for the change of name. The changing of a name is almost universally a symbolic act of stripping a person of his former identity (note for example the tendency among modern peoples to assign a new formal identification, usually a number, to both prisoners of war and domestic convicts). The slave’s former name died with his former self. The significance of the new name, however, varied from one kind of slave culture to another. Among most kin-based societies the slave took the clan name of his new master. This was the first act in the creation of fictive kin ties. The situation was different, however, among that small group of kin-based societies where the slave was not incorporated into the household economy but was exploited separately, in a protocapitalist sector, and in most of the advanced premodern slave systems. Here the new name was often a badge of inferiority and contempt. Sometimes the names were either peculiarly or characteristically servile. A Greek name in republican Rome, for example, often indicated slave status or ancestry, and many traditionally Roman names eventually became favorite slave names, cognomens such as Faustus, Felix, Fortunatus, and Primus.75 In Russia masters and slaves used the same names to a greater degree, which is understandable in the light of the local origins of most slaves: nonetheless, certain names such as Kondratii and Matrona became typical slave names.76 In other societies such as China, those of the ancient Near East and pharaonic Egypt, the absence of family names was the surest mark of slavery.77 Much more humiliating, 75  There were, however, many peculiarly servile names, the best-known being perhaps “Rufio.” This and other names suggest the national origins of the slaves, but as Gordon, Solin, and others have pointed out, it is dangerous to draw conclusions about the ethnic origins of Roman slaves on the basis of the available distribution of ethnic names. Slaves were often named for the place of purchase, which tells us nothing about their origin—a good case in point being the common slave name “Corinthus.” Greek or hellenized names were often taken for cultural reasons. In an exceptional case a captive was allowed to keep his original name, the most famous example being Spartacus. Whatever the new name, the overwhelming tendency was for the slave’s master or superior to select it. Principally for this reason slave names “do not often take the form of nicknames derived from physical characteristics.” See Mary L. Gordon, “The Nationality of Slaves under the Early Roman Empire,” in Finley, Slavery in Classical Antiquity, pp. 171–211; Lily Ross Taylor, “Freedman and Freeborn in the Epitaphs of Imperial Rome,” American Journal of Philology 82 (1961): 113–132; and, more recently, Heikki Solin, Beiträge zur Kenntnis der griechischen Personennamen in Rom (Helsinki: Societas Scientiarum Fennica, 1971). 76  Hellie, Slavery in Russia, mspp. XI-19–XI-27. 77  On the ancient Near East see Isaac Mendelsohn, Slavery in the Ancient Near East (Oxford: Oxford University Press, 1949), p. 31; on China see Pulleyblank, “The Origins and Nature of Chattel Slavery in China,” p. 217; on Egypt see Bakir, Slavery in Pharaonic Egypt, pp. 103–107, 114.

118

Patterson

however, were those cases in which insult was added to injury by giving the slave a name that was ridiculous or even obscene. Among the Duala of the Cameroon, slaves were given such names as “Irritation”; and among the Aboh of Nigeria, there were names like “Bluebeard” and “Downcast.”78 The Nootka of the northwest coast of America, the Icelanders, and the Kachin of highland Burma are all typical of peoples who took special delight in giving to female slaves names that demeaned both their status and their sex.79 Much the same pattern existed in the Americas, where the assignation and use of names was an important focus of conflict between masters and slaves. In the U.S. South slaves were sometimes whipped for using the forenames of important whites. The pompous classical names preferred by many planters were resented by most slaves, except when they were reminiscent of African names. Slaves usually changed their surnames after manumission, although sometimes, for purposes of protection, they kept the names of their ex-masters if they were important persons. Apparently many slaves selected their own surnames, which they used among themselves.80 In doing so they often took the names (or “entitles” as they called them) of distant ancestors or former masters, in a direct symbolic rejection of their present master. Herbert G. Gutman insists that most slaves had surnames, and that the choice of a different name involved, on the one hand, a rejection of the “intimacy” of the ties of paternalism claimed by the master and, on the other hand, “served to 78  Ralph A. Austen, “Slavery among the Coastal Middlemen: The Duala of the Cameroon,” in Miers and Kopytoff, Slavery in Africa, p. 312. See also K. Nwachukwu-Ogedengbe, “Slavery in Nineteenth Century Aboh (Nigeria),” in the same volume, p. 140. 79  Edward Sapir and Morris Swadesh, Native Accounts of Nootka Ethnography (Bloomington: Indiana University Research Center in Anthropology, Folklore, and Linguistics, 1955), p. 177. When a Nootka slave escaped or was ransomed, a potlatch was given for him and he was assigned a new name. Carl O. Williams, Thraldom in Ancient Iceland (Chicago: University of Chicago Press, 1937), pp. 35–36. 80  Genovese, Roll, Jordan, Roll, pp. 443–450. See also Newbell N. Puckett, “American Negro Names,” Journal of Negro History 23 (1938): 35–48. On the significance of name-changing upon emancipation see Ira Berlin, Slaves without Masters (New York: Vintage Books, 1976), pp. 51–52. On the struggle to retain African names in colonial South Carolina, the linguistic compromises worked out between masters and slaves (especially in the use of African day-names), and the ultimate disappearance of African names with the Americanization of the slave population see Peter H. Wood, Black Majority: Negroes in Colonial South Carolina (New York: Alfred A. Knopf, 1974), pp. 181–186. For an enslaved African’s account of his reactions with each new master see Olaudah Equiano, The Interesting Narrative of the Life of Olauda Equiano, or Gustavus Vasa, The African, Written by Himself (Norwich, England: Printed and Sold by the Author, 1794), pp. 62, 87.

Authority, Alienation, And Social Death

119

shape a social identity independent of slave ownership.”81 This has become a highly contentious subject, one that has generated more heat than the points at issue merit. My own reading of the literature, including slave narratives and interviews, suggests that while there were many variations both within and between regions, most slave surnames in the United States were those of the owners and changed with a change of owner. Nor can the fact that slaves had no legal claim to surnames be dismissed as irrelevant “legalism” any more than can the fact that they had no legal claim to their own persons or labor. The situation in Latin America was similiar to that uncovered in South Carolina by Peter Wood: the masters chose the names, but during the colonial period often selected African names; later the African names were replaced by Spanish ones. Thus in Colombia: The Spanish usually retained the bozales’ African tribal names, or their place of origin in Africa, as the blacks’ surnames. Second-generation slaves might retain this African surname but usually either had no surnames, took the surnames of their masters, or were designated criollos (born in America).82 A census taken in Colombia in 1759 showed that almost 40 percent of the slaves had only one name; 30 percent had the surname Criollo, and the remainder had African tribal or regional surnames such as Mina, Congo, Mandingo, and Caraba. Blacks “were more likely to assume their owners’ surnames following manumission than while they remained in captivity.”83 Much the same pattern existed in other parts of Latin America. In Mexico, for example: All African slaves ... were given a first name and were identified by that name. The names most commonly used included Juan, Anton, Francisco, Diego, Sebastian, and Hernando for males, and Maria, Isabel, Magdalena, Ana, and Catalina for females. Some slaves had a last name as well (usually that of the master)—slaves who were given only a first name were often identified by the addition of their tribal or their national origin ...

81   The Black Family in Slavery and Freedom (New York: Pantheon, 1976), pp. 230–256. 82  William F. Sharp, Slavery on the Spanish Frontier: The Colombian Chocó, 1680–1810 (Norman: University of Oklahoma Press, 1976), p. 114. 83  Ibid., pp. 114–115.

120

Patterson

Other slaves, such as Juan Viejo (old man) and Juan Tuerto (one-eyed), had a nickname appended to the first name.84 The pattern of naming in the Caribbean also was very similar to that of Spanish America and colonial South Carolina. In Jamaica, African day-names and tribal names were either selected in their pure form or adapted as English names. During the nineteenth century these African names acquired pejorative meanings: Quashee, a day-name that originally meant “Sunday” in Akan, came to signify a stupid, lazy slave; and Cudjo, which was the Akan day name for “Monday,” came to mean a drunkard.85 Even a change to purely English or Creole names did not involve any lessening of degradation: slaves were given either classical names such as Phoebe and Cyrus, or insulting nicknames. On Worthy Park estate, for example, they had such names as Beauty, Carefree, Monkey, Villain, and Strumpet. These names were certainly imposed on the slaves by their masters or overseers, for as Craton notes: “To a significant degree, all these single slave names were distressingly similar to those of the estate’s cattle, so that it is almost possible to confuse one list with another in the Worthy Park ledgers.”86 Toward the end of the eighteenth century an increasing number of slaves acquired a surname and usually at the same time changed their forename. This was permitted after baptism and may well have been one of the major incentives for Christianization in Jamaica. Whatever the reason, by the time of abolition most slaves had two names, usually English, with the surname being that of respected whites on the plantation or in the area.87 Where children acquired surnames, these were rarely given before their tenth year “and very often these names reflect those of the whites on the estates (even when they were not the fathers).”88 Finally there were the French Antilles. While naming practices there were similar in broad outline to those in the British Caribbean, there were a few noteworthy variations.89 Slaves were given a new name on the slave ships 84  Colin A. Palmer, Slaves of the White God; Blacks in Mexico, 1570–1650 (Cambridge Mass.: Harvard University Press, 1976), p. 39. 85  David De Camp, “African Day-Names in Jamaica,” Language 43 (1967): 139–149. See also Patterson, The Sociology of Slavery, pp. 174–181. 86  Michael Craton, Searching for the Invisible Man: Slaves and Plantation Life in Jamaica (Cambridge, Mass.: Harvard University Press, 1978), p. 157. 87  Ibid., p. 158. 88  Higman, Slave Population and Economy in Jamaica, p. 173. 89  My discussion here of the French Antilles relies heavily on Gabriel Debien, Les esclaves aux Antilles françaises, XVIIe–XVIIIe siècles (Basse-Terre, Fort-de-France: Société d’histoire de la Guadeloupe et Société d’histoire de la Martinique, 1974), pp. 71–73.

Authority, Alienation, And Social Death

121

during the passage from Africa, yet among themselves they used their African names. A few days after arriving on the plantation each slave was given a nickname, which became his official name and was the one used by the planters. Apparently slaves continued among themselves to use other names as their Christian names, with the planters’ names becoming their surnames. This tendency was much greater among males than females, the women for the most part using the single name given by their masters. The slaves had a third name, acquired after baptism—usually that of a saint. This name was rarely used by the slaves themselves and almost never by the masters. Its main role was to indicate baptismal status. As for the names themselves, French masters too used names of classical figures and names from literature. The blacks themselves apparently preferred names from the military lexicon such as Alerte, Jolicoeur, Sans-souci, and Fanfaron. The nicknames or second names given by the masters referred either to some physical characteristic of the slave (Longs-Bras, Conquerico, Torticolis, Hautes-Fesses) or to their area of origin (Fantu, Mina, Senegal). In some instances the African day-name was used, as in the British Caribbean and colonial South Carolina, but the masters, being French, insisted on a translation, so that the slaves were called Mercredi, Vendredi, and so on. There was the same tendency in the French Antilles for African first names to be replaced by Creole names with the passing of the eighteenth century. The slaves late in that century had more opportunities to choose their names because of the much higher proportion of absentee owners and the rapid turnover of overseers. When they had a choice, they almost never selected the names of their owners: instead they used the names of ancestors who had belonged to another master, or an area of Africa, or colonial heroes and theatrical and literary figures known to be abolitionists, or—most commonly—of saints. The slave’s name was only one of the badges of slavery. In every slave-holding society we find visible marks of servitude, some pointed, some more subtle. Where the slave was of a different race or color, this fact tended to become associated with slave status—and not only in the Americas. A black skin in almost all the Islamic societies, including parts of the Sudan, was and still is associated with slavery. True, there were white slaves; true, it was possible to be black and free, even of high status—but this did not mean that blackness was not associated with slavery.90 Perceived racial differences between masters 90  See Bernard Lewis, “The African Diaspora and the Civilization of Islam,” in Martin L. Kilson and Robert I. Rotberg, eds., The African Diaspora (Cambridge, Mass.: Harvard University Press, 1976), pp. 37–56.

122

Patterson

and slaves could be found in a significant number of other societies ranging from the Ethiopians, the Bemba, and even the Lozis of Africa, to the Gilyaks and Lolos of eastern Asia. Another way slaves were identified was by the ornaments they were either obliged or forbidden to wear. Usually a special kind of clothing was specified among peoples like the Ashanti and Chinese, and among peoples such as the Ibos as well, certain forms of jewelry were forbidden. Tlingit slave women could not wear the lip plug favored by free women. Obvious racial distinctions made it unnecessary to enforce clothing prohibitions on the slaves of the Americas and other areas of the modern world, although there were such rules in some areas.91 The Greeks did not require their slaves to wear special clothes, but apparently (as in America) the slaves’ style of dress immediately revealed their status.92 Rome is fascinating in this regard. The slave population blended easily into the larger proletariat, and the high rate of manumission meant that ethnicity was useless as a means of identifying slaves. A ready means of identification seemed desirable, however, and a special form of dress for slaves was contemplated. When someone pointed out that the proposal, if carried out, would lead slaves immediately to recognize their numerical strength, the idea was abandoned.93 The presence of tattoos also identified slaves. They were universal in the ancient Near East, although apparently removable.94 Surprisingly few societies 91  During the eighteenth century the South Carolina masters attempted to regulate slave clothing by law, but the attempt was abandoned because of a lack of interest of the masters in its enforcement. For a detailed discussion of the limited symbolic role of costume in the slave culture of South Carolina see Joyner, “Slave Folklife,” pp. 206–219. Joyner found no evidence to support Genovese’s claim that slaves preferred the color red. Over time the rough cotton “osnaburgs” became identified as “nigger cloth.” American slaves, too, soon developed peculiar styles of dressing. See Genovese, Roll, Jordan, Roll, pp. 550–561. In Mauritius slaves were not permitted to wear shoes. The colonists declared that doing so “was tantamount to proclaiming their emancipation” (Burton Benedict, “Slavery and Indenture in Mauritius and Seychelles,” in Watson, Asian and African Systems of Slavery, p. 141). 92  For further discussion see Victor Ehrenberg, The People of Aristophanes (New York: Schocken Books, 1962), p. 184. 93  We do not know when this incident, referred to by Seneca, occurred. See Seneca, On Mercy (Cambridge, Mass.: Harvard University Press, Loeb Classical Library, 1970), 1.24.1. Plautus also refers to the slaves’ different style of dress although, of course, the setting is supposedly Greece. Plautus, Amphitryo, 114, in Plautus, The Rope and other Plays, ed. and trans. E. F. Watling (New York: Penguin Books, 1964), p. 232. 94  There is still some controversy concerning exactly what the abbuttum or Babylonian slave-mark was, though it is certain that it was a mark of degradation. For a discussion

Authority, Alienation, And Social Death

123

in the premodern world branded slaves and when they did, as in China, Hellenistic Egypt (where it was eventually forbidden by law), and Rome, only incorrigible runaways were marked. In late medieval and early modern Europe, however, branding of galley and other public slaves was the norm. In France, from the middle of the sixteenth century, persons condemned to galley slavery were first publicly whipped and then the letters gal were burned into their shoulders. Between 1810 and 1832, when branding was abolished, all public slaves (especially those sent to the Bagnes) were branded with the letters tp (Travaux perpétuels).95 The branding of public slaves was not abolished in Russia until 1863. The kátorshniki were branded in a particularly grisly manner: the letters kat were punctured on their cheeks and forehead, and gunpowder was rubbed into the wounds.96 Throughout the Americas slaves were routinely branded as a form of identification right up to the second half of the eighteenth century. Thereafter, although branding became mainly a form of punishment used on runaways and insubordinate slaves, it did not disappear as a means of identification, even in the United States. As late as 1848 a Kentuckian master identified a runaway female slave by announcing that she was branded “on the breast something like L blotched.”97 And South Carolina not only allowed branding until 1833, but mutilated slave felons by cropping their ears.98 Branding as a customary form of identification only began to decline in the Caribbean during the last decades of the eighteenth century under abolitionist and missionary pressure. The lp mark with which slaves were branded on their shoulders in Worthy Park during the eighteenth century is still used today as a means of identifying the estate’s cattle.99 Latin America showed much the same pattern, except that branding of runaways as a form of identification continued until well into the nineteenth century and may even have increased in Cuba during the expansive years at the middle of the century. Occasionally the branding of slaves backfired. In the Minas Gerais area of Brazil, runaway slaves who formed quilombos, or Maroon communities, were branded f on their shoulders if and when recaptured. of the subject see G. R. Driver and John C. Miles, eds., The Babylonian Laws (Oxford: Clarendon Press, 1960), vol. 1, pp. 306–309, 422–423. 95  Sellin, Slavery and the Penal System, pp. 49–50. 96  Ibid., p. 120. 97  Cited in Kenneth Stampp, The Peculiar Institution (London: Eyre & Spottiswoode, 1964), p. 185. 98  Ibid., p. 205. 99  Craton, Searching for the Invisible Man, p. 198.

124

Patterson

Among the slaves themselves, however, the f brand became “a badge of honor rather than of infamy,” and recaptured slaves proudly displayed it to their more cautious but admiring fellow sufferers. When the masters learned of this they replaced branding with a more gruesome form of punishment: the Achilles tendon on one foot was severed.100 Sometimes it was the absence of marks that identified slaves, as among the Yorubas who forbade slaves to scar themselves with Yoruba tribal marks; at other times it was the presence of such tribal marks that immediately betrayed the slaves, as among the Ashanti, who did not tattoo themselves like the many neighboring peoples they captured and enslaved. And one could always tell a Mende slave woman by the fact that her hands were not black with dye, since only nonslave women had the leisure and prerogative to dye cloth.101 There is one form of identification that deserves special attention, since it is found in the great majority of slaveholding societies: this is the shorn or partly shorn head. In Africa we find the shorn head associated with slaves among peoples as varied as the Ila and the Somali. In China, in highland Burma, among the primitive Germanic peoples, the nineteenth-century Russians, the Indians of the northwest coast, and several of the South American and Caribbean tribes, the heads of slaves were shorn (in the ancient Near East so was the pubic hair of female slaves). In India and pharaonic Egypt slaves wore their hair shorn except for a pigtail dangling from the crown. The Mossi of West Africa were unusual in that the head of the slave was periodically shaved by the master considering selling him, and the practice strongly influenced his final decision on the matter. According to A. A. Dim Delobsom: “Depending on where the hair starts to grow, whether well back on the head, at the forehead, or near the ears, the interpretation varies as to how the slave is to be regarded: as a dangerous being; as a lucky or unlucky influence on the family owning him.”102 Numerous other examples could be cited. The shaving of the slave’s head was clearly a highly significant symbolic act. Of all the parts of the body, hair has the most mystical associations.103 On the private or individual level, 100  C. R. Boxer, The Golden Age of Brazil 1695–1750 (Berkeley: University of California Press, 1969), p. 172. 101  Kenneth Little, “The Mende Farming Household,” Sociological Review 40 (1948): 38. 102  A. A. Dim Delobsom, L’empire du Mogho-Naba: Coutumes des Mossi de la Haute-Volta (Paris: Domat-Montchrestien, 1943), p. 64. 103  On the psychoanalysis of hair and hairdressing see Charles Berg, The Unconscious Significance of Hair (London: Allen & Unwin, 1951). For a sympathetic critique of this work and a useful statement of the anthropological symbolism of hair see E. R. Leach, “Magical Hair,” in John Middleton, ed., Myth and Cosmos (Garden City, N.Y.: Natural History Press, 1967), pp. 77–108. For a general study of hair and hair symbolism see Wendy Cooper, Hair:

Authority, Alienation, And Social Death

125

there is hardly a culture in which hair is not, for males, a symbol of power, manliness, freedom, and even rebellion; and for women, the crowning expression of feminine beauty. The shorn head is, conversely, symbolic of castration—loss of manliness, power, and “freedom.” Even in modern societies we tend to shave the head of prisoners, although the deep symbolic meaning is usually camouflaged with overt hygienic explanations. On the public or social level, the shorn head in premodern societies usually signified something more: it was a common symbol of transition, especially in the case of mourning the dead. The association between death, slavery, and the shorn head was made explicit for us by the Callinago Caribs of the Lesser Antilles, many of whom were wiped out by the Spaniards soon after their conquest of the islands. Raymond Breton, who visited them in the middle of the seventeenth century, wrote as follows: The women cut their hair upon the death of their husbands, and husbands cut their hair upon the death of their wives. The children cut their hair upon the death of their father or mother. The hair is cut for the period of a year. The slaves have their hair cut all the time and are never allowed to let it grow. They have their hair cut to the neck which means that they are in mourning (emphasis added).104 It is not unreasonable to conclude that the shorn head of the slave was one aspect of a stark symbolic statement: the man who was enslaved was in a permanent condition of liminality and must forever mourn his own social death. How then do we explain the absence of the shorn head in the large-scale slave systems of the Americas? The answer, I feel, is highly revealing of the symbolic role of hair not only in slave relations but in race relations as well. First, there is the obvious fact that the masters were white and the slaves Sex, Society, and Symbolism (London: Aldus Books, 1971). And for a useful review of the anthropological literature on hair “as private asset and public symbol” see Raymond Firth, Symbols: Public and Private (Ithaca, N.Y.: Cornell University Press, 1973), pp. 262–298. As important as the content in hair symbolism is the structural principle that hair shaving always implies a transitional status. I shall have more to say about this in a later chapter. 104   Raymond Breton, “Father Raymond Breton’s Observation of the Island Carib: A Compilation of Ethnographic Notes Taken from Breton’s Carib-French Dictionary Published in 1665,” trans. and ed. Marshall McKusick and Pierre Verin (New Haven: HRAF, 1957?), p. 42 (manuscript). See also Raymond Breton and Armand de la Paix, “Relations de l’Ile de la Guadeloupe,” in Joseph Rennard, ed., Les Caraibes, La Guadeloupe, 1635–1656 (Paris: Librairie Generale et Internationale, 1929), pp. 45–74; Irving Rouse, “The Carib,” in Steward, Handbook of the South American Indians, vol. 4, pp. 552–553.

126

Patterson

black—a somatic difference that obviated the need for the more common badges of slavery. Contrary to the common view, it was not so much color differences as differences in hair type that become critical as a mark of servility in the Americas. Color, despite its initially dramatic impact, is in fact a rather weak basis of ranked differences in interracial societies.105 There are several reasons. For one thing, the range of color differences among whites and among blacks is greater than is normally thought. Dark Europeans, especially Latins, are not far removed from many Africans who come from areas other than the classic West African “jet-black” zone. The differences diminish even more when we take into account the permanent suntan acquired by most whites working in the tropics. Furthermore, the color differences are quickly blurred by miscegenation, which diminishes the significance of color much faster than is usually imagined. Very soon, therefore, in all slave societies of the Americas, there were numbers of slaves who were in fact lighter than many European masters: the probability that the mulatto slave offspring of an African mother and a very blond Cornish or Irish father was lighter than the average dark Welsh overseer was significantly above zero. Within a couple of generations the symbolic role of color as a distinctive badge of slavery had been greatly muted—though, of course, not eliminated. Variations in hair were another matter. Differences between whites and blacks were sharper in this quality than in color and persisted for much longer with miscegenation. Hair type rapidly became the real symbolic badge of slavery, although like many powerful symbols it was disguised, in this case by the 105  There is a vast literature on this subject. For a good overview see the papers collected in Magnus Mörner, ed., Race and Class in Latin America (New York: Columbia University Press, 1970). See also Charles Wagley, “On the Concept of Social Race in the Americas,” Actas del XXXIII Congreso Internacional de Americanistas 1 (1959): 403–417; and Carl N. Degler, Neither Black nor White: Slavery and Race Relations in Brazil and the United States (New York: Macmillan Co., 1971). For an excellent study of the complexities of race, class, and color in a Caribbean slave society see Verena Martinez-Alier, Marriage, Class, and Colour in Nineteenth-Century Cuba (Cambridge: Cambridge University Press, 1974). For an interesting but flawed theory of the role of somatic perception in the development of race relations during and after slavery see Harry Hoetink, The Two Variants of Caribbean Race Relations (London: Oxford University Press, 1967). See also David Lowenthal, West Indian Societies (New York: Oxford University Press, 1972), esp. chap. 7; Florestan Fernandes, “Slaveholding Society in Brazil,” in Vera Rubin and Arthur Tuden, eds., Comparative Perspectives on Slavery in New World Plantation Societies (New York: New York Academy of Sciences, 1977), pp. 311–342; and Leslie B. Rout, Jr., The African Experience in Spanish America (New York: Cambridge University Press, 1976), esp. chaps. 5 and 12.

Authority, Alienation, And Social Death

127

linguistic device of using the term “black,” which nominally threw the emphasis to color. No one who has grown up in a multiracial society, however, is unaware of the fact that hair difference is what carries the real symbolic potency.106 In the Americas, then, blacks’ hair was not shorn because, very much like the Ashanti situation where the slaves came with a readymade badge (their tribal taboos), leaving the hair as it was served as a powerful badge of status. Shaving it would have muted the distinction. Significantly, in those mixed-blood slaves where the hair type was European, we find a reversion to the premodern tendency of resentment of the slave’s long hair on the part of the masters, not to mention excessive pride on the part of the slave. A telling instance of this comes from nineteenth-century Barbados. In 1835 the governor issued an order to the effect that all slaves convicted of crimes “shall have their hair cut off, and their heads washed, for the better promotion of cleanliness.” This was a new practice, coming less than four years before the complete abolition of slavery. The governor, following European practice, no doubt introduced the order for genuinely hygienic purposes. However, it provided masters and, more frequently, mistresses with a golden opportunity to put “uppity” mixed female slaves in their place—as we learn from a September 1836 entry in the journal of John Colthurst, special magistrate of Barbados: Speaking of the practice of shaving the heads of apprentices, a young quadroon woman who had conducted herself very improperly to her mistress, was brought up about a fortnight before my arrival in the island, and convicted by my predecessor of insubordination, and sentenced to labour on the tread mill for fourteen days, and her head (as a matter of course) to be shaved. This was accordingly done, and on the expiration of her punishment, she was sent home to her mistress, in all respects tamed and amenable, until she found she was laughed at by her fellow servants for the loss of her hair which, like all others of her particular complexion, is usually extremely beautiful, and of wavy and glossy black, and in the utmost profusion and great length. To replace her hair, she purchased false curls, and exhibited a beautiful front. Ere long, however, the circumstance of the original shaving of her head, and which she of course laid all 106  For a discussion of the social psychology of hair and hair color, and the priority status of “good hair” over “good color,” see the work of the Jamaican sociologist Fernando Henriques, Family and Colour in Jamaica (London: MacGibbon & Kee, 1968), esp. chaps, 1, 2, 13, and 14. As Henriques observes: “A dark person with good hair and features ranks above a fair person with bad [that is, African-type] hair and features” (p. 55).

128

Patterson

to her mistress’ account, created another quarrel for which she was again brought up before me, in full curl. The charge was proven and another punishment was the consequence—solitary confinement for six days. If this woman’s head had not been shaved in the first instance, it is clear there would not have been any necessity of a second application to the Special Magistrate. Therefore my objection to punishments of degrading nature, for it appeared in evidence upon this trial that whenever she put her hand to her head, after her return home from her first punishment, and found it bald, she flew into a rage, and swore she would be revenged.107 No doubt the female slaves of ancient Mesopotamia must have flown into similar though silent and repressed rages when they felt their shorn pubic hair, as did the male slaves of all the premodern slaveholding systems when they felt their bald or half-shorn heads. In the Americas the master class thought it achieved the same objective by making African hair the badge of servility. With mixed-race mulatto slaves they may well have succeeded; but with those who retained their African features the degree of symbolic success was questionable. As the shrewd magistrate Colthurst commented: “The negro laments over the loss of his lamb’s wool much more than any fashionable young man in England would, having lost the most exquisite crop of hair in the world.”108 Unfortunately it was the mulattoes who were to define the symbolic meaning of hair in postemancipation and modern Caribbean societies. But that is another story.109

Fictive Kinship

I have several times referred to the practice of incorporating the slave as a fictive kinsman of his master in kin-based societies, and even in many of the more complex premodern systems. It is time to clarify exactly what this means. 107  Woodville K. Marshall, ed., The Colthurst Journal: Journal of a Special Magistrate in the Islands of Barbados and St. Vincent, July 1835–September 1838 (Millwood N.Y.: K.T.O. Press, 1977), p. 100. Resentment of the beautiful curls of some mulatto women also existed in South Africa. 108  Ibid. 109  See Rex M. Nettleford, Identity, Race, and Protest in Jamaica (N.Y.: William Morrow and Co., 1972), esp. chaps. 1, 3, and 5. On Trinidad see Bridget Brereton, Race Relations in Colonial Trinidad, 1870–1900 (Cambridge: Cambridge University Press, 1979), and for the modern period Selwyn D. Ryan, Race and Nationalism in Trinidad and Tobago (Toronto: University of Toronto Press, 1972).

Authority, Alienation, And Social Death

129

On the surface the relationship appears to be a straightforward adoption. All over the world we find the master being addressed as “father” and the slave as “son” or “daughter,” and in matrilineal societies we find the term for the social father being used (that is, the term for “mother’s brother,” while slaves are referred to by the master as “sister’s son”). This fictive kin relation extends also to other members of the master’s family. It would be a great mistake, however, to confuse these fictive kin ties with the claims and obligations of real kinship or with those involving genuine adoption. Some anthropologists are rather careless about making this distinction.110 Relations, we are told, are always warm and intimate; it is difficult to detect any difference between the “adopted” slave and other young members of the family. No wonder some interpreters have concluded that slavery does not exist in these traditional societies, or that the traditional patterns of servitude are best called something else. In order to avoid confusion it is best that we distinguish between two kinds of fictive kinship, what I shall call adoptive and, following Meyer Fortes, “quasi-filial.”111 Fictive kin ties that are adoptive involve genuine assimilation by the adopted person of all the claims, privileges, powers, and obligations of the status he or she has been ascribed. Fictive kin ties that are quasi-filial are essentially expressive: they use the language of kinship as a means of expressing an authority relation between master and slave, and a state of loyalty to the kinsmen of the master. In no slaveholding society, not even the most primitive, is there not a careful distinction drawn between the genuinely adopted outsider (who by virtue of this act immediately ceases to be an outsider) and the quasi-filial slave (who is nonetheless encouraged to use fictive kin expressions in addressing the master and other members of his family). Thus among the pre-European Cherokees, for instance, a captive who was not tortured and put to death was either adopted or enslaved and there was no confusion on the matter. Persons adopted were “accorded the same privileges ... as ... those whose membership derived from birth.”112 Of the Tallensi slaves of West Africa, Fortes wrote: “Homeless and kinless, they must be endowed with a new social personality and given a definite place in the

110  See, for example, James B. Christensen, Double Descent among the Fanti (New Haven: Human Relations Area Files, 1954), p. 96; and J. S. Harris, “Some Aspects of Slavery in Southeastern Nigeria,” Journal of Negro History 27 (1942): 96. 111  Meyer Fortes, The Web of Kinship among the Tallensi (London: Oxford University Press, 1949), p. 25. 112  Perdue, Slavery and the Evolution of Cherokee Society, p. 11.

130

Patterson

community. But the bond of actual paternity cannot be fabricated; the fiction is a makeshift and always remains so.”113 At best, the slave was either viewed as an illegitimate quasi-kinsman or as a permanent minor who never grew up. He might be “of the lineage,” but as the Imbangala of Angola illustrate, he was never in it.114 Among the Ashanti, children of slaves remained slaves “forever” in spite of the adoption of the master’s clan name, and while such children were preferred for political purposes (and well treated), their slave origins were never forgotten. They were laughed at in private, and people referred to them as having a “left-handed” clan affiliation. Old family slaves who became too familiar were put in their place, as several Ashanti proverbs indicate. For example: “If you play with your dog, you must expect it to lick your mouth.”115 The Imuhag group of Tuaregs is instructive in this respect. We find here the standard pattern of fictive kin assimilation and the slave’s adoption of the master’s clan name. However, a slave’s status as a fictive daughter did not get in the way of the master’s taking her as his concubine or even his wife. Furthermore, the social distance between free and slave was great, in spite of the fictive kin bond. Masters in general distrusted their slaves, both male and female.116 Female slaves were frequently accused of witchcraft, and we know from the anthropological psychology of witchcraft that such accusations invariably reflect an underlying fear and distrust of the accused.117 Even where there was considerable intermarriage between slave and free, in this way replacing fictive kinship with real, the assimilation of the slave was still not assured. As Polly Hill points out, the assimilation of gandu slaves (those on special slave farms) into the Nigerian Hausa society “was probably quite limited owing to the breakup of most gandu estates by the time the grandsons had reached marriageable age, if not before.”118 One of the problems with many anthropological accounts of slavery in kin-based societies is that the emphasis on the structural aspects of social life 113  Meyer Fortes, The Dynamics of Clanship among the Tallensi (London: Oxford University Press, 1945), p. 52. 114  J. C. Miller, “Imbangala Lineage Slavery,” in Miers and Kopytoff, Slavery in Africa, p. 213. 115  Rattray, Ashanti Law and Constitution, pp. 40–41. 116   André Bourgeot, “Rapports esclavagistes et conditions d’affranchissement chez les Imuhag,” in Meillassoux, L’esclavage en Afrique précoloniale, pp. 85, 90. 117  On the psychology of witchcraft in West Africa see M. J. Field, Search for Security: An Ethnopsychiatric Study of Rural Ghana (London: Faber & Faber, 1960). See also the classic paper of S. F. Nadel on the subject, “Witchcraft in Four African Societies: An Essay in Comparison,” American Anthropologist 54 (1952): 18–29. 118   Rural Hausa: A Village and a Setting (Cambridge: Cambridge University Press, 1972), p. 42.

Authority, Alienation, And Social Death

131

often leads to a neglect of the purely human dimension. This is a serious drawback when it comes to understanding the real meaning of slavery, especially for slaves. Precisely because economic and class differences between masters and slaves were often not marked, the interpersonal and psychological dimensions of powerlessness became all the more important. It was deeply humiliating to be a slave in a kin-based society, and the indignity was no less because unaccompanied by class differentiation. Indeed, it may have hurt a good deal more. The latifundia slave could at least explain his degradation in terms of the economic parasitism and exploitation of his master. The slave in the kinbased society had no such external explanations. His degradation sprang from something presumably innate to his very being. And the degradation heaped upon him came in little ways, sometimes minor, sometimes cutting, but with the cumulative effect of a piranha assault. Occasionally an anthropologist gives us a rare glimpse of this aspect of exploitation in a kin-based society. In his fine study of the Cubeo Indians of the northwestern Amazon, Irving Goldman records the following incident in the life of a servant girl who had been “adopted”: The little girl, about nine, was addressed as “daughter” but held the status in the household of a servant. She took on the heaviest chores and was almost never free to play. Her lowly status was truly stigmatized by her lack of possessions. She was the only child among the Cubeo whom I have ever seen unadorned ... The children in the household enjoyed beating her as a way of teasing her, rather than wickedly. She took their pinchings and cuffings good-naturedly, on the whole, and had learned to pretend not to notice. Once, in the presence of the headman, her “father,” the children were overdoing their teasing. She looked imploringly at the headman. Finally, she caught his eye and he said to her, “It is all right for you to run away.” He saw no need to reprimand his own children.119 The distinction between adoptive and quasi-filial kinship helps us to understand why it is that even in the highly capitalistic slave systems of the Americas it was still possible to find the master-slave relationship expressed in “kinship” terms. Indeed, quasi-filial kinship became embroiled in the ongoing covert struggle for authority and dignity between masters and slaves, and it was often difficult to distinguish between genuine expression of affection, sheer duplicity, and psychological manipulation. 119   The Cubeo: Indians of the Northwestern Amazon (Urbana: University of Illinois Press, 1963), Illinois Studies in Anthropology, no. 2, p. 130.

132

Patterson

Two examples will illustrate. In the U.S. South the masters encouraged children to see them as the “Big Pappy,” always benevolent, kind, and indulgent. Strict discipline was left to, and expected from, the slave child’s parents. The slave children grew up making unfavorable comparisons between real parental authority and the quasi-filial paternalism of the master. The resulting erosion of the paternal bond was, of course, reinforced by the mortifying subjection of the slave parents to punishment before their children. As Genovese concludes: “If the tendency to worship the master and scorn the parents did not take a greater toll than it apparently did, the credit belongs to the slave parents, whose love for their children went a long way toward offsetting the ravages inherent in this scenario.”120 We find quite a different scenario in Jamaica. In the absence of a cohesive master-class culture, relations between masters and slaves either lacked authority or were on the verge of losing it. Slaves, even here, employed quasi-filial kin terms, but often in sardonic ways, with their aggressive intent only lightly veiled. When the popular gothic novelist Monk Lewis, an absentee slave owner from England, toured his plantations in the early nineteenth century, he was overwhelmed by the reception from the slaves: In particular, the women called me by every name they could think of. “My son! my love! my husband! my father! You no me massa, you my tat” [father], said one old woman.121 Lewis might have been temporarily overwhelmed, but he was hardly deceived, as he later noted. Nor were any but the most naive of the masters who were so addressed. The use of quasi-filial kin terms not as an expression of loyalty or of subordination, but as a thinly disguised form of sarcasm signaled the failure of authority in this most brutal of slave systems.

Religion and Symbolism

The social death of the slave and his peculiar mode of reincarnation on the margin of his master’s society was reinforced by the religious institutions of kin-based societies. As we have seen, the slave was usually forced to reject his own gods and ancestral spirits and to worship those of his master. Even so, he 120   Roll, Jordan, Roll, p. 514. 121   Journal of a West Indian Proprietor, Kept during a Residence in the Island of Jamaica (London: 1834), p. 240.

Authority, Alienation, And Social Death

133

was frequently excluded from community-wide ritual practices: while it was all right for him to worship his master’s ancestral spirits, he was not allowed to participate in cults that were associated with political power and office. Among more advanced slaveholding systems religion played an even greater role in the ritual process of incorporating the slave to his marginal status. Most ritual activities became the specialized preserve of religious institutions. And in both its structural and ritual aspects religion reflected the more centralized nature of political power.122 In the same way that the state had to develop a specialized set of laws to deal with the secular problems of the slave, so the state cult needed to develop a more specialized set of rules and beliefs to represent the condition of slavery. Religion never played the important role in the development of Greek slavery that it did among the Roman, Islamic, or many Christian peoples. The practice of having the slave worship at the Greek family hearth continued well into the classical period. This hardly met the religious needs of the slaves any more than it would have sufficed for their masters. But slaves again were largely excluded from the extrahousehold religious cults of their masters. What is more, restraints were placed on their attempts to develop their own cults. The religious isolation and confinement of their slaves hardly bothered the Greek masters, for they did not care for any form of incorporation of slaves into the Greek community. Franz Bömer, one of the leading authorities on the religious lives of slaves in antiquity, tells us: The fact is that Greek slaves, and not only those from Delphi but from everywhere ... wander like creatures who are dumb, like human bodies without face or profile, without individuality or self-consciousness, and most important, without a noticeable expression of any religious life, be it collective or personal ... The slaves of the Greeks are diametrically opposed to the religious wealth and vivaciousness of the slaves in Rome, who, in fact, could even convince foreign slaves to forget the gods of their

122  For the classic statement of this view see Max Weber, The Sociology of Religion, trans. Ephraim Fischoff (Boston: Beacon Press, 1964), esp. chaps. 1, 3, and 14. See in particular Weber’s discussion of the means by which Rome, like China and other Far Eastern states, developed “more inclusive associations, especially of the political variety,” while retaining the power and significance of familial religious organizations and ancestral gods (pp. 15–16). See also Robert N. Bellah, Beyond Belief (New York: Harper & Row, 1970), chap. 2.

134

Patterson

native lands and accept Roman ways. Roman religion was stronger in the world of the little man.123 The contrast may be a little overdrawn, but the basic point is certainly correct. Roman slaves had more freedom in every part of their lives than Greek slaves. The Greek polis was an ethnically exclusive unit, whereas Rome was, from relatively early on, an ethnically and politically open systern. It was not just slaves who were excluded from the Greek community, but all foreigners. There were three respects, however, in which Greek religion aided in the adjustment of the slave to his social death. Along with women, slaves were allowed to participate in the state cult of Eleusis. The second important representation of slavery in Greek religion was the saturnalia-type festivals associated with a variety of cults. During these festivities (the oldest being the Cronia ritual) there was a reversal of roles in which slaves ate, drank, and played with their masters.124 The late British anthropologist Max Gluckman has suggested that such rites of reversal both vented feelings of tension in conflictridden relationships and reaffirmed the Tightness of the established order: “The acceptance of the established order as right and good, and even sacred, seems to allow unbridled license, even rituals of rebellion, for the order itself keeps this rebellion within bounds. Hence to act the conflicts, whether directly or by inversion or in other symbolical forms, emphasizes the social cohesion within which the conflicts exist.”125 It may be speculated that these rituals of reversal involved not just a means of releasing the tension inherent in the master-slave relationship, and thereby maintaining order, but emphasized the social death of the slave and his total alienation from Greek life. By playing the master, the slave came to realize, however fleetingly, what it was really like to be not just a free man, but more, a truly free man—that is to say, a Greek. When the playing was over and the roles were reversed to normal, the slave would know then with the sinking feeling of the morning after that socially and politically he was dead. The master, in his turn, learned from the role reversal not compassion for his slave, but the bliss it was to be free and Greek. The Cronia, then, was really a death and resurrection ritual: for the master, it was an affirmation of the life principle and freedom; for the slave, it was a confirmation of his living death, powerlessness, and degradation.

123   Untersuchungen über die Religion der Sklaven in Griechenland und Rom (Wiesbaden: Akademie Mainz, 1960), vol. 2, p. 144. 124  Ibid., vol. 3, pp. 173–195. 125   Custom and Conflict in Africa (Oxford: Basil Blackwell, 1955), p. 125; also chap. 5.

Authority, Alienation, And Social Death

135

The third, perhaps most important, way in which Greek religion related to the condition of slavery was by sacred manumission. The problem of manumission will be discussed at length in a later chapter; I am concerned here only with the role of religion in its legitimization. Sacred manumission was the technique of selling the slave to a god who, by not exercising his proprietary powers, allowed the slave to behave like a free man. The interesting thing about this practice is how secular it actually was. Religion was brought in as a means of legitimizing the manumission transaction only where formal legal mechanisms were absent. Where (as in Athens) legal mechanisms existed, we find no trace of sacred manumission. Bömer demolishes the traditional view that Apollo was a defender of slaves and the great symbol of Greek humanity. The idea of finding freedom in servitude to a god remained alien to Greek thought. The slave who was sold to Apollo was not given his freedom by the god; he merely acquired a de facto freedom by virtue of the fact that the god did not exercise his proprietary powers. This was a neat way of solving the problem created by the naturalistic theory of slavery. If the slave was by nature fit for nothing else, how could he become free? If he was socially dead, how could he be made socially alive? It was not possible. Thus selling the slave to a god salvaged the idea of his slaveness and the permanence of his servile status. Apollo was no defender of slaves, no oasis of universal humanity in the desert of Greek chauvinistic tyranny; on the contrary, he was the ideological salvation of the most inhuman product of the Greek mind—the Aristotelian notion of innate slavishness. Bömer’s brilliant exposure of this false pretender to the sacred throne of humanism deserves to be quoted at length: The light that surrounded Apollo was cold and hard, and this coldness and hardness characterized his essence. He was no “divine friend of man” who could console the unlucky, the wounded and the homeless. These people found help later from Asclepius and Sarapis, and often consciously turned away from Apollo. This ruthless aspect, not the humane one, of the Delphic god revealed itself simultaneously in the enslavement of small groups ... and in the Delphic form of sacral manumission.126 A fascinating aspect of Apollo is the fact that this god, who became the very embodiment of the “Hellenic Spirit,” was of non-Greek origin. This has intrigued and puzzled students of Greek religion, especially the fact that the god was in all likelihood of barbaric, Asiatic origins. The main support for the Asiatic origin of Apollo, W. K. C. Guthrie tells us, is “the fact that at most of his great 126  Bömer, Untersuchungen über die Religion der Sklaven, vol. 3, p. 44.

136

Patterson

cult-centers in the mainland of Greece he appears as an intruder.”127 That the most Greek of Greek gods should be of barbaric origin offers ample room for speculation; equally tantalizing is the thought that there may be some connection between the intruder status of this god in the realm of the supernatural and the significance of his social role in the life of the vast number of intrusive slaves who were so essential to the socioeconomic fabric of Greek civilization. Rome was different, and the slaves’ religious life a great deal better. Not that Roman masters were any less cruel; they may have been even more brutal. Rather, Rome had a culture that was far more inclusive, with institutions that were incomparably more flexible, and in no area more so than religion. In primitive Rome and even as late as midrepublic times, slaves participated in the religion of the household, especially in the Lares cult. Originally the head of the cult was the paterfamilias. But as the latifundia replaced the household farm, the master withdrew from this role. By Cato’s day the slave villicus or overseer directed the cult. With urbanization and the further growth of the latifundia, toward the end of the republic the Lares cult became increasingly attractive to slaves and freedmen.128 The saturnalia and matronalia (festivals in honor of Mars and Hera originally celebrated by married women) were also important ritual supports for the slaveholding system from early times, the former quite possibly influenced by Greek traditions.129 As the gesellschaft principle of social organization replaced the gemeinschaft principle in Roman life, ritual specialization increased further. The slave-oriented cults, however, could only initiate the new slaves into the slave sector. There remained the urgent need to incorporate the slave and still more, his descendants, into the wider community. Several kinds of religious organizations were adapted to meet both the specific ritual needs of the slaves and the wider superstructural problem of somehow representing the slave system in supernatural terms. There were, first, the interclass cults. In Jupiter, Juno, and especially Silvanus, we find originally Roman deities who were associated by the slaves with eastern counterparts with which they were more familiar. Many of the cults were of foreign origin—a good number of them brought to Rome by the slaves themselves. Most notable was Mithras, famous for the rapidity with which it

127  W. K. C. Guthrie, The Greeks and Their Gods (Boston: Beacon Press, 1950), p. 84; also chap. 7. 128  Bömer, Untersuchungen über die Religion der Sklaven, vol. 1, pp. 32–86. 129  For a good discussion of the saturnalia see E. O. James, Seasonal Feasts and Festivals (London: Thames & Hudson, 1961), pp. 175–177.

Authority, Alienation, And Social Death

137

attained popularity and the equality of master and slave in the performance of ritual practices.130 In the institution of the collegia, which constituted the organizational aspect of worship, the slave found not only a church but “a social club, a craft guild, and a funeral society”;131 and in holding one of the many offices, he or she experienced some vicarious sense of importance. The names of some of these colleges are very revealing. In the light of what we have said about slavery as a state of social death, it is not unreasonable to suppose that when the members of one college called themselves “comrades in death,” they were referring not solely to their coming physical death.132 Finally, there was the role religion played in relating the slave and slavery to the wider sociopolitical order. Here it was the state cults that were critical. According to Börner, during the republican era Jupiter Libertas had a special appeal to slaves because of the association of the god with freedom, but the evidence is slender and controversial.133 Of much greater interest was the phenomenon of emperor worship and the extraordinary role of the slaves and ex-slaves in the imperial cults. The earliest of these, the Augustan Lares, was in fact a revival of the dying Lares cult to which the emperor added his own imprint. Keith Hopkins argues that this cult had been started by exslaves, Augustus simply institutionalizing the informal local celebrations into a state cult devoted partly to his worship. “The cult provided rich ex-slaves, as organizers of the cult, with a prestigious and public outlet for social display. And it allowed emperor worship to flourish at street level.”134 It was not long, however, before emperor worship was accepted at all levels of society. It was a major legitimizing force among slaves for the simple reason that the emperor’s cult introduced into Roman law the alien principle of asylum for slaves. The granting of the right of appeal to Caesar’s statue was one of the few ways in which the state intervened between master and slave. The state was, of course, sensitive to this intrusion on the authority of the master, and in practice very few slaves attempted such an appeal. But in enhancing the authority of the emperor in the eyes of all, including even the meanest of slaves, the legitimacy of the system as a whole was reinforced. What the master lost in individual 130  Bömer, Untersuchungen über die Religion der Sklaven, vol. 1, pp. 87–98. 131  R. H. Barrow, Slavery in the Roman Empire (London: Methuen & Co., 1928), p. 164. 132  Ibid., p. 168. 133  Bömer, Untersuchungen über die Religion der Sklaven, vol. 1, pp. 110–171. 134   Conquerors and Slaves (Cambridge: Cambridge University Press, 1978), pp. 212–213. See also Robert E. A. Palmer, Roman Religion and Roman Empire: Five Essays (Philadelphia: University of Pennsylvania Press, 1974), pp. 114–120.

138

Patterson

authority, the slave system as a whole gained, embodied as it was in the divine protective power of the deified emperor.135 Still, as Moses Finley has pointed out: “In so one-sided a relationship, in a world in which there was little hope of material success for the majority of the free population (let alone the slaves), and in which the earthly power was now pretty close to despotism, fear rather than love was often the dominating emotion behind worship, at best fear and love together. Religion became increasingly centered on salvation in the next world, whereas it had once been chiefly concerned with life in this one.”136 Among the religions of salvation, Christianity was to emerge slowly, then dramatically, over the next three hundred years as the religion par excellence, one that could forge a moral order which appealed to and united emperor and subject, master and slave.137 A discussion of the means by which it achieved this is beyond the scope of the present work. It is generally accepted that Christianity found many of its earliest converts among the slave populations of the Roman Empire, although the fact is surprisingly difficult to authenticate.138 What is certain, however, is that the slave experience was a major source of the metaphors that informed the symbolic structure of Christianity.139 The most cursory examination of “the three terms which are the keywords” of the Apostle Paul’s theology (according to J. G. Davies) immediately reveals the extraordinary role of the slave experience as a metaphoric source. These key words are redemption, justification, and reconciliation.140 Redemption quite literally means release from enslavement. Through Christ the believer is emancipated from sin. Justification means that the believer has been judged 135  Hopkins, Conquerors and Slaves, chap. 5. 136   Aspects of Antiquity (New York: Viking Press, 1969), p. 207. 137  For the classic analysis of the religions of salvation see Weber, The Sociology of Religion, chaps. 9–12. 138  Early Christianity, as Weber pointed out, was essentially a religion of urban artisans “both slave and free” (ibid., p. 95). See also Ernst Troeltsch, The Social Teachings of the Christian Churches, vol. 1 (London: Macmillan & Co., 1931); A. D. Nock, Early Gentile Christianity and Its Hellenistic Background (New York: Harper Torchbooks, 1957); Philip Carrington, The Early Christian Church, vol. 1 (Cambridge: Cambridge University Press, 1957). For a good recent treatment of Constantine’s conversion and its consequences see J. H. W. G. Liebeschuetz, Continuity and Change in Roman Religion (Oxford: Clarendon Press, 1979), pp. 277–308. 139  See Ambrogio Donini, “The Myth of Salvation in Ancient Slave Society,” Science and Society 15 (1951): 57–60. For a review of slavery in Christian thought see Davis, The Problem of Slavery, esp. chaps. 3 and 4. 140  J. G. Davies, “Christianity: The Early Church,” in R. C. Zaehner, ed., The Concise Encyclopedia of Living Faiths (Boston: Beacon Press, 1959), p. 56.

Authority, Alienation, And Social Death

139

and found not guilty, in much the same manner as the slave who has received the most perfect of manumissions, the restoration of his natality with the legal fiction that he had been wrongfully enslaved. “Reconciliation or Atonement means the bringing together of those who have been separated,” in much the same way that the manumitted slave is reborn as a member of a community. Paul in fact went so far as to use the idea of adoption to describe the relationship between redeemed man and God. “Redeemed, justified, reconciled, man is elevated from the status of slave to that of son, and becomes ‘an heir through God’ of the promised salvation.”141 What Ambrosio Donini calls “the myth of salvation” became the unifying master concept of organized Christianity, and it is most powerfully evoked in the dominant symbol of the religion, that of the death and resurrection of Christ.142 Man fell into spiritual slavery because of his original sin. Slavery, which on the level of secular symbolism was social death, became on the level of sacred symbolism spiritual death. When, however, we question what Christ’s crucifixion meant, we find two fundamentally different symbolic interpretations. One explanation, which has profoundly conservative spiritual and social implications, held that Christ saved his followers by paying with his own life for the sin that led to their spiritual enslavement. The sinner, strictly speaking, was not emancipated, but died anew in Christ, who became his new master. Spiritual freedom was divine enslavement. Here was a confluence of two old ideas: the Near Eastern and Delphic notion of freedom through sale to a god, and the Judaic idea of the suffering servant and sacrificial lamb. It was not a very tidy symbolic statement, and it accounts in part for the occasional impenetrability of Paul’s theology. He had this interpretation in mind, for example, when he made remarks such as the following: “The death that he died, he died unto sin once: but the life that he liveth, he liveth unto God.”143 There was a far more satisfactory and at the same time more liberating symbolic interpretation of the crucifixion. The slave, it will be recalled, was someone who by choosing physical life had given up his freedom. Although he could, of course, have kept his freedom and died, man lacked the courage to make such a choice. Jesus, “his savior,” by his death made this choice for him. It is this feature that was completely new in the religious behavior and death 141  Ibid., pp. 55–58. 142  For one of the best-known statements of the view that Christianity is primarily a religion built on the response to Jesus’ crucifixion see John Knox, The Death of Christ (London: Collins, 1967). See also Christopher F. Evans, Resurrection and the New Testament (Naperville, Ill.: Alec R. Allenson, 1970). 143  Rom. 6:10.

140

Patterson

of Jesus. What it meant in symbolic terms was that Jesus did not redeem mankind by making mankind his slave in the manner of the old pagan religions. Rather, he annulled the condition of slavery in which man existed by returning to the original point of enslavement and, on behalf of the sinner about to fall, gave his own life so that the sinner might live and be free. It is remarkable that Paul held also to this radically different interpretation of the crucifixion. The contradiction was directly paralleled by the well-known ethical contradiction of his theology. As Maurice Goguel has pointed out, Paul had two irreconcilable religious ethics.144 One was the pre-Christian and essentially Judaic ethic of law and judgment, in which obedience to divine law, and judgment according to one’s social and religious actions, were of the essence. The other was the ethic of the justified man. In this ethic Christ’s death redeemed mankind of the burden of sin; the believer, through faith, was immediately emancipated. The first ethic corresponded to the conservative use of the slave metaphor; the second to the more liberal conception of slavery and emancipation. Paul tried to hold both these positions at the same time and thereby placed the believer, as Goguel points out, in the impossible position of one “who must struggle to realize in fact what he is in principle.”145 And he asks: “How can we now speak of a judgment for those who are in Christ, and therefore cannot be subject to condemnation?”146 The answer was to abandon the liberal view of emancipation and to canonize the essentially pre-Christian interpretation of salvation as reenslavement to a god, in the triumph of the conception of the believer as the slave of God and of Christianity as a theological transmutation of the order of slavery. Whatever other factors explain Christianity’s conquest of the Roman world, there seems little doubt that the extraordinary way in which its dominant symbolic statements and meanings are informed by the experience of slavery was a major contributing factor. For the same reason too, Christianity was to provide institutional support and religious authority for the advanced slave systems of medieval Europe and of the modern Americas. Christianity was not alone among the major world religions in legitimizing slavery. Earlier we noted the contradiction in Islam between the rationalization of slavery as a means of converting the unbeliever and the continuing enslavement of the converted. We find the same contradiction in Judaism and 144  Undoubtedly the most probing modern analysis of Paul’s ethical dualism is Maurice Goguel’s The Primitive Church, trans. H. C. Snape (London: Allen & Unwin, 1964), pp. 425–455. 145  Ibid., p. 428. 146  Ibid., p. 449.

Authority, Alienation, And Social Death

141

Christianity. The slave, in the city of the Christian God, was declared an insider, an integral part of the brotherhood of man in the service of God; but the slave, in the city of man, remained the archetypical outsider, the eternal enemy within, in a formalized state of marginality. At first sight the contradiction is not obvious. Indeed, the opposite seems to be the case: the exclusion of the slave on the secular level was symbolically compensated for by his inclusion in the sacred community. The contradiction between marginality and integration, which slavery created, was apparently resolved by relegating each to a separate domain of cultural existence. But this theological solution on the part of a monotheistic slaveholder class works only where there is hegemonic imposition of a rigid dualism in the socioreligious ideology. This was exactly what happened in medieval Christendom under the conservative spell of Saint Augustine.147 But Judaism and Islam were too this-worldly and too strongly monistic for such an interpretation to be taken seriously.148 And the rise of Protestantism dealt a death blow to the neat symbolic compromise of the Middle Ages. Augustinian dualism lingered in the symbolic representation of Latin American slavery: hence the apparent anomaly that has baffled so many Anglo-American historians, that of a Catholic church stoutly declaring slavery a sin, yet condoning the institution to the point where it was itself among the largest of slaveholders.149 The Anglican masters of the Caribbean avoided the problem altogether by abandoning religion or making a mockery of it, both for themselves and for their slaves—clergymen in nineteenth-century Jamaica being “the most finished debauchers in the land.”150 As Richard S. Dunn has pointed out, the 147  On the chronic dualism of Saint Augustine see Karl Jaspers’ brilliant critique in his Plato and Augustine (New York: Harvest Books, 1957), esp. pp. 109–119. Still valuable is J. N. Figgis, The Political Aspects of St. Augustine’s City of God (London: Longmans, 1921). See my own interpretation in Orlando Patterson, Ethnic Chauvinism: The Reactionary Impulse (New York: Stein & Day, 1977), pp. 231–241. 148  See Weber, The Sociology of Religion, esp. chaps. 9 and 10. 149  We need not become involved here with the spent debate on the role of the Catholic church in Latin American slavery. See Davis, The Problem of Slavery, pp. 223–261; also Rout, The African Experience in Spanish America, chap. 5; Boxer, The Golden Age of Brazil, chaps. 5–7; and Palmer, Slaves of the White God, chap. 4, esp. the cases cited on pp. 113–114. 150  See William Lou Mathieson, British Slavery and Its Abolition (London: Longmans, 1926), pp. 109–114; Elsa V. Goveía, Slave Society in the British Leeward Islands at the End of the Eighteenth Century (New Haven: Yale University Press, 1965), pp. 263–310; and Patterson, The Sociology of Slavery, pp. 33–51. Although formally Catholic, the situation in the French Caribbean was more like that of the British islands. See Debien, Les esclaves aux Antilles françaises, pp. 249–295.

142

Patterson

refusal of the English planters in the West Indies to convert their slaves to Christianity, in contrast with contemporary Latin masters, “can largely be explained by Protestant versus Catholic conversion techniques.”151 Protestantism by its very nature demanded the liberating conception of the crucifixion, with its emphasis on personal choice and freedom. Realizing this, the West Indian masters did everything possible to keep their slaves in ignorance of their creed—giving in only when, a few decades prior to abolition, they found their policy to be too easy a target in the propaganda war of the abolitionists. How then do we account for the Protestant slave South where, during the late eighteenth and the nineteenth centuries, both masters and slaves were highly religious? It is clear that the special version of Protestantism that triumphed in the South and the peculiar socioeconomic features of the system together explain its unusual course of development. Until nearly the end of the eighteenth century the U.S. South did not differ markedly from other Protestant slave systems. Masters were generally hostile to the conversion of their slaves, fearing—like their Caribbean counterparts— that the nature of their creed with its emphasis on instruction in the gospels, personal choice, and spiritual liberation would, if adopted by their slaves, undermine the masters’ authority. As late as 1782 slaves in Georgia were still being whipped savagely for preaching,152 and while Albert Raboteau may have overstated the case in claiming that “the majority of slaves ... remained only minimally touched by Christianity by the second decade of the nineteenth century,” he was not far wrong.153 Two major developments explain the remarkable change that took place during the nineteenth century. One was the great religious awakening that culminated in the religious conversion of the South from classical Protestantism to revivalist fundamentalism.154 The second was the emergence in the South between 1790 and 1830 of a full-fledged slave system, a total commitment to the institution as an essential feature of the region’s socioeconomic order, and the realization that if slavery was to function effectively the system had to be reformed. In Genovese’s words, “whereas previously many slaveholders had 151   Sugar and Slaves: The Rise of the Planter Class in the English West Indies, 1624–1713 (Chapel Hill: University of North Carolina Press, 1972), p. 249. 152  Albert J. Raboteau, Slave Religion (New York: Oxford University Press, 1980), p. 141. 153  Ibid., p. 149. 154  See John B. Boles, The Great Revival, 1787–1805 (Lexington: University Press of Kentucky, 1972); also his Religion in Antebellum Kentucky (Lexington: University Press of Kentucky, 1976). On the ways in which Christian conscience was reconciled to bondage see H. Sheldon Smith, In His Image, But—(Durham, N.C.: Duke University Press, 1972).

Authority, Alienation, And Social Death

143

feared slaves with religion, now they feared slaves without religion even more. They came to see Christianity primarily as a means of social control ... The religious history of the period formed part of the great thrust to reform slavery as a way of life and to make it bearable for slaves.”155 Fundamentalist Protestantism was peculiarly suited to such a reform. Its emphasis on conversion as a sudden spiritual transformation rather than the result of reflection and instruction; its oral rather than literary missionary techniques; its other-worldliness, especially its insistence on salvation as a purely spiritual change, the rewards of which are to be achieved in the hereafter; its emphasis on piety and obedience, and on the sinfulness of the world and the flesh; made it a creed that the masters could confidently regard as a support for, rather than a subversion of, their authority.156 Nevertheless, it would be simplistic to interpret the role of religion in the slave South solely in terms of an opiate for the masses, a device used by the master class as an agent of social control. In the final analysis it was indeed just that, and there is abundant evidence that the master class cynically devised a “theology of slavery” in a crude attempt to rationalize the system. But as recent studies have shown, slaves quickly recognized the crude ideological strategy of their masters. Olli Alho’s detailed analyses of the slave narratives “indicate that the carefully constructed theology of slavery built up by the whites became in many plantations nothing more than a joke” among the slaves.157 The slaves found in fundamentalist Christianity paths to the satisfaction of their own needs, creating the strong commitment to Christianity that has persisted to this day. In so doing they created an institutional base that provided release and relief from the agonies of thralldom, and even offered some room for a sense of dignity before God and before each other. Having said all this, I must emphasize that the religion they experienced was the same as that of their masters in all its essential doctrinal and cultic aspects; that while the spirituals they sang may have had a double meaning with secular implications, it is grossly distorting of the historical facts to claim that they were covertly revolutionary in their intent; and, most important of all, it is irresponsible to deny that however well religion may have served the slaves, in the final analysis it did entail a form of accommodation to the system. 155  Genovese, Roll, Jordan, Roll, p. 186. 156  See Edgar T. Thompson, “God and the Southern Plantation System,” in S. S. Hill, ed., Religion and the Solid South (Nashville, Tenn.: Abingdon Press, 1972), pp. 51–91; and James L. Peacock, “The Southern Protestant Ethic Disease,” in J. K. Morland, ed., The Not So Solid South (Athens: University of Georgia Press, 1971). 157   The Religion of the Slaves (Helsinki: Finnish Academy of Science and Letters, 1976), p. 139.

144

Patterson

In all of this I am in complete agreement with Genovese’s penetrating interpretation of the role of religion in the slave South.158 Where I differ from him, and from others such as Lawrence W. Levine159 and Albert J. Raboteau who with equal skill and persuasion have emphasized the creative and positive side of religion for the slave, is in my interpretation of the specific means by which fundamentalist Christianity became at one and the same time a spiritual and social salvation for the slaves and an institutional support for the order of slavery. To appreciate where we differ it is necessary to return to the nature of Christianity and to specify the peculiar doctrinal features of fundamentalism. Pauline Christianity, as we saw, was theologically dualistic, containing an ethic of judgment and an ethic of the justified person that were in constant tension with each other. These two ethics in turn were symbolically expressed in two contrasting interpretations of Jesus’ crucifixion. Roman Catholicism resolved the tension by eliminating what I call the liberating pole of Pauline dualism, emphasizing the ethic of judgment and obedience; classic Protestantism resolved it by eliminating the conservative pole and by strongly reviving the ethic of the justified person. What then is distinctive in fundamentalism? My answer is that it restored both poles and returned fully to Pauline dualism with all its contained tension and its contextual shifting from one ethical and symbolic pole to the other. If we do not understand this distinctive doctrinal feature of fundamentalism, we cannot fully appreciate how the religion could have spiritually sustained both slaves and masters as well as the system as a whole. We will also fail to comprehend the symbolic life of the slaves themselves. If we next seek the major doctrinal and symbolic components of slave religion, we find that the fundamentalism of the slaves was, like that of all southerners, essentially Pauline in its overwhelming preoccupation with Christ and the crucifixion and in its ethical and symbolic dualism, its paradoxical tension between the ethic of judgment and the ethic of the redeemed sinner. Further, it is precisely this dualism that explains the apparent paradox that the religion of the slaves, doctrinally one with that of their masters, nonetheless allowed for the spiritual support of both groups and of the system as a whole.

158  Genovese, Roll, Jordan, Roll, bk. 2, pt. 1. 159  Lawrence W. Levine, Black Culture and Black Consciousness (New York: Oxford University Press, 1978), chaps. 1 and 3.

Authority, Alienation, And Social Death

145

Jesus and his crucifixion dominate the theology of the slaves and not, as recent scholars have claimed, the Israelites and Exodus story.160 Not only is the theme of the crucified Christ explicitly central and dominant, but even when figures from the Old Testament are referred to (including Moses), closer examination reveals that the allusion is really to Jesus. Although Alho does not make the connection to Pauline theology, it is striking that his most important finding concerns the dualistic conception of Jesus in the religion of the slaves—that of Jesus as Messiah King and Jesus as comforting savior. He concludes his interpretation with a reference to an insightful contemporary observer: “The difference between the two main identities of Jesus reminds one of what T. W. Higginson wrote in his camp diary about the religious behavior of his black soldiers; softness, patience, and meekness on the one hand, hardness, energy, and daring on the other, seems to be reflected in the dualistic way in which the spirituals picture the figure and roles of Jesus.”161 We can now explain how fundamentalism, a single religion, performed the contradictory roles it did in the slave South. Both masters and slaves adhered to Pauline ethical dualism, with its sustained “eschatological dissonance.”162 And in exactly the same way that Paul and the early Christians shifted from one pole of their doctrinal dualism to another as occasion and context demanded, so did the masters and their slaves. Thus the masters, among themselves, could find both spiritual and personal dignity and salvation in the ethic of the justified and redeemed sinner. The crucified Jesus as redeemer and liberator from enslavement to sin supported a proud, free group of people with a highly developed sense of their own dignity and worth. Similarly, the slaves in the silence of their souls and among themselves with their own preachers, could find salvation and dignity in this same interpretation of the crucified Lord. When the theologian Olin P. Moyd insists that “redemption is the root and core motif of black theology” and that it means essentially liberation from sin and confederation within the fellowship of black worshippers, it is, I suspect, to this end of the Pauline dualism that he is referring.163 As with the masters, the slave dualism had another pole. This is the ethic of law, judgment, and obedience, the ethic that found symbolic expression in 160  See Levine, Black Culture and Black Consciousness, p. 33; and Raboteau, Slave Religion, p. 250. Alho’s careful study of the spirituals and slave narratives does not support the view that the Exodus myth or identification with the children of Israel were the dominant themes in the religious beliefs of slaves. See Alho, The Religion of the Slaves, pp. 75–76. 161  Alho, The Religion of the Slaves, p. 79. 162  Goguel, The Primitive Church, pp. 454–455. 163   Redemption in Black Theology (Valley Forge, Pa.: Judson Press, 1979).

146

Patterson

the other Jesus, the more Judaic Messiah King who judges, who demands obedience, and who punishes the wicked and rewards the righteous. This is the Jesus who saves not by annulling slavery but by divine enslavement. To live with this Jesus demands, as Goguel tells us, watchfulness, obedience, and stoic acceptance. Both masters and slaves held also to this conception of Jesus and, like Paul and the early Christians, shifted to this symbolic code in dealing with, and coming to terms with, all authority relations—not only the one between master and slave but, among the masters between male and female, upper class and working class, parent and child, and among the slaves between parent and child. In this way fundamentalism, by reverting to Pauline dualism, provided the slave South with the perfect creed, one much more subtle in its support for the system than most of the masters thought. The crude theology of slavery that the masters tried unsuccessfully to preach in the plantation mission was really quite unnecessary. Nor was it necessary for master and slave to have two separate religions. Christianity, after Paul, had already constructed an extraordinarily shrewd creed with a built-in flexibility that made it possible for emperor and slave to worship the same god without threatening the system, but also without denying all dignity to the oppressed. In the U.S. South there developed the last and most perfectly articulated slave culture since the fall of the Roman Empire. The religion that had begun in and was fashioned by the Roman slave order was to play the identical role eighteen hundred years later in the slave system that was to be Rome’s closest cultural counterpart in the modern world. History did not repeat itself; it merely lingered.

The Anthropology of Slavery: The Womb of Iron and Gold Claude Meillassoux Introduction Current research seems to be concerned less with slavery as a social system than with the definition of the slave. While the convergence of semantics and law helps to identify the phenomenon, it is unable to characterize slavery as an institution. In African societies, as in ancient societies (Vidal-Naquet, 1965–7) the words translated as ‘slave’ can also be applied to wider categories, sometimes to all those who are or have been in any sort of temporal or religious subjugation, in relation to an elder, a sovereign, a protector or a spiritual guide. These words may mean, variously, subservient, submissive, dependent, subject to, and may even refer to a disciple. Conversely, most slave societies have a wide vocabulary covering different conditions of subjugation which no longer have any equivalent in our language and are uniformly translated by the word ‘slave’. In legal terms, the slave is described as an object,1 a possession, therefore alienable and subject to his or her master. But given the nature of his or her exploitation, the assimilation of a human being to an object, or even to an animal, is an untenable and contradictory fiction. If the slave were in practice treated in this way, slavery would present no advantages over the use of material tools or over the breeding of cattle; and in fact, slaves are not used as the objects or animals to which the ideological fiction tends to reduce them. All their tasks, even porterage, call for the use of reason, if only to a limited extent, and their utility and productivity grow in proportion to the use of their intelligence. Efficient slave management implies a greater or lesser recognition of the slave’s capacities as Homo sapiens, and thus a constant shift towards notions of obedience and duty which renders S ource: Meillassoux, Claude, “Introduction,” in Claude Meillassoux, (Alide Dasnois transl.), The Anthropology of Slavery: The Womb of Iron and Gold, Chicago: University of Chicago Press, 1991, 9–22. © Claude Meillassoux, 1991, The Anthropology of Slavery: The Womb of Iron and Gold, and The Athlone Press, an imprint of Bloomsbury Publishing Plc. 1  In primitive Roman law, the slave is explicity an object and not a person (Monier, 1947: 211).

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_007

148

Meillassoux

the slave indistinguishable, in strictly legal terms, from other categories of dependants. Pubescent girls, cadets, wives and pawned people are, like the slave, subject to the absolute power of the head of the family. They can be beaten, sold and even killed. The obligation to work weighs on all those, free-born or subservient, who are dependent on a master, a ‘patriarch’ or a sovereign. On the other hand, certain slaves enjoy privileges—wealth, rank, office— which put them in what appears to be a superior position; and of these, it is said that they are ‘kin’. The soldier-slave, the henchman, the rich slave, who benefit indirectly from the labour of other slaves or even of free men, or who themselves own slaves, are not expected to work. In terms of the slave-object fiction, as in the situations described above, the only institutional relationship relevant to the slave which is recognized by law is his relationship with his master. The legal definition of the slave therefore applies to this—strictly individual—relationship. In these circumstances, the law both ratifies and conceals social relations, sanctioning them in the form most suited to preserve the interests of those for whom the law is conceived and worded. It cannot, therefore, be seen as the objective expression of social reality; nor can it contain an explanation of this reality. In expressing the slave relation as individual, the law fixes the limits within which the authority of the master over the slave can be exercised; thus the individual relation masks and neutralizes the class relation. The individual relation merely reflects a personalized and individualized conception of authority, based on patriarchal ideology. In fact, in strictly legal—that is, individual—terms, the definition of the slave necessarily depends to some extent on that of the free man or woman, because of this implicit ideological reference. This explains the apparently infinite variety of the conditions of individual slaves—a variety which cannot be explained by the legal principle, which is itself indeterminate. Even the most pertinent and most widely applicable criterion—that of the legal alienability of the slave, irrespective of his or her condition—lacks precision. Certain individuals who are not slaves can be alienated, and not all slaves are, in fact, alienable. The weakness of the legalistic approach is that it considers alienability as a characteristic specific to slaves. Yet alienability has meaning only in the context of institutions which make alienation possible: wars of capture or ‘slave markets’—that is, the set of mechanisms and operations through which a set of individuals can be deprived of social personality, transformed into livestock, sold as merchandise and exploited or employed in such a way that the cost of capture or purchase can be covered. But alienation merely represents the

The Anthropology of Slavery

149

transcendental state of the slave. It takes place only when the slave cannot perform productive or functional tasks, for whatever reason, in the society in which he finds himself. Alienation is merely the effect and confirmation of a process of depersonalization which has already been inflicted on the slave through capture. Final alienation takes place on the ‘sacrificial’ altar as well as on the market—that is, in religious rites as well as in commercial transactions. The state of the slave is expressed in his relation to these institutional social frameworks, not in his individual relation to his master. All the African societies examined in this study are linked, directly or indirectly, to the market. Some supply it, others are supplied from it. In the final analysis, the real or potential fate of the slaves—their state [see Glossary], in other words—is necessarily defined with respect to the market, even if all slaves are not necessarily directly subject to the market at all times. It is through the market that the common state of slaves, as a social class, is defined, and it is with respect to the market that the different, changing and individual condition of each slave is defined according to the mode of insertion of the slave in each slave society. The individual relation of the slave to his or her master cannot be explained outside this context. The distinction between the state and the condition of the slave, inherent in this process, is one of the keys to understanding the problem. Because of the importance of this distinction, this study is divided into three parts: the first covers the whole of the economic space of slavery which structures the state of the slave; the other two are devoted to the political and economic forms which slavery takes in the two main types of African societies in which it is present, military aristocracies and merchant societies. In an important collaborative work on African slavery, Miers and Kopytoff (1977) take a very different approach. They offer in their introduction a genetic explanation of slavery which seems to me to take legalism, functionalism and economism to extremes. Miers and Kopytoff consider that what they call ‘minors’ (children, young people, women) are in a dependent position in the family and that, on the other hand, the kinship system allows for the transfer of dependency; and they see slavery as the extension of this double phenomenon to aliens. As a result, the thrust of their argument turns around what they call the ‘slavery-kinship continuum’ and their theory of ‘the transfer of rights-in-persons’. In the first place, they discover that ownership has a peculiar meaning in Africa in that it entails not only ‘rights-in-things’, but a set of ‘rights-in-things-and-in-persons’. Next they discover what they consider to be another feature peculiar to African culture, though they do not specify to which type of society they refer: the fact

150

Meillassoux

that ‘the concept of rights-in-persons and transactions in them ... constitute some of the basic elements on which kinship systems are constructed’. Such transactions, they write, are a formal part of African concepts of kinship.... The transactability of such rights, as discrete and separate items, is also remarkable. Moreover, transfers of such rights are normally made in exchange for goods and money, and the transfer may cover the total rights-in-a-person. As a result, such phenomena as kinship, adoption, the acquisition of wives and children, are all inextricably bound up with exchanges that involve precise equivalence in goods and money.... Consequently, to say that what makes a person a slave is the fact that he is ‘property’, is in fact to say ‘a slave is a person over whom certain [unspecified] rights are exercised’. (1977:11) This theory, according to Miers and Kopytoff, is likely to surprise ‘the Westerner’, who cannot conceive that rights can be divided up and applied to individuals as well as to things. What surprises me personally is that Miers and Kopytoff do not see that their explanation rests on the strict application of Western notions of law and of liberal economics. In our society, property is a set of rights, usus, fructus and abusus, which can, indeed, be attributed separately to different parties or persons. What is more, in domestic society it is not property which is in question but patrimony, which is governed by completely different rules of transfer. It is no longer possible today to defend the ‘vulgar’ materialist thesis that bridewealth is ‘an acquisition of rights’ over children or wives ‘in exchange for a precise equivalent in goods or in money’—a purchase, in other words. Not only do Miers and Kopytoff forget that the matrimonial transaction can take place, and did in fact take place, in numerous African societies without bridewealth, but the notion of equivalence of individuals with goods is not relevant to domestic societies. What is correct in Miers and Kopytoff s theory is that kinship relations are constantly manipulated. What is false is that they are manipulated against currency through purchase. In matrimonial relations, the only equivalent to a pubescent woman is another pubescent woman, with the same potential fertility. The concept of bridewealth does not bear children. When the two terms of a transaction are identical, intermediate goods (if they exist) have no intrinsic value and cannot be exchanged for themselves. Only when these goods enter the commercial circuit outside the community and are produced for exchange can they acquire an intrinsic value and communicate their saleability to the matrimonial circuits, resulting in the transformation of individuals into commodities. The effect of this transformation cannot

The Anthropology of Slavery

151

be attributed to ‘kinship’. There is no ‘continuum’ between the two levels, but rather a qualitative change. Miers and Kopytoff believe that ‘rights-in-persons’ are communicated to the slave system; in fact, the reverse is true: the saleability of slavery contaminates and reifies kinship relations. The theory of rights-in-persons yet again introduces the principles of conservative classical economics into historical situations for which they are even less suited than in our days. Miers and Kopytoff see ‘the roots of the servile institution in the need for wives and children, the wish to enlarge one’s group ... to have clients, servants and retainers’ (p.67). This need grows with ‘the infinite urge to absorb more consumer goods ... much in the same way [as in our] modern consumer societies’. These needs and wishes are satisfied, as Adam Smith would have us believe, thanks to ‘the human propensity for truck and barter’ (p.67). It would be difficult to go further than Miers and Kopytoff in using economic motivations to interpret social phenomena! Why, in these circumstances, would certain populations wish to ‘sell’ their children? If the desire of peoples is to ‘enlarge their group’, why would the majority be prepared to alienate their dependants, and thus to impoverish themselves in absolute terms for the benefit of a small fraction? And where is it possible to find examples of this? It is true that there are parents, driven by hunger, who are forced to sell their children, but this takes place in a context where saleability is already active as a direct or indirect result of trade. Within the original domestic economy, nothing, as has been noted above, can replace a human being as a producer or reproducer except another, identical, human being. If the ‘propensity to barter’ is the motor of exchange, it can allow only for the barter of one man for another man or of one woman for another woman. How can this explain the accumulation of human beings for the benefit of a few? In return for what ‘riches’ would one be tempted to give up the greatest of all riches? The sale of kin is neither ‘traditional’ nor compatible with the organization of kinship. This weakness in Miers and Kopytoff’s approach results in the assimilation of slavery to kinship, when in fact the two institutions are strictly antinomic. If, by the purely ideological extension of kinship, the slave is sometimes assimilated to a sort of perpetual cadet, with the obligations of a dependant in terms of customary notions of morality, he still never acquires the essential prerogatives—those bound up with paternity. His status of non-kin stems from the specificity of slave exploitation and its mode of reproduction. Ignorance of this point blinds one to the contours of slavery, for it is in fact slavery which highlights its opposite, franchise. The free man is defined only in relation to the slave. Society is transformed by the very fact of the introduction of slaves within its midst. It becomes a class society, if it was not one already. New rules are established and the old rules remain only in order to perpetuate the

152

Meillassoux

domination of the free classes and their reproduction as such. The presentation of slavery as an extension of kinship implies approval of the old paternalistic idea which has always been used as moral backing for slavery. It means falling into the trap of an apologist ideology in which the slave-owner tries to pass off those he exploits as his beloved children. Although Miers and Kopytoff lean heavily on economism and naive materialism in their interpretation of servitude and its transformations, they declare: ‘we don’t need to appeal to an economic raison d’être for the existence of slavery’! Perhaps what they mean is that slaves are not necessarily used as producers, which is true. However, the economic scope of slavery is not limited to the productive use of slaves, nor to the profit which it can generate. Whatever their employment, slaves are imported at a cost: that of war or that of export goods. Putting slaves to work involves a choice between commercial and natural reproduction. Those slaves who are not assigned to agricultural labour must be fed, and at low levels of productivity this is a domestic problem which is difficult to resolve. It implies—assuming that the masters do not work for their own slaves!—the exploitation of a different fraction of free men or of other slaves, as well as institutions which can extract their surplus-product and transfer it. When they are used in political activity or in war, slaves take part in the establishment of the political class and act as the means of its economic domination—which becomes even more necessary, since idle slaves must be provided for. In merchant society, the condition of slaves, who can be manipulated at will, changes further in terms of their relation (and that of their products) to the market, through the articulation of their food production and profits, through their entry as a stolen means of production into the general economy, through the nature of the product which enables them to be replaced, and so on. Because they do not consider the practical aspects of slave management, Micrs and Kopy-toff arc blind to its economic implications. It seems clear— though they do not explicitly say so—that Miers and Kopytoff imagine that by dismissing ‘the economic raison d’être of slavery’ they can also dismiss the ‘Marxist’ interpretation, in the commonly held belief that historical materialism can be reduced to the same economistic causality which they themselves use, apparently unwittingly. What, then, does historical materialism—and particularly what do Marx and Engels—contribute to an understanding of slavery? The contributions of the two authors vary in quality in their different works. Engels in particular worked on the conditions leading to the emergence of slavery (1884/1954: 145–63). He sees slavery as the effect (following the dissolution of the gentle order) of three main divisions of labour:

The Anthropology of Slavery

a.

b.

c.

153

The division between agriculture and pastoralism, which gives rise to regular exchange, the emergence of money, and an increase both in production and in the productivity of labour. With an increase in work there is an increasing demand for the producers that war provides.2 The separation of craftwork from agriculture. The value of labour-power increases and men are themselves introduced into exchange as objects of exchange. Slavery becomes an essential component of the social system and war becomes a permanent sector of industry. The separation of town and country which favours the development of a merchant class, the differential accumulation of wealth and its concentration in the hands of a class which takes over the producers by increasing the number of slaves: slavery becomes the dominant form of production.

Marx writes of slavery as such even less than Engels does. He deals with it only in comparison with other modes of production. In order to understand the thrust of his ideas on the subject, his writings in the Formen3 need to be distinguished from those in Capital. In the Formen, Marx almost always associates slavery with serfdom. His remarks are more suggestive than workable, and are often confused. Sometimes he sees slavery as ‘the development of property based on tribalism’;4 sometimes as the result of the extension of the family, in which case slavery is ‘latent’ (ibid. 122–3).5 Sometimes he seeks the origins of slavery in the appropriation of the means of subsistence (ibid. 101); sometimes in conquest (ibid. 93). Marx does not resolve the question of the possible endogenous development of slavery, nor of its historical emergence through contact between civilizations. He does not make clear the organic link between slaves as a class and their masters, in spite of a pertinent comment (ibid. 95) on the historical nature of the individualization of class relations;6 he does not distinguish clearly between the systems of subordination which are set up between kin in the relations of agricultural production and those which result from capture. Some of his comments help us to understand the ‘juridical’ confusion between subjects, family

2  This is also Finley’s thesis (1981): he considers that the demand for slaves precedes supply. 3  This is the name usually given to a fragment of the Grundrisse, in Marx (1964). 4  Marx, ibid. 5  It is precisely this position which I have criticized in Miers and Kopytoff’s work (1977). 6  ‘But man is individualized only through the process of history.’ The same is true of apparently individualized relations such as that which binds the slave to his or her master (Marx, 1964).

154

Meillassoux

dependants and slaves (ibid. 92), but they do not resolve the problem of the specificity of the slave relation. Marx’s observations in Capital (Marx, 1867/1970), while they are less dense and do not constitute a theoretical corpus which would lead to the classification of slavery as a mode of production, are characterized by a rigour which is not to be found in the Formen. Marx broadly distinguishes two main forms of slavery: in the first, so-called ‘patriarchal’ slavery, ownership of an individual may be an ‘accident’ and in any case the labour of the slave is directed towards the direct production of the means of subsistence (Capital III, ch. XX: 332)—in other words, use-value. With the effects of trade, patriarchal slavery can develop into a system geared to the production of surplus-value (Capital I, ch. X: 236) in which the slave is subject to fiercer and fiercer exploitation, as exchange-value develops (Capital I, ch. X: 236). It is this form of slavery, which produces exchange-value, to which Marx most often refers. Although Marx still, in Capital, often associates slavery with ‘serfdom’, he characterizes the former by the fact that it requires an initial outlay of money which he assimilates to fixed capital (Capital II, ch. XX: 483). The benefit gained by the owner is seen as interest on the capital advanced, or as anticipated, capitalized surplus-value, or as profits (if capitalist concepts are dominant), or as rent: ‘Whatever it may be called, the available surplus-product appropriated ... is here the normal and prevailing form’ (Capital III, ch. XLVII: 804). But the advance on fixed capital invested in the purchase of the slave forces the owner to invest further capital in exploitation of the slave. Thus, there is necessarily a choice to be made between the purchase of more slaves and that of the means of production which would increase their productivity. In the first case (that of slavery in ancient times), as Engels states (1884: 111), the number of slaves may be considerable: up to eighteen times the number of free men. The relation between masters and slaves, ‘which appears as the motor of production’, would still exclude the reification of the relations of production (Capital III, ch. XLVIII: 831). Because of the inequality of these social relations, Marx notes à propos of Aristotle, slavery obscures the equality of tasks and thus the conception of value in the minds of men (Capital I, ch. I: 59). When, as in American slavery, the labour of the slave is mediated by investments (however limited, as Marx notes), domination over men tends once again to be achieved through domination over things (Engels, 1884: 208). The surplus-labour of the slave increases as soon as it ‘is no longer a question of obtaining from him a certain quantity of useful products’ (Capital I, ch. X: 236). The slave trade enables the easy replacement of slaves by ‘foreign Negroes’ whose lifespan is less important than their productivity (Capital I, ch. X: 266). Marx touches on the problem of reproduction, emphasizing that in America

The Anthropology of Slavery

155

natural growth was insufficient and the slave trade was necessary to meet the needs of the market (Capital II, ch. XX: 483). Engels similarly notes that the slaves of Rome reproduced themselves to a very limited extent and that colossal supplies of slaves ensured by war were a precondition for the development of the great landed estates. The continual wars the Germans waged among themselves, like those between the Saxons and the Normans, were also designed to supply the slave market (1884: 200). In fact, Roman slavery disappeared with the decline of trade and of the towns, and the development of colonies and of serfdom (ibid. 1398; 1877: 362). It should be noted that Marx, and especially Engels, referring to slaves in ancient Europe, point to the social role of certain slaves—because they did not belong to any gens—as favourite slaves, with access to wealth, honours and high rank; these slaves constituted the basis for a new nobility among the Franks and the Germans. This nobility was doubtless associated with those military suites who waged war on their own account and through wars of plunder supplied the slave market. Despite the absence of a systematic study of the phenomenon of slavery, many of these reflections are still relevant. A few points should, however, be corrected. So-called ‘patriarchal’ slavery should not—because of its ‘accidental’ character—be identified as a class relation and does not in itself lead to a slave system of production. As I see it, this is not, strictly speaking, slavery, but rather an isolated phenomenon of subservience. The distinction between subsistence slavery, which produces a rent in food, and slavery which generates profit must be retained, but the two are not always mutually exclusive. While subsistence slavery does dominate military and ancillary slavery in aristocratic and military societies, it still continues to be an indispensable basis for the creation of profit in merchant slavery. Marx and Engels suggest, as a general economic condition for the emergence of slavery, an increase in productivity such that labour-energy is able to supply more than what is necessary for the subsistence of the producer.7 But in any society adults always produce more than they themselves need, in order to feed younger generations. It is this share, reserved for reproduction, which is appropriated by the slave-owner when he tears the adult out of his milieu to produce for him and finally when he replaces him, once he is no longer 7  The same reasoning can be found in the work of contemporary authors such as Leslie White (1969: 128): ‘Slavery did not exist during the hundreds of thousands of years before Neolithic times because culture had not developed sufficiently to make it possible for a producer to be more than self-supporting.’ Strictly speaking, this assertion implies the extinction of the human race before its emergence.

156

Meillassoux

productive, with another slave. Slavery is thus the only mode of production which allows the human surplus-product to be appropriated independently of increases in the productivity of labour over the level of ‘simple reproduction’. Serfdom, on the contrary, necessitates higher productivity, since the serf has to ensure, at the very least, both his own ‘simple reproduction’ and that of his masters. The weakness of Marx and Engels’s reasoning lies in their repeated confusion of slavery with serfdom—a confusion which also affects their arguments on the problem of value8 and on the relationship between slavery and kinship. On the other hand, as Engels suggests, it is possible that relations between wide-ranging nomads and sedentary agriculturalists favoured slavery. Cattleraising nomads are, at one and the same time, economically dependent on sedentary agriculturalists who cultivate the subsistence goods they need, and militarily and logistically dominant through their control over animal energy. This energy—which enables the herds to feed themselves on the move—also provides a means of transport for long-distance trade or can be offered as a service in exchange for agricultural goods. It also provides speedy mounts which are efficient instruments of plunder and abduction. Not all cattle-raisers are involved in pillage, but wide-ranging nomads, better mounted and in contact with other sedentary peoples beyond the desert, can combine plunder with abduction and transport their surplus booty to distant clients. In this way the contact between pastoralists and sedentary peoples provides the opportunity for subservience, and nomadism supplies its logistics. But this contact still does not explain the demand for slaves from the client populations—in other words, the ‘genesis’ of slavery. The underlying cause of this mode of exploitation and of the confrontations between peoples which result from it is probably to be found in a centurieslong historical process. Slavery is a period in universal history which has affected all continents, sometimes simultaneously, sometimes successively. Its ‘genesis’ is the sum of all that happened during an indeterminate period in many places. African trade in slaves to the Maghreb, then to Europe, which is the origin of slavery in black Africa, merely took over from the trade in slaves which had lasted for centuries in Asia, on the European continent and around the Mediterranean. The Slavs supplied their contingent of ‘slaves’, the Esclavons theirs of ‘esclaves’; our ancestors the Gauls regularly sold their English captives to the Romans; the Vikings took captives and sold them while carrying 8  The exchange of commodities at their approximate value and not at their ‘price of production’ takes place, as in slavery, ‘so long as the means of production involved in each branch can be transferred from one sphere to another only with difficulty ...’ (Capital III, 1; ch. X: 177).

The Anthropology of Slavery

157

on their coastal trade. Muslim and Christian pirates took each other captive. The process of enslavement had been under way for a long time, and in order to explain it in Africa it is necessary to explain its emergence on the Euro-Asian continent. Yet paradoxically it is in Africa, the last of the continents to supply the slave trade, that some seek an explanation for the origins of slavery, on the basis of an endogenous development of societies which are still suspected of primitivism and isolation and which are, therefore, laboratories for retarded fantasies. It is not possible in this study to reconstitute a history of the emergence of slavery in the world. This work is not a general theory of slavery, but a theoretical essay on slavery based on my knowledge of a small part of Africa. But the history of slavery in Africa is indispensable to any understanding of the facts which accompany it: it is this history which reveals the specificity of the slave mode of production, gives an economic meaning to war and provides a means of interpreting certain forms of power. It shows that the phenomenon of slavery is embedded in a social and political complex with considerable geographical scope. The anthropological dimension is significant only in the context of the economy and the demography of all the people concerned, both the plundered and the plundering. This perspective offers a constantly widening dimension to research on slavery, a dimension which I have not, in this study, exhausted. In the regions of Africa where I have worked, slavery is within the reach of living memory. However, it is not present in all its variants. It is therefore tempting to seek data elsewhere. In some of my examples I have succumbed to this temptation, but I have tried to stop short of unruly comparisons. This study has shown me that the notion of slavery in Africa covers such a wide variety of situations that a similar heterogeneity can be suspected elsewhere—in ancient times or in the Americas, for example. Slavery, rigorously defined, may have universal characteristics, but its definition, which is the main object of this study, must be generally accepted if a real discussion is to take place. I have, therefore, decided not to undertake a criticism of the classical works on slavery in other periods or in other regions: first because their authors’ criteria for differentiation are not the same as mine, and secondly because, not being on my own ground, I would not have been able to locate areas of agreement and disagreement with any certainty. At a number of points, this work refers to my previous writing and I have reproduced parts of the argument for the sake of clarity. In order to make the exposition more readable, I have also compiled an alphabetical glossary of the notions and concepts used in the body of the text (see p. 335). This exercise has shown me how easy it is to drift away from one’s own vocabulary, and has

158

Meillassoux

forced me to take up the work again several times to steer it back to coherence. I am not sure that I have succeeded. Finally, I would like to emphasize that, in their essentials, my information and my analysis have been fed by the very rich contributions of my colleagues in a previous collaborative work.9 Bibliography Abitbol, M., 1979, Tombouctou et les Arma, Maisonneuvre & Larose. Aguessy, C., 1956, ‘Esclavage, colonisation et tradition au Dahomey (sud)’, Présence africaine, 6: 58–67. Aguessy, H., 1970, ‘Le Dan-Homè au XIXe siècle était-il une société esclavagiste?’, Rev. franç. d’Et. pol. afr., 50: 71–91. Akindele, A. and Aguessy, C., 1953, Contribution à l’Étude de l’histoire de l’ancien royaume de Porto-Novo, Dakar, ifan. Akinola, G. A., 1972, ‘Slavery and Slave Revolts in the Sultanate of Zanzibar in the Tenth Century’, Jl of the Hist. Soc. of Nigeria, 6: 215–27. Al-Bakri [1968], ‘Routier de l’Afrique blanche et noire du Nord-Ouest’, trad. notes V. Monteil, B. IFAN, B, 30, 1: 39–116. Al-Bekri, see Al-Bakri, El-Bekri. Alencastro, L.-F. de, 1981, ‘La traite négrière et les avatars de la colonisation portugaise au Brésil et en Angola’, Cah. du CRIAR, 1: 9–76. Almeida-Topor, H. d’, 1984, Les Amazones, Paris, Rochevignes. Al-Omari [1927], Masàlik el Afsàr, I: L’afrique moins l’Egypte (transl. GaudefroyDemombynes), Paris, P. Geuthner. Alpers, E. A., 1975, Ivory and Slaves in East Central Africa, London, Heinemann. Alpers, E. A., 1983, ‘The Story of Swema: Female Vulnerability in 19th Century East Africa’, in Robertson and Klein, 1983: 185–99. Amselle, J.-L., 1977, Les négociants de la savane, Paris, Anthropos. Amselle, J.-L., Doumbia, Z., Kuyate, Y., Tabure, M., 1979, ‘Littérature orale et idéologie: la geste des Jakite-Sabasni du Ganan (Wasolon-Mali)’, Cah. Et. afr., 19 (1–4): 73–6; 381–439. Antoine, R., 1974, ‘Adventure d’un jeune négrier français d’après un manuscrit inédit du XVIIIe siècle’, Notes africaines. Arhin, K., n.d., Peasants in Asante in the Nineteenth Century, 39 p. mult. Aristotle [1958], Economics, Paris, Libr. Phil. J. Vrin. 9  Meillassoux, ed., L’esclavage en Afrique pré-coloniale, Paris, Maspero, 1975.

The Anthropology of Slavery

159

Aubin-Sugy, C., 1975, Economic Growth and Secular Trends in the Pre-Colonial Sudanic Belt, Ph.D., New York, Columbia U., Faculty of History. Augé, M., 1969, ‘Statut, pouvoir et richesse: relations lignagères, relations de dépendance et rapports de production dans la société alladian’, Cah. d’Etudes afr., 9, 3 (35), 1969: 461–81. Augé, M., 1975, Les faiseurs d’ombres, in Meillassoux (ed.), 1975b: 455–76. Austin, M. and Vidal-Naquet, P., 1972, Economies et sociétés en Grèce ancienne, Paris, A. Colin. Aymard, M., 1983, ‘Autoconsommation et marchés: Chayanov, Labrousse ou Le RoyLadurie?’, Annales, 38, 6, Nov.–Dec. 1983: 1392–411. Azevedo, M. J., 1980, ‘Precolonial Sara Society in Chad and the Threat of Extinction due to Arab and Muslim Slave Trade, 1870–1917’, Journal of African Studies, 7, 2: 99–108. Ba, A.-H. and Daget, J., 1962, L’Empire peul du Macina (I), Paris, Mouton & Co. Ba, A.-H. and Kesteloot, L., 1968, ‘Une épopée peule: Silimaka’, L’Homme, 8, 1: 5–36. Baduel, C., Meillassoux, C., 1975, ‘Modes et codes de la coiffure ouest-africaine’, L’Ethnographie, n.s., 69 (1975–1): 11–59. Baer, G., 1967, ‘Slavery in 19th-century Egypt’, Jl of Afr. Hist., 8, 3: 417–41. Balandier, G., 1965, La vie quotidienne au royaume du Kongo, Paris, Hachette. Balde, M.-S., 1975, ‘L’esclavage et la guerre sainte au Fuuta-Jalon’, in Meillassoux (ed.), 1975b: 183–220. Barnet, M. (ed.), 1967, Esclave à Cuba, Paris, Gallimard. Barry, B., 1972, Le royaume du Waalo, Paris, Maspero. Barth, Henry, 1857–8, Travels and Discoveries in North and Central Africa, London, Frank Cass [1965], 3 vols. Bataille, H., 1967, La part maudite, Paris, Ed. de Minuit. Bathily, A., 1975, ‘A Discussion of the Traditions of Wagadu’, Bull. IFAN. B. 37, 1: 1–94. Bathily, A., 1985, Guerriers, tributaires et marchands, doctoral thesis, Etat ès lettres, Dakar, 3 vols. Bathily, A. and Meillassoux, C., 1976, Lexique soninke (sarakole)-français, Dakar, Centre Linguistique appliquée. Battuta, Ibn [1968], Voyages d’Ibn Battûta (transl. C. Defremery and B.-R. Sanguinetti, Paris, Anthropos, reissue of 1854 edition (preface by V. Monteil), 4 vols. Bay, E. G., 1983, ‘Servitude and Worldly Success in the Palace of Dahomey’, in Robertson and Klein, 1983: 340–68. Bazin, J., 1972, ‘Commerce et prédation: l’Etat bambara de Segou et ses communautés maraka’, Conference on Manding Studies, 27 p. mult. Bazin, J., 1975, ‘Guerre et servitude à Ségou’, in Meillassoux (ed.), 1975b: 135–81. Bazin, J., 1979, ‘La production d’un récit historique’, Cah. Et. africaines, XIX, 1–4: 435–84. Bazin, J., 1982, ‘Etat guerrier et guerres d’Etat’, in Bazin and Terray (eds), 1982: 319–72.

160

Meillassoux

Bazin, J. and Terray, E. (eds), 1982, Guerres de lignages et guerres d’Etats en Afrique, Ed. des Archives Contemporaines. Becker, C. and Martin, V., 1975, ‘Kayor et Baol: Royaumes sénégalais et traite des esclaves au XVIIe siècle’, Revue franç. d’Hist. d’OM, 62, 226–7; 270–300. Bedaux, R. M. A. et al., 1978, ‘Recherches archéologiques dans le Delta intérieur du Niger’, Paleohistoria, 20: 91–220. Bellat, Cap., 1893, Renseignements historiques sur le Sansanding et le Macina, Arch. Dakar, 1 G 184. Benveniste, E., 1969, Le vocabulaire des institutions indo-européennes, Ed. de Minuit, 2 vols. Berlin, I., 1980, ‘Time, Space and the Evolution of Afro-American Society on British Mainland, North America’, Amer. Hist. Rev., 85, 1: 44–78. Berlioux, E.-F., 1870, La traite orientale, Paris, De Guillaumin. Bernardi, B., Poni and Triulzi, 1978, Fonti orali, Anthropologia e Storia, Milan, F. Angeli. Bernus, E., 1976, ‘Les Touaregs et la guerre’, communication Séminaire EPHESS, Paris. Binger, Cap., 1892, Du Niger au golfe de Guinée, Paris, Hachette, 2 vols. Blachère, R. (transl.), 1980, Le Coran, Paris, G.-P. Maisonneuve & Larose. Blanc, R., 1957, Manuel de recherche démographique en pays sous-développés, Paris, Service des Statistiques des Territoires d’Outre-Mer. Bloch, Marc, 1947, ‘Comment et pourquoi finit l’esclavage antique’, Annales, 1, Jan.–Mar. 1947: 30–40. 2, Apr.–Jun. 1947: 161–70. Bloch, Marc, 1939 [1968], La société féodale, Paris, Albin Michel. Bloch, Marc, 1955, Les caractères originaux de l’histoire rurale française, Paris, A. Colin. Bloch, Marc, 1963, ‘Les colliberti: étude historique sur la formation de la classe servile’, in Mélanges historiques, SEVPEN, 1: 385–451. Bloch, Maurice, 1975, ‘Property and the End of Affinity’, in Bloch, M. (ed.), 1975: 203–28. Bloch, Maurice (ed.), 1975, Marxist Analysis and Social Anthropology, London, Malaby Press, ASA Studies. Bohannan, P., 1963, Social Anthropology, NY, Holt, Rinehart & Winston. Bonnafé, P., 1975, ‘Les formes d’asservissement chez les Kukuya d’Afrique centrale’, in Meillassoux (ed.), 1975b: 529–56. Bonte, P., 1975, ‘Esclavage et relations de dépendance chez les Touaregs Kel Gress’, in Meillassoux (ed.), 1975b: 49–76. Bosnian, G., 1705, Voyage de Guinée (transi, from Dutch), Utrecht. Botte, R., 1974, ‘Processus de formation d’une classe sociale dans une société africaine récapitaliste’, Cah. Et. afr., 56, 14/4: 605–26. Bouche, D., 1968, Les villages de liberté en Afrique noire française, Paris, Mouton. Bourgeot, A., 1975, ‘Analyse des rapports de production chez les pasteurs et les agriculteurs de l’Ahaggar’, in Monod, T. (ed.), 1975: 263–83.

The Anthropology of Slavery

161

Boutillier, J.-L., 1968, ‘Les captifs en AOF (1903–1905)’, Bull IFAN, B, 30, 2. Boutillier, J.-L., 1975, ‘Les trois esclaves de Bouna’, in Meillassoux (ed.), 19756: 253–80. Bovill, G. W., 1968, The Golden Trade of the Moors, London, OUP. Bowdich, T. E., 1819, A Mission from Cape Coast Castle to Ashantee, London. Bradbury, R. E., 1957, The Benin Kingdom and the Edo speaking Peoples of South-Western Nigeria, London, IAI. Bradbury, R. E., 1967, ‘The Kingdom of Benin’, in Forde and Kaberry, 1967: 1–35. Brisson J.-P., 1959, Spartakus, Paris, Club Français du Livre. Brun, Th. and Layrac, C., 1979, ‘Fragilité de l’ajustement production-consommation en zone soudano-sahélienne’, Cah. Nutr. Diét., XIV, 1: 33–9. Brunet, R., 1981, ‘Géographie des Goulag’, L’Espace géographique, 3: 215–32. Brunschwig, R., 1960, ‘Abd’, in L’Encyclopédie de l’Islam (2nd French edn), Paris, Brill, Leide & Besson, I: 25–41. Bucher, H. H., 1979, ‘Liberty and Labor: The Origins of Libreville Reconsidered’, Bull. IFAN, B, 41, 3: 478–96. Bugner, L., 1976, L’image du Noir dans l’art occidental, Paris, Office du Livre, 3 vols. Burnham, P., 1980, ‘Raiders and Traders in Adamawa: Slavery as a Regional System’, in Watson (ed.), 1980: 43–72. Burton, R. F., 1864, A Mission to Glegle, King of Dahomy, London, 2 vols. Busnot (Père), 1690 [1925], Récit d’aventure au Maroc au temps de Louis XIV, Paris, Ed. Pierre Roger. Butler, J. (ed.), 1964, Boston Univ. Papers on African History (vol. 1), Boston, Boston U. Press. Ca da Mosto, A., 1895, Relation des voyages à la côte occidentale d’Afrique (1455–1457), Paris, Leroux. Cahen, C., 1968, L’Islam, des origines au début de l’Empire ottoman, Paris, Bordas. Caillié, R., 1830 [1965], Journal d’un voyage à Tembouctou et à Jenné dans l’Afrique centrale, reprinted Anthropos, 3 vols. Cairnes, J. E., 1863 [1958], The Slave Power, reprinted David & Charles, Newton Abbot. Camps, G., 1970, ‘Recherches sur les origines des cultivateurs noirs du Sahara’, Revue de l’Occident musulman, 7: 35–45. Canot, Cap., 1854 [1938], Vingt années de la vie d’un négrier, Paris, Mercure de France. Centre d’Etudes juridiques comparatives (ed.), 1979, Sacralité, pouvoir et droit en Afrique, Paris, Ed. CNRS. Centre de Recherche d’Histoire ancienne (ed.), 1972, Actes du Colloque 1971 sur l’esclavage, Paris, Les Belles-Lettres, ‘Ann. litt, de l’Univ. de Besançon’. Centre de Recherche d’Histoire ancienne (ed.), 1976, Textes, politique, idéologie Cicéron, Paris, Les Belles-Lettres, ‘Ann. litt, de l’Univ. de Besançon’. Centre de Recherche d’Histoire ancienne (ed.), 1979, Terre et paysans dépendants dans les sociétés antiques, Paris, CNRS.

162

Meillassoux

Centre of African Studies (ed.), 1977, 1981, African Historical Demography, Univ. of Edinburgh. Chastanet, M., 1983, ‘Les crises de subsistances dans les villages soninke du cercle de Bakel, de 1858 à 1945’, Cah. Et. afr., 23 (89–90), 1–2: 5–36. Chauveau, J.-P., 1978, ‘Contribution à la géographie historique de l’or en pays baule (Côte-d’Ivoire)’, J. des Afric., 48, 1: 15–70. Chauveau, J.-P., 1979, Kokumbo et sa région, Paris, ORSTOM. Chauveau, J.-P., 1980, ‘Spécialisations écologiques. Etat et réalisation de la valeur par les échanges à longue distance’, Cah. Et. afr., 77–8; 161–7. Chayanov, A. V., 1925 [1966], The Theory of Peasant Economy, Home-wood, R. D. Irwin, Inc. Chilver, E. M. and Kaberry, P. M., 1967, ‘The Kingdom of Kom in West Cameroon’, in Forde and Kaberry, 1967: 123–51. Ch’ü Ts’ung-tsu, Han Social Structure, vol. I: Han Dynasty China, Seattle, Un. of Washington Press. Cissoko, S. M., 1968, ‘Famines et épidémies à Tombouctou et dans la boucle du Niger du XVIe au XVIIIe siècle’, Bull. IFAN, B, 30, 3: 806–21. Cissoko, S. M., 1969a, ‘Traits fondamentaux des sociétés du Soudan occidental, du XVIIe siècle au début du XIXe s.’, Bull. IFAN, B, 31, 1: 1–30. Cissoko, S. M., 1969b, ‘La royauté (Mansaya) chez les Mandingues occidentaux d’après leurs traditions orales’, Bull. IFAN, B, 31, 2: 325–38. Cissoko, S. M., 1975, Tombouctou et l’Empire songhay. Dakar. Nouv. Ed. Afr. Cissoko, S. M. and Sambou, K., 1974, Recueil des traditions orales des Mandingues de Gambie et de Casamance, Niamey, Centre Rég. pour la Trad. or. CLAD, Lexique wolof-français, Dakar, IFAN, 3 vols. Clarence-Smith, W. G., 1979, ‘Slaves, Commoners and Landlords in Bulozi’, Jl of Afr. Hist., 20, 2: 219, 234. Coale, A. J. and Demeny, P., 1966, Regional Models-life Tables and Tables of Population, Princeton University Press. Contamine, P., 1976, ‘Guerre et christianisme médiéval’, in Glénisson J, (ed.), 1976. Cooper, F., 1977, Plantation Slavery on the East Coast of Africa, New Haven, Yale UP. Cooper, F., 1979, ‘The Problems of Slavery in African Studies’, Jl of Afr. Hist., 20, 1: 103–25. Cooper, F., 1981, ‘Islam and cultural hegemony’, in Lovejoy (ed.), 1981: 271–307. Cooper, J., 1876, Le continent perdu ou l’esclavage et la traite en Afrique (1875), Paris, Hachette. Coquery, C. (ed.), 1965, La découverte de l’Afrique, Paris, Julliard, ‘Archives’. Coquery-Vidrovitch, C., 1971, ‘De la traite des esclaves à l’exportation de l’huile de palme et des palmistes au Dahomey, XIXe siècle’, in Meillas-soux, C. (ed.), 1971: 107–23. Cornu, R. and Lagneau, J., 1969, Hiérarchies et classes sociales, Paris, A. Colin.

The Anthropology of Slavery

163

Corro, B., 1975, L’esclavage dans les Andes, Mém, de maîtrise ethnologie, Un. de Bordeaux II. Coser, L. A., 1964, ‘The Political Functions of Eunuchism’, Amer. Soc. Rev., 29, 6: 880–5. Coulibaly, T., 1979, ‘La Communautée noire d’Algérie’, Peuples noirs, Peuples africains, 2, 9: 115–23. Crone, P., 1980, Slaves on Horses: The Evolution of the Islamic Policy, London, Cambridge U. Press. Cuoq, J. M., 1975, Recueil des sources arabes concernant l’Afrique occidentale du VIII e au XVI e siècle, Paris, CNRS. Curtin, P. D. and Vansina, J., 1964, ‘Sources of the XIXth Cent. African Slave-Trade’, Jl of Afr. Hist., 2: 185–208. Curtin, P. D. (ed.), 1966, Narratives of the West African Slave Trade, Madison, Un. of Wisconsin Press. Curtin, P. D., 1967, Africa Remembered, Madison, Un. of Wisconsin Press. Curtin, P. D., 1969, The Atlantic Slave Trade, Madison, Un. of Wisconsin Press. Curtin, P. D., 1975, Economic Change in Pre-colonial Africa, Madison, Un. of Wisconsin Press. Cuvillier, A., 1958, Manuel de sociologie, Paris, PUF, 2 vols. Daaku, K. Y., 1970, Trade and Politics on the Gold Coast, 1600–1720, Oxford, Clarendon Press. Daaku, K. Y., 1971, ‘La traite des Noirs et la société africaine’, in Perspective nouvelle sur l’histoire africaine, Paris, Présence, Africaine: 152–8. Dalby, D. (ed.), 1970, Language and History in Africa, London, F. Cass. Dalzel, A., 1793, The History of Dahomey, London, G. and W. Nicol. Daubigney, A., 1976, ‘Contribution à l’étude de l’esclavagisme: la propriété chez Cicéron’, in Centre de Recherche d’Histoire ancienne, 1976: 13–72. Daumas, Lt-Col., E., 1845, Le Sahara algérien, Paris, Fortin-Masson. Daumas, Gen. E., 1857, Le Grand Désert (suivi du Code de l’esclavage chez les Musulmans), Paris, M. Levy Frères. Daumas, Gen. E., 1858, Les chevaux du Sahara et les mœurs du désert, Paris, M. Levy Frères. Debien, G.: 1961–1967, ‘Les origines des esclaves aux Antilles’, Bull. IFAN, B, 23, 3–4, 1961: 363–87. 25, 1–2, 1963: 1–38. 25, 3–4, 1963: 215–65. 26, 1–2, 1964: 166–211. 26, 3–4, 1964: 601, 675. 27, 3–4, 1965: 755–99. 29, 3–4, 1967: 536–58. Debien, G., Delafosse, M., Thilmans, G., 1978, ‘Journal d’un voyage de traite en Guinée, à Cayenne, aux Antilles fait par Jean Barbot en 1678– 1679’, Bull. IFAN, B, 40, 2: 235–395. Deherme, P., 1908, Rapport sur la captivité en Afrique Occidentale, Archives Dakar, K25. De Jonghe, E., 1949, Les formes d’asservissement dans les sociétés indigènes du Congo belge, Bruxelles, Inst. roy. col. belge.

164

Meillassoux

Delafosse, M., 1912, Haut-Sénégal—Niger, Paris, Larose, 3 vols. Delaunay, D., 1984, De la captivité à l’exil, Bondy, ORSTOM. Delcourt, A., 1952, La France et les Etablissements français au Sénégal entre 1713 et 1763, Mémoire IFAN, no. 17, Dakar. Delobsom, A.-A.-D., 1933, L’empire du Mogho-Naba, Paris, Domat-Mon-chrestien. Demougeot, A., 1949, ‘L’esclavage et l’émancipation des Noirs au Sénégal’, Tropiques, n.s., 47, 312: 10–17. Depestre, René, Minerai noir, Poèmes, Paris, Présence Africaine. Dermenghem, E., 1956, Mahomet et la tradition islamique, Paris, Seuil. Derrick, J., 1975, Africa’s Slaves Today, London, Allen & Unwin. Deschamps, H., 1971, Histoire de la traite des Noirs de l’Antiquité a nos jours, Paris, Fayard. Devisse, J., 1970, ‘La Question d’Audagust’, Tegdaoust I: 109–57. Diabaté, M.-M., 1970a, Janjon, Paris, Présence Africaine. Diabaté, M.-M., 1970b, Kala Jata, Bamako, Ed. Populaires, coll. ‘Hier’. Diallo, O., 1947, ‘Les Kore et les Simo au Fouta Dialon’, Notes afr., 33: 2–5. Dieng, A.-A., 1974, Classes sociales et mode de production féodal en Afrique de l’Ouest, Dakar, 15 p. mult. Dieng, A.-A., 1983, Contribution à l’étude des problèmes philosophiques en Afrique noire, Paris, Nubia. Dinafla, Lexique bambara-français, Bamako, DINAFLA. Diop, A.-B., 1981, La société wolof Paris, Karthala. Diop, A.-S., 1978, ‘L’impact de la civilisation manding au Sénégal. La génèse de la royauté gelwar au Siin et au Saalum’, Bull. IFAN, B, 40, 4: 689–707. Diop, C. A., 1960, L’Afrique noire précoloniale, Paris, Présence Africaine. Diop, M., 1971–2, Histoire des classes sociales dans l’Afrique de l’Ouest, Paris, Maspero, 2 vols. Dockes, P. and Servet, J.-M., 1980, Sauvages et ensauvagés, PU de Lyon. Dorman, W., 1973, Serfs, Peasants and Socialists, Berkeley, Un. of California Press. Doutressoule, G., 1940, ‘Le cheval au Soudan français’, B. IFAN, B, 2, 3–4: 324–46. Dubois, F., 1897, Tombouctou la Mystérieuse, Paris, Flammarion. Duby, G., 1973, Guerriers et paysans, Paris, Gallimard. Duby, G., 1981, Le chevalier, la femme et le prêtre, Paris, Hachette, ‘Pluriel’. Dumestre, G. (ed.), 1979, La geste de Ségou, Paris, A. Colin, ‘Les Classiques africains’. Dumestre, G. and Kesteloot, L. (eds), 1975, La prise de Dionkoloni, Paris, A. Colin, ‘Classiques africains’. Dunbar, R. A., 1977, ‘Slavery and the Evolution of XIXth Century Damagaram’, in Miers and Kopytoff, 1977: 155–77. Dupire, M., 1970, Organisation sociale des Peuls, Paris, Pion. Edrisi [1866], Description de l’Afrique et de l’Espagne (transi. R. Dozy and M.-J. de Goeje), Leyde, Brill.

The Anthropology of Slavery

165

Ekholm, K., 1972, Power and Prestige, the Rise and Fall of the Kongo Kingdom, Uppsala, Skriv Service. El-Bekri, A.-O., 19th cent. [1965], Description de l’Afrique septentrionale (transl, de Slane), Paris, A. Maisonneuve. Elwert, G., 1973, Wirtshaft und Herrshaft von Dahomey, Munich, Kommission Verlag K. Rinner. Emmer, P., Mettas, J., Nardin, J.-E. (eds), 1975, ‘La traite des Noirs par l’Atlantique: nouvelles approches’, special issue, Rev. franç. d’OM, 42: 226–7. Engels, F., 1877–8 [1950], Anti-Dühring, Paris, Ed. Sociales. Engels, F., 1884 [1954], L’origine de la famille, de la propriété privée et de l’Etat, Paris, Ed. Sociales. Engerman, S. L. and Genovese, E. D. (eds), 1975, Race and Slavery in the Western Hemisphere: Quantitative Studies, Princeton, Princeton University Press. Engerman, S. L., 1981, ‘L’esclavage aux Etats-Unis et aux Antilles anglaises: quelques comparaisons économiques et démographiques’, in Mintz S. (ed.), 1981: 223–46. Es-Sa’di, A. [1964], Tarikh es-Soudan, Paris, Leroux. Ezeanya, S. N., 1967, ‘The osu System in Igboland’, Jl Relig. in Afr., 1, 1: 35–45. Fabre, M., 1970, Esclaves et planteurs dans le Sud américain, Paris, Julliard, ‘Archives’. Fabre, J. D., 1964, ‘Some Thoughts on State Foundation in the Western Sudan before the 17th Cent.’, in J. Butler (ed.), 1964: 17–34. Fage, J. D., 1969, ‘Slavery and the Slave-Trade in the Context of West African History’, Jl of Afr. Hist., 10, 3: 393–404. Fage, J. D., 1974, State and Subjects in SubSaharan African History, Raymond Dart Lectures, Witwatersrand Univ. Fage, J. D., 1980, ‘Slave and Society in Western Africa, c. 1445–c. 1700’, Jl of Afr. Hist., 21, 3: 289–310. Farias, P. F. de M., 1980, ‘Model of the World and Categorial Models: The “Enslavable Barbarian” as a Model of Classificatory Label’, Slavery and Abolition, 1, 2: 115–31. Fayet, C., 1931, Esclavage et travail obligatoire. La main-d’œvre volontaire en Afrique, Paris, Libr. Générale de Droit et de Jurisprudence. Filipowiak, W., 1979, Etudes archéologiques sur la capitale médiévale du Mali, Warsaw, Museum Narodowe w. Szczenie. Finley, M. I., 1960, Slavery in Classical Antiquity, Cambridge, Heffer. Finley, M., 1969, ‘The Idea of Slavery: Critique of David Brion Davis: The Problem of Slavery in Western Culture’, in Foner and Genovese, 1969: 260. Finley, M. I., 1973, The Ancient Economy, Berkeley, Univ. of California Press. Finley, M. I., 1979, Slavery and the Historians, Histoire sociale, 12, 24: 247–61. Finley, M. L, 1981, Esclavage antique et idéologie moderne, Paris, Ed. de Minuit. Fisher, A. G. B. and Fisher, N. J., 1970, Slavery and Muslim Society in Africa, London, Hurst.

166

Meillassoux

Fisher, H. J. and Rowland, V., 1971, ‘Firearms in the Central Sudan’, Jl of Afr. Hist., 12: 215–39. Fohlen, Cl., 1975, ‘L’esclavage aux Etats-Unis’, Revue franç. d’Hist. d’Outre-mer, 52: 226–7; 372–83. Foner, L. and Genovese, E. D., (eds), 1969, Slavery in the New World, Englewood Cliffs, NJ, Prentice Hall. Forde, D., 1951, The Yoruba Speaking People of South Western Nigeria, London, Int. Afr. Inst. Forde, D. and Kaberry, P. M. (eds), 1967, West African Kingdoms in the 19th century, London, IAI, OUP. Fraser, D. and Cole, H. M. (eds), 1972, African Art and Leadership, Madison, Un. of Wisconsin Press. Freyre, G., 1933, Maîtres et esclaves, Paris, Gallimard. Frossard, M., 1789 [1889], La cause des esclaves nègres et des habitants de la Guinée, Lyon (micro-ed., Hachette). Fuglestad, F., 1977, ‘Quelques réflexions sur l’histoire et les institutions de l’ancien royaume du Dahomey et de ses voisins’, Bull. IFAN, B., 39, 3: 493–517. Gabriel, J. and Ben-Tovim, G., 1978, ‘Marxism and the Concept of Racism’, Economy and Society, 7, 2: 118–55. Gaden, H., 1931, Proverbes et maximes peuls et toucouleurs, Paris, Inst. d’Ethnologie. Gallais, J., 1967, Le delta intérieur du Niger, Dakar, IFAN, 2 vols. Galliéni, Ct, 1885, Voyage au Soudan français, Paris, Hachette. Gamble, D. P., 1957, The Wolof of Senegambia, London, Int. Afr. Inst. Garlan, Y., 1982, Les esclaves en Grèce ancienne, Paris, Maspero. Garnsey, P., 1976, ‘Peasants in Ancient Roman Society’, Jl of Peasant Studies, 3, 3: 221–35. Gaston-Martin, 1948, Histoire de l’esclavage dans les colonies françaises, Paris, PUF. Gautier, A., 1985, Les sœurs de solitude, Paris, Ed. Caribéennes. Gemery, H.-A. and Hogendorn, J. S., 1981, ‘La traite des esclaves sur l’Atlantique: essai de modèle économique’, in S. Mintz (ed.), 1981: 18–45. Genovese, E. D., 1968, Economie politique de l’esclavage, Paris, Maspero. Genovese, E. D., 1974, Roll, Jordan, Roll. The World the Slaves Made, New York, Pantheon. Geoffroy de Villeneuve, A., 1814, De l’Afrique, Paris, Nepveu, 4 vols. Gerbeau, H., 1967, ‘Un mort-vivant: l’esclavage’, Prés, africaine, 61, 1: 180–98. Gerbeau, H., 1970, Les esclaves noirs, Paris, A. Balland. Girard, J., 1969, Genèse du pouvoir charismatique en Basse-Casamance (Sénégal), Dakar, IFAN. Giraud, M., 1979, Races et classes en Martinique, Paris, Anthropos. Giraud, M. and Jamard, L.-L., 1976, Les Antillais et le travail, Paris, Grim-SCA, MSH. Gisler, A., 1965, L’esclavage aux Antilles françaises (XVII e–XIX e siècle), Contribution au problème de l’esclavage, Fribourg, Ed. Univ.

The Anthropology of Slavery

167

Gleave, M. B. and Prothero, R. M., 1971, ‘Population Density and Slave Raiding, A Comment’, Jl of Afr. Hist, 12, 2: 319–24. Glélé, M.-A., 1974, Le Danxome, Paris, Nubia. Glénisson, J. (ed.), 1976, La guerre au Moyen Age, Paris, Pons. Gomila, J., 1976, ‘Définir la population’, in Jacquard, 1976: 5–36. Goody, J. (ed.), 1968, Literacy in Traditional Society, Cambridge, CUP. Goody, J., 1971, Technology, Tradition and the State in Africa, London, IAI, OUP. Goody, J., 1980, ‘Slavery in Time and Space’, in Watson (ed.), 1980: 16–42. Grace, J., 1975, Domestic Slavery in West Africa, London, Frederick Muller Ltd. Graham, J. D., 1965, ‘The Slave-Trade, Depopulation and Human Sacrifice in Benin’, Cah. Et. afr., 5 (2), 18: 317–34. Granderye, Cap., 1947, ‘Notes et souvenirs sur l’occupation de Tombouctou’, Rev. Hist, col, 1947, 34: 87–131. Grant, D., 1968, The Fortunate Slave, London, OUP. GROMSCA, 1975, Reproduction des hiérarchies sociales et action de l’Etat, Paris, MSH, Ecology et Sciences humaines (mult.). Groupe universitaire de Recherche inter-Caraïbes, 1969, Le passage de la société esclavagiste à la société postesclavagiste, Pointe-à-Pitre. Gueye, M., 1966, ‘La fin de l’esclavage à Saint-Lous et à Gorée en 1848’, Bull. IFAN, B., 28, 3–4: 637–56. Gueye, M., 1965, ‘L’affaire Chautemps (avril 1904) et la suppression de l’esclavage de case au Sénégal’, Bull. IFAN, B, 27, 3–4: 543–59. Guillaume, E., 1895, Le Soudan en 1894, la vérité sur Tombouctou. L’esclavage au Soudan, Paris, Savine. Guillaume, H., 1974, Les nomades interrompus, Niamey, ‘Etudes nigériennes’, no. 35. Gutman, H., 1981, ‘Familles et groupes de parenté chez les Afro-Américains en esclavage dans les plantations de Good Hope (Car. du S.) 1760–1860’, in Mintz (ed.), 1981: 141–72. Hafkin, N. and Bay, E. G., 1976, Women in Africa, Stanford, Stanford Univ. Press. Halphen, L. (ed.), 1964, Histoire anonyme de la première croisade, Paris, Soc. d’Ed. Les Belles-Lettres. Harms, R., 1978, Competition and Capitalism: the Bobangi Role in Equatorial Africa’s Trade Revolution, Ph.D. Dissertation, Un. of Madison. Harms, R., 1983, ‘Sustaining the System: Trading Towns along the Middle Zaïre’, in Robertson and Klein (eds), 1983: 95–110. Harries, P., 1981, ‘Slavery, Social Incorporation and Surplus Extraction: the Nature of Free and Unfree Labour in SE Africa’, Jl of Afr. Hist, 22, 3: 309–30. Hecquard, H., 1853, Voyage sur la Côte et l’intérieur de l’Afrique occidentale, Paris, Bénard & Cie.

168

Meillassoux

Heers, J., 1981, Esclaves et domestiques au Moyen Age, Paris, Fayard. Héritier, F., 1975, ‘Des cauris et des hommes: production d’esclaves et accumulation de cauris chez les Samo (Haute-Volta)’, in Meillassoux (ed.), 19756: 477–508. Herskovits, M. J., 1978, Dahomey, an Ancient West African Kingdom, New York, Augustin, 2 vols. Hindess, B. and Hirst, P. Q., 1975, Pre-Capitalist Modes of Production, London, Routledge & Kegan Paul. Hogendorn, J. S., 1977, ‘The Economics of Slave Use on two “Plantations” in the Zaria Emirate of the Sokoto Caliphate’, Int. Jl of Afr. Hist. Stud., 10, 3: 369–83. Hogg, P., 1973, The African Slave Trade and its Suppression (A Classified Bibliography), London, Frank Cass. Holas, B., 1949, ‘A propos de l’étymologie du komo’, Notes afr., 42: 49. Hopkins, A. G., 1973, An Economic History of West Africa, London, Longman. Hopkins, K., 1978, Conquerors and Slaves, London, CUP. Horton, R., 1975, ‘On the Rationality of Conversion’, Africa, 45, 3: 219–35. 45, 4: 373–99. Horton, W. R. G., 1954, ‘The Ohu System of Slavery in a Northern Ibo Village’, Africa, 24, 4: 311–36. Hunkandrin, L., 1964, ‘L’esclavage en Mauritanie’, Ed. dahoméennes, 3: 31–49. Hunwick, J. O., 1970, ‘The Term Zanj and its Derivatives in West African Chronicles’, in Dalby (ed.), 1970: 102–8. Hunwick, J. O., 1973, ‘The Mid-Fourteenth-Century Capital of Mali’, J. Afr. Hist, XIV, 2: 195–208. Ibn Khaldun [1852], Histoire des Berbères, Algeria. Idrissi, see: Edrisi. Igbafe, P. A., 1975, ‘Slavery and Emancipation in Benin, 1897–1945’, Jl of Afr. Hist., XVI, 3: 409–29. Imbert, J., Sautel, G. and Boulet-Sautel, M., 1957, Histoire des institutions et des faits sociaux, Paris, PUF, 2 vols. Innes, G., 1974, Sunjata, Three Mandinka Versions, London, SOAS. Isaac, E., 1980, ‘Genesis, Judaism and the “Sons of Ham”’, Slavery and Abolition, 1, 1: 3–17. Isaacman, B. and A., 1977, ‘Slavery and Stratification among the Sena of Mozambique: A Study of the Kaporo System’, in Miers and Kopytoff (eds), 1977: 105–19. Isert, P.-E., 1973, Voyages en Guinée et dans les îles Caraïbes en Amérique, Paris, Maradan. Itoh, M., 1985, La crise mondiale, Paris, EDI. Izard, M., 1970, Introduction à l’histoire du royaume mossi, Paris, Ouagadougou, Recherches Voltaïques, 2 vols. Izard, M., 1975, ‘Les captifs royaux dans l’ancien Yatenga’, in Meillassoux (ed.), 1975: 281–97. Izard, M., 1976, ‘Formes d’organisation militaire dans la Haute-Volta de la seconde moitié du XIXe siècle’, communication, Séminaire EHESS, Paris.

The Anthropology of Slavery

169

Jacquard, A. (ed.), 1976, L’étude des isolais, espoirs et limites, Paris, INED. Jacques-Meunie, D., 1947, ‘Les oasis de Lektaoua et des Mehamid’, Hesperis, 3–4: 410–12. Jacques-Meunie, D., 1958, ‘Hiérarchie sociale au Maroc présaharien’, Hesperis, 3–4, 239–67. Jacques-Meunie, D., 1961, Cités anciennes de Mauritanie, Paris, Klincksieck. Jamard, J.-L. (n.d.), Les systèmes socio-économiques antillais, leurs modalités de constitution et de transformation, Paris, CORDES. Jamard, J.-L., 1982, ‘Le mode de production esclavagiste en Guadeloupe et en Martinique en rapport avec le mouvement de la société française’, Archipelago, 1: 57–95. Jaulin, R., 1971, La mort sara, Paris, UGE. Jawara, M., 1976–7, Monographie du royaume de Jara du XVIe au XVIIIe siècle, Mémoire, Bamako, Ecole normale supérieure. Jean Leon l’Africain, 1550/1956, Description de l’Afrique, Paris, Adrien-Maisonneuve. Jewsiewicki, B., 1981, ‘The social context of Slavery in Equatorial Africa during the 19th and 20th Centuries’, in Lovejoy (ed.), 1981: 41–72. Johnson, M., 1970, ‘The Cowrie Currencies of West Africa’, Jl of Afr. Hist., 11, 1: 17–49. 11, 3: 331–54. Johnson, M., 1976, ‘The Economic Foundation of an Islamic Theocracy: The Case of Masina’, Jl of Afr. Hist., 17, 4: 481–96. Johnson, M., 1980, ‘Polanyi, Peukert and the Political Economy of Dahomey’, Jl of Afr. Hist., 21, 3: 395–9. Jonckers, D. de, 1981, Organisation socio-économique des Minianka du Mali, thesis, Fac. Sc. soc, Univ. Libre de Bruxelles. Jones, A. and Johnson, M., 1980, ‘Slaves from the Windward Coast’, Jl Afr. Studies, 21, 1: 17–34. Josefsson, C., 1981, Kuba Slavery, unpublished, Göteborg Univ. Kaba, L., 1981, ‘Archers, Musketeers and Mosquitoes: The Moroccan Invasion of the Sudan and the Songhay Resistance (1591–1612)’, Jl of Afr. Hist., 22: 457–75. Kake, I. B., 1969, ‘L’aventure des Bukhara (Prétoriens noirs) au Maroc au XVIIIe siècle’, Prés, afr., 70, 2: 69–74. Kamissoko, W., 1975, L’empire du Mali (transi. Y.-T. Cissé), 1st International Colloquium, Bamako, SCOA. Kâti [1964], Tarikh cl-Fettach (transi. O. Houdas and M. Delafosse), Paris, E. Leroux. Kea, R. A., 1971, ‘Firearms and Warfare on the Gold and Slave Coast (16th to 19th Cent.)’, Jl Afr. Hist., 12: 185–213. Keim, C. A., 1983, ‘Women in Slavery among the Mangbetu, 1800–1910’, in Robertson and Klein (eds), 1983: 144–60. Kersaint-Gilly, F. de, 1924, ‘Essai sur l’évolution de l’esclavage en Afrique occidentale française. Son dernier stade au Soudan français’, Bull. Com. Et. hist, scient. AOF, 9: 669–78.

170

Meillassoux

Kesteloot, L. (ed.), 1972, Da Monzon de Ségou, Paris, F. Nathan, 4 vols. Kesteloot, L. (ed.), 1978, ‘Le mythe et l’histoire dans la formation de l’Empire de Ségou’, Bull. IFAN, B, 40, 3: 578–681. Kimba, I., 1979, Guerres et sociétés, thesis, UER Geog., Hist. and Soc. Sc., Paris VIII. Klein, H. S., 1983, ‘African Women in the Atlantic Slave Trade’, in Robertson and Klein (eds), 1983: 29–38. Klein, M. A., 1971, ‘Slavery, the Slave-Trade and Legitimate Commerce in the Late 19th Cent’., Et. d’Hist. Afr., 2: 5–28. Klein, M. A., 1972, ‘Social and Economic Factors in the Muslim Revolution in Senegambia’, Jl of Afr. Hist, 13, 3: 419–41. Klein, M. A., 1977, ‘Servitude among the Wolof and Sereer of Senegambia’, in Miers and Kopytoff (eds), 1977: 335–63. Klein, M. A., 1978, ‘The Study of Slavery in Africa’, Jl of Afr. Hist., 19, 4: 599–610. Klein, M. A. (n.d.), From Slave to Metayer in the French Sudan: An Effect at Controlled Social Change, 22 p. mult. (unpub.). Klein, M. A., 1983, Slavery, Forced Labour and French Rule in Colonial Guinea, Un. of Toronto, 15 p., mult. (unpub.). Klein, N., 1981, ‘The Two Asantes’, in Lovejoy (ed.), 1981: 149–67. Klein, N., Asante Slavery, mult. Kodjo, N.-G., 1976, ‘Contribution à l’étude des tribus dites serviles du Songaï’, Bull IFAN, B, 38, 4: 790–812. Kohler, J.-M., 1972, Les migrations des Mosi de l’Ouest, Paris, ORSTOM, ‘Travaux et Documents’, no. 18. Konare Ba, A., 1977, Sonni Ali Ber, Niamey, ‘Etudes nigériennes’. Koubbel, L., 1968, ‘Histoire de la Vallé du Niger supérieur et moyen du VIIIe au XVIe siècle: quelques réflexions sur le découpage chronologique’, Notes et Doc. voltaïques, 1, 4: 13–28. Kouloub, Out el, 1958, Ramza, Paris, Gallimard. Kouroubari, A., 1959, ‘Histoire de l’imam Samory’, Bull. IFAN, B, 21, 3–4: 544–71. Kula, W., 1970, Théorie économique du système féodal, Paris, Mouton. Labouret, H., 1955, ‘Le servage, étape entre l’esclavage et la liberté en Afrique occidentale’, Deutsche Afr. Studien, 1955: 147–53. Lacoste, Y., 1966, Ibn Khaldoun, Paris, Maspero. Lacroix, L., 1967, Les derniers négriers, Paris, Inter Presse. Lamiral, D. H., 1789, L’Afrique et le peuple affriquain, considérés sous tous leurs rapports avec notre commerce et nos colonies, Paris, Dessenne. Lanfry, J., 1966, Gadames, Algeria. Lange, D., 1972, ‘L’intérieur de l’Afrique occidentale d’après Giovanni Lorenzo Anania (XVIe siècle)’, Cah. d’Hist. mondiale, 14, 2: 299–351.

The Anthropology of Slavery

171

Lange, D., 1978, ‘Progrès de l’Islam et changement politique au Kanem du XIe au XIIIe siècle: un essai d’interprétation’, Jl of Afr. Hist., 19, 4: 495–514. La Roncière, C. de, 1933, Nègres et négriers, Paris, Portiques. Latour, E. de, 1981, Une aristocratie coloniale, thesis, Paris VIII. Latour, E. de, 1982, ‘La Paix destructice’, in Bazin and Terray (eds), 1982: 235–68. Laurentin, A., 1960, ‘Femmes Nzakara’, in Paulme (ed.), 1960: 121–72. Law, R., 1971, ‘The Constitutional Trouble of Oyo in the 18th Cent.’, Jl of Afr. Hist., 12, 1: 25–44. Law, R., 1975, ‘A West African Cavalry State: The Kingdom of Oyo’, Jl of Afr. Hist., 16, 1: 1–16. Law, R., 1976, ‘Horses, Firearms and Political Power in Pre-colonial West Africa’, Past and Present, 72: 112–32. Law, R., 1977a, The Oyö Empire: 1600–1836, Oxford, Clarendon Press. Law, R., 1977b, ‘Royal Monopoly and Private Enterprise in the Atlantic Trade. The Case of Dahomey’, Jl of Afr. Hist., 18, 4: 555–77. Lefebvre des Noëttes, 1924, La force animale à travers les âges, Paris, Berger-Levrault. Lefebvre des Noëttes, 1931, L’attelage, le cheval de selle à travers les âges. Contribution à l’histoire de l’esclavage, Paris, Picard, 2 vols. Lefebvre des Noëttes, 1932, ‘Autour du vaisseau de Boro-Boudan’, La Nature, 2885: 49–58. Leff, N. N., 1974, ‘Long Term Viability of Slavery in a Backward Closed Economy’, Jl of Interdisciplinary History, 5: 103–8. Legassick, M., 1966, ‘Firearms, Horses and Samorian Army Organization 1870–1898’, Jl Afr. Hist., 7: 95–115. Le Goff, J., 1965, La civilisation de l’Occident médiéval, Paris, Arthaud. Le Hérissé, 1911, L’ancien royaume de Dahomey, Paris, Larose. Lengellé, M., 1967, L’esclavage, Paris, ‘Que sais-je?’. Levtzion, N., 1971, ‘A 17th Century Chronicle by Ibn al Makhtar: A Critical Study of Tarikh al-Fattash’, Bull of the SOAS, 34, 3, 1971: 571–93. Levtzion, N., 1971, ‘Mahmud Kati fut-il l’auteur du Tarikh al-Fattash?’ Bull IFAN, B, 33, 4: 665–74. Levtzion, N., 1973, Ancient Ghana and Mali, London, Methuen & Co. Lévy-Bruhl, H., 1931, ‘Théorie de l’esclavage’, Revue générale du Droit, 55, 1931: 1–17. Leynaud, E. (n.d.) [1961], Les cadres sociaux de la vie rurale dans la Haute Vallée du Niger, Paris, BDPA, 2 vols. mult. Littré, E., 1877 [1959], Dictionnaire de la langue française, Gallimard/ Hachette, 7 vols. Lombard, J., 1967, ‘The Kingdom of Dahomey’, in Forde and Kaberry (eds): 70–92. Louis-Joseph, C., 1976, Capital marchand, esclavage et sucre aux Antilles françaises, 1635–1848, unpub. ms.

172

Meillassoux

Lovejoy, P. E., 1978, ‘Plantation in the Economy of Sokoto Caliphate’, Jl of Afr. Hist., 19, 3: 341–68. Lovejoy, P. E. (ed.), 1981, The Ideology of Slavery, Beverly Hills, Sage. Lovejoy, P. E., 1981, ‘Slavery in thé Sokoto Caliphate’, in Lovejoy (ed.), 1981: 201–13. Lovejoy, P. E. (ed.), 1983, Transformation in Slavery, London, CUP. Lovejoy, P. E. and Baier, S., 1975, The Desert Side Economy of the Central Sudan’, Intl Jl of Afr. Hist. Study, 8, 4: 551–81. Loyer, P. G. de, 1660–1715, in Roussier, 1935. Ly, A., 1958, La compagnie du Sénégal, Paris, Présence Africaine. Ly, B., 1967, ‘L’honneur dans les sociétés ouolof et toucouleur du Sénégal’, Prés, africaine, 61, 1: 32–67. McCall, D. F., 1967, The Horse in West African History’, Congrès international des Africanistes, Dakar. McCall, D. F. and Bennett, N. R. (eds), 1971, Aspects of West African Islam, Boston. MacCormack, C. P., 1977, ‘Wono: Institutionalized Dependency in Sherbro Descent Group’, in Miers and Kopytoff (eds): 181–200. Machiavelli, N., 1571 [1938], Le Prince, Paris, Ed. de Cluny. MacLeod, W. C., 1928, ‘Economic Aspects of Indigeneous American Slavery’, Am. Anthrop., n.s., 30: 632–50. MacLeod, W. C., 1929, ‘The Origin of the Servile Labor Groups’, Am, Anthrop., n.s. 30: 89–113. McCaskie, T. C., 1980, ‘Office, Land and Subjects in the History of the Manwere Fekuo of Kumase’, Jl of Afr. Hist., 21, 2: 189–208. McDougall, E. A., 1980, The Ijil salt Industry, Ph.D. Un. of Birmingham. Magalhaes-Godinho, V., L’économie de l’Empire portugais aux XV e et XVI e siècles, Paris, SEVPEN. Mage, E., 1868, Voyage dans le Soudan occidental (Sénégambie-Niger), Hachette, Paris. Mahibou, S.-M. and Triaud, J.-L., 1983, Voilà ce qui est arrivé, Paris, CNRS. Malowist, M., 1966, ‘Le commerce de l’or et des esclaves au Soudan occidental’, Africana Bull., 4: 49–72. Manning, P., 1969, ‘Slaves, Palm-oil and Political Power on the West African Coast: A Historical Hypothesis’, Afr. Hist. Stud., 2, 2: 279–88. Manning, P., 1975, ‘Un document sur la fin de l’esclavage au Dahomey (Rapport de l’administrateur Charles, 1904)’, Notes africaines, 147: 88–91. Manning, P., 1981, Slavery, Colonialism and Economic Growth in Dahomey, 1640–1960, Cambridge, CUP. Manning, P. (n.d.), The Enslavement of Africans: A Demographic Model, 71 p. mult., unpub. Marchai, F., Rabut, O., et al., 1972, ‘La conjoncture démographique: l’Afrique, l’Amérique latine et l’Asie’, Population, 27, 6: 1076–117.

The Anthropology of Slavery

173

Markov, W. (ed.), Etudes africaines, Leipzig. Marty, P., 1920–21, Etudes sur l’Islam et les tribus du Soudan, Paris, Leroux, 4 vols. Marx, K. [1964], Pre-capitalist Economic Formations (with an introduction by Eric Hobsbawm), London, Lawrence & Wishart. Marx, K., 1857–1858 [1969], Fondements de la Critique de l’économie politique, Paris, Anthropos. Marx, K., 1867, Capital, London, Lawrence & Wishart, 1970. Marx, K., 1861–1865 [1970], La guerre civile aux Etats-Unis, Paris, Union Gén. d’Ed. Mason, M., 1971, ‘Population Density and Slave-raiding. A Reply’, Jl of Afr. Hist., 12, 2: 324–7. Mason, M., 1973, ‘Captive and Client Labour in the Economy of the Bida Emirate’, Jl of Afr. Hist., 14, 3: 453–71. Maugham, R., 1961, Les esclaves existent toujours, Paris, Ed. Universitaires. Mauny, R., 1961, Tableau gégraphique de l’Ouest africain au Moyen Age, Mém. IFAN, Dakar. Mauny, R., 1967, ‘Le livre de bord du navire Santa-Maria da Comeiça (1522)’, Bull. IFAN, B. 29, 34: 512–35. Maurin, J., 1975, ‘Remarques sur la notion de puer à l’époque classique’, Bull, de l’Ass. Guillaume-Budé, 4: 221–30. Mazillier, 1894, ‘Rapport sur la captivité dans les cercles du Soudan’, Archives nationales du Sénégal, K14. Mbodj, M., 1978, Un exemple d’économie coloniale, le Sine-Saloum de 1877 à 1940, thesis, Paris VIII. M’Bow, A. M., Ki-Zerbo, J. and Devisse, J., 1965, La traite négrière du XVIIIe au début du XIXe siècle, Paris, Hatier. Meillassoux, C., 1964, Anthropologie économique des Gouro de Côte-d’Ivoire, Paris, Mouton. Meillassoux, C., 1966, ‘Plans d’anciennes fortifications (tata) en pays malinke’, Jl Soc. Afr., 36, 1: 29–43. Meillassoux, C., 1968, ‘Ostentation, destruction, reproduction’, Economies et Sociétés, II, 4: 760–72. Meillassoux, C. (ed.), 1971a, The Development of Indigenous Trade and Markets in West Africa, Oxford, OUP. Meillassoux, C., 1971d, ‘Le commerce, pré-colonial et le développement de l’esclavage à Gumbu du Sahel (Mali)’, in Meillassoux (ed.), 1971a: 182–95. Meillassoux, C., 1971e, ‘Review of Technology, Tradition and the State in Africa by J. Goody’, Africa, 41, 4: 331–3. Meillassoux, C., 1972b, ‘L’itinéraire d’Ibn Battuta de Walata à Malli’, Jl of Afr. Hist, 13, 3: 389–96. Meillassoux, C., 1972d, ‘Les origines de Gumbu du Sahel’, Bull. IFAN, B, 34, 2: 268–98.

174

Meillassoux

Meillassoux, C., 1972f, ‘Réexamen de l’itinéraire d’Ibn Battuta entre Walata et Malli’, Conference on Manding Studies, London. Meillassoux, C., 1972g, ‘Où donc était Suleyman? (Réponse à J. O. Hunwick sur l’emplacement de la capitale du Mali)’, Conference on Manding Studies, London. Meillassoux, C., 1975a, ‘Etat et condition des esclaves à Gumbu (Mali) au XIXe siècle’, in Meillassoux (ed.), 1975b: 221–52. Meillassoux, C. (ed.), 1975b, L’esclavage en Afrique précoloniale, Paris, Maspero. Meillassoux, C., 1978a, ‘L’interprétation légendaire de l’histoire du jonkoloni (Mali)’, in Bernardi et al. (eds), 1978: 347–92. Meillassoux, C. 1978d, ‘Correspondence (On Slavery)’, Economy and Society, 7, 3: 322–31. Meillassoux, C., 1979d, ‘Historical Modalities of the Exploitation and Overexploitation of Labour’, Critique of Anthropology, 4, 13/14: 9–16. Meillassoux, C., 1979e, ‘Le mâle en gésine’, Cah. Et. afr., 19, 1–4 (73–76): 353–80. Meillassoux, C., 1981, Maidens, Meal and Money: Capitalism and the Domestic Economy, Cambridge, CUP. Meillassoux, C., 1983, ‘The Economic Basis of Demographic Reproduction’, Journal of Peasant Studies, 11, 1: 50–61. Meillassoux, C., 1984, ‘Habitat e modi di vida’, in Spini, A. and S. (eds), 1984: 5–9. Meillassoux, C. and Niaré, A.-C., 1963, ‘Histoire et institutions du Kafo de Bamako’, Cah. Et. africaines, 4, 2 (14): 186–226. Meillassoux, C., Doucoure, L. and Simagha, D., 1967, Légende de la dispersion des Kusa (Epopée Soninké), Dakar, IFAN. Meillassoux, C. and Silla, A., 1978, ‘L’interprétation légendaire de l’histoire du Jonkoloni’, in Bernardi, B. et al. (eds), 1978 347–92. Mercadier, F.-J.-G., 1971, L’esclave de Timimoun, Paris, Ed. France-Empire. Mettas, J., 1978, Répertoire des expéditions négrières françaises au XVIII e siècle, Paris, Soc. franç. d’Hist. d’oM. Meunier, D., 1980, ‘Le commerce du sel de Taoudeni’, Jl Soc. des Afr., 50, 2: 133–44. Meyers, A., 1971, ‘Slavery in Hausa-Fulani Emirates’, in McCall, D. F. and Bennett, N. (eds), 1971. Michelet, J., 1861, La sorcière, Paris, Bibliothèque Mondiale. Miers, S. and Kopytoff, I. (eds), 1977, Slavery in Africa, Madison, Un. of Wisconsin Press. Miller, J. C., 1977, ‘Imbangala Lineage Slavery (Angola)’, in Miers and Kopytoff, 1977: 205–31. Miller, J. C., 1977, Slavery: A Teaching Bibliography, Brandeis, Crossroad Press and suppl. in Slavery and Abolition, 1980, 1, 1: 65–110. Mintz, S. (ed.), 1981, Esclave = Facteur de production, Paris, Dunod. Molin, Mgr, 1959, Recueil de proverbes bambaras et malinkes, Bamako, Les Presses Missionnaires.

The Anthropology of Slavery

175

Mollien, Th.-G. [1967], L’Afrique occidentale en 1818, Paris, Calmann-Lévy. Monier, R., 1947, Manuel élémentaire de droit romain, Paris, Domat-Mont-chrestien. Monod, T., 1975 (ed.), Les sociétes pastorales en Afrique tropicale, London, OUP. Monteil, C., 1902–3, Carnets, unpub. documents, Arch. V. Monteil. Monteil, C., 1915, Les Khassonké, Paris, E. Leroux. Monteil, C., 1924 [1977], Les Bambara de Ségou et du Kaarta, Paris, Larose. Monteil, C., 1927, Le coton chez les Noirs, Com. Et. Hist. Sci., AOF, Paris, Larose. Monteil, C., 1929, Les empires du Mali, Paris, Maisonneuve & Larose. Monteil, C., 1932 [1971], Une cité soudanaise, Djenné, Paris, Anthropos. Monteil, C., 1966, ‘Fin de siècle a Médine (1898–1899)’, B. IFAN, B, 28, 1–2: 82–172. Monteil, V., 1963, ‘Contribution à la sociologie des Peuls: “Fonds Vieillard” de I’IFAN’, Bull. IFAN, B, 25, 3–4: 351–414. Monteil, V., 1964, L’Islam noir, Paris, Seuil. Monteil, V., 1967, ‘The Wolof Kingdom of Kayor’, in Forde and Kaberry (eds), 1967: 260–82. Monteil, V., 1968a, ‘Un cas d’économie ostentatoire: les griots d’Afrique noire’, Economies et Sociétés, 2, 4: 772–91. Monteil, V., 1968b, ‘Introduction aux voyages d’Ibn Battuta (1325–1353)’, Bull, IFAN, B, 30, 2: 444–62. Montrat, M., 1935, ‘Notes sur les Malinke du Sankaran’, Outre-mer, 7: 107–27. Moraes, N.-I. de, 1978, ‘La campagne négrière du San-Antonio-c-as-Almas (1670)’, Bull, IFAN, B, 40, 4: 708–17. Moraes Farias, P. F. de, 1974, ‘Great States Revisited’, Jl of Afr. Hist., 15, 3: 479–88. Moraes Farias, P. F. de, 1980, ‘Model of the World and Categorial Models: The “Enslavable Barbarians” as a Mobile Classificatory Label’, Slavery and Abolition, 1, 2; 115–31. Mörner, M., 1980, Buy or Breed, Bucharest, Int. Congress of Historical Sciences. Morton-Williams, P., 1967, ‘The Yoruba Kingdom of Oyo’, in Forde and Kaberry (eds), 1967: 36–69. Mousnier, J., 1957, Journal de la traite des Noirs, Paris, Ed. de Paris. Mousnier, R., 1969, Les hiérarchies sociales de 1405 à nos jours, Paris, PUF. Munson, P., 1972, ‘Archeology and the Prehistoric Origins of the Ghana Empire’, Conf. on Manding Studies, London, SOAS, 1972. Nadel, S. F., 1942, Black Byzantium, London, OUP. N’Dyaye, B., 1970, Les Castes au Mali, Bamako, Editions Populaires. N’Diaye, M, (ed.), 1978, ‘Histoire de Ségou par Cheik Moussa Kamara’, Bull. IFAN, 40, 3: 458–88. Newbury, C. W., 1960, ‘Une ancienne enquête sur l’esclavage et la captivité au Dahomey’, Zaïre, 14, 1: 53–67. Niane, D. T., 1960, Soundjata ou l’épopée mandingue, Paris, Présence Africaine.

176

Meillassoux

Niane, D. T., 1974, ‘Histoire et tradition historique du Manding’, Présence africaine, 89, 1: 59–74. Nicolas, F. J., 1977, ‘L’origine et la signification du mot HARTANI et de ses équivalents’, Notes africaines, 156: 101–6. Nieboer, H. J., 1900, Slavery as an Industrial System, The Hague, Martinus Nijhoff. Northrup, D., 1976, ‘The compatibility of the Slave and Palm Oil Trades in the Bight of Biafra’, Jl of Afr. Hist., 17, 3: 353–64. Northrup, D., 1981, ‘Slavery in a Yoruba Society in the 19th Century’, in Lovejoy (ed.), 1981: 101–22. Nwachukwu-Ogedengbe, 1977, ‘Slavery in Nineteenth Century Aboh (Nigeria)’, in Miers and Kopytoff (eds), 1977: 133–52. Nyendal, D. Van, 1705, ‘Lettre’, in Bosman, 1705: 458. Obichere, B. I., 1978, ‘Women and Slavery in the Kingdom of Dahomey’, Rev. franç. d’Hist. d’OM, 65, 238: 5–20. OTahey, R. S., 1973, ‘Slavery and the Slave-Trade in Dār-Fūr’, Jl of Afr. Hist., 14, 1: 29–43. Olivier de Sardan, J.-P., 1969a, Les voleurs d’hommes, Niamey, ‘Etudes nigériennes’, 25. Olivier de Sardan, J.-P., 1969b, Systèmes des relations économiques et sociales chez les Wogo du Niger, Paris, Inst. d’Ethno. Olivier de Sardan, J.-P. 1973, ‘Personnalité et structures sociales (le cas sonxai)’, in Colloques internationaux du CNRS, La notion de personne en Afrique noire: 421–46. Olivier de Sardan, J.-P., 1973, ‘Esclavage d’échange et captivité familiale chez les Songhay-Zerma’, Jl Soc. des Afr., 43, 1: 111–51. Olivier de Sardan, J.-P., 1975, ‘Captifs ruraux et esclaves impériaux du Songhay’, in Meillassoux (ed.), 1975: 99–134. Olivier de Sardan, J.-P., 1976, Quand nos pères étaient captifs, Paris, Nubia. Olivier de Sardan, J.-P., 1982, Concepts et conceptions Songhay-Zarma, Paris, Nubia. Olivier de Sardan, J.-P., 1984, Les sociétés Songhay-Zarma (Niger-Mali), Paris, Karthala. Ortoli, H., 1939, ‘Le gage des personnes au Soudan français’, Bull. IFAN, B, 1, 1: 313–24. Panoff, M., 1970, ‘Du suicide comme moyen de gouvernement’, Les Temps modernes, 27, 288: 109–30. Panoff, M., 1979, ‘Travailleurs, recruteurs et planteurs dans l’archipel Bismarck de 1885 à 1914’, Jl de la Soc. des Océanistes, 35, 64: 159–73. Park, M. [1960], Travels (R. Miller, ed.), London, Dent & Sons. Pasquier, R., 1967, ‘A propos de l’émancipation des esclaves au Sénégal en 1848’, Rev. franç. d’Hist. d’OM, 54, 194–7; 188–208. Patterson, O., 1979, ‘On Slavery and Slave Formations’, MLR, II, 7: 31–67. Paulme, D., 1940, L’organisation sociale des Dogon, Paris, Domat-Montchrestien. Paulme, D. (ed.), 1960, Femmes d’Afrique noire, Paris, Mouton. Peristiany, J. G. (ed.), 1965, Honour and Shame, London, Weidenfeld & Nicolson.

The Anthropology of Slavery

177

Peroz, E., Cap., 1896, Au Soudan français, Paris, Calmann-Lévy. Perrot, C.-H., 1975, ‘Les captifs dans le royaume Anyi du Ndenye’, in Meillassoux, C. (ed.), 1975b: 351–88. Perrot, C. H., 1983, Les Anyi-Indenye et le pouvoir aux XVIII e et XIX e siècles, Abidjan, CEDA. Person, Y., 1968–75, Samori, une révolution dyula, Dakar, IFAN, Mém. no. 80, 3 vols. Peytraud, L., 1897 [1973], L’esclavage aux Antilles françaises avant 1789, Pointe-a-Pitre, Desormeaux. Piault, M., 1975, ‘Captifs du pouvoir et pouvoir des captifs’, in Meillassoux (ed.), 1975b 321–50. Piault, M., 1982, ‘Le héros et son destin’, Cah. Et. afr, 87–88, 22, 3–4; 403–40. Pietri, Cap., 1885, Les Français au Niger, Paris, Hachette. Pipes, D., 1981, Slave Soldiers and Islam. The Genesis of a Military System, London, CUP. Polanyi, K., 1964, ‘Sortings and “Ounce Trade” in the West African Slave Trade’, Jl of Afr. Hist., 5, 3: 381–94. Polanyi, K., 1965, Primitive, Archaic and Modern Economics, New York, Doubleday, Anchor. Polanyi, K., 1966, Dahomey and the Slave-Trade, Seattle, Un. of Washington Press. Pollet, E. and Winter, G., 1971, La société Soninke, Brussels, Inst. de Sociologie, Ed. de l’Un. de Bruxelles. Pontie, G., 1973, Les Guiziga du Cameroun septentrional: l’organisation traditionnelle et sa mise en contestation, Paris, Mém. ORSTOM, no. 65. Postma, J., 1972, ‘The Dimension of the Dutch Slave Trade from Western Africa’, Jl Afr. Hist., 13, 2: 237–48. Pouillon, J. and Maranda, P. (eds), 1970, Echanges et communications, Paris, Mouton. Pruneau de Pommegorge, 1789, Description de la Nigritie, Amsterdam. Quesne, J.-S., 1819, Histoire de l’esclavage en Afrique de P.-J. Dumont, Paris, Pillet Aîné. Quimby, L. G., 1975, ‘History as Identity: The Jaaxanke and the Founding of Tuuba’, Bull. IFAN, B, 37: 619–45. Quinn, C. A., 1972, Mandingo Kingdoms of the Senegambia, Evanston, Northwestern Un. Press. Raffenel, A., 1846, Voyage dans l’Afrique occidentale (1843–1844), Paris, Arthus Bertrand. Randies, W. G. L., 1968, L’ancien Royaume du Congo des origines à la fin du XIXe siècle, Paris, Mouton. Randles, W. G. L., 1975, L’Empire du Monomotapa du XV e au XIXe siècle, Paris, Mouton. Rattray, R. S., 1923, Ashanti, Oxford, Clarendon Press. Rattray, R. S., 1929 [1969], Ashanti Law and Constitution, Oxford, Clarendon Press. Reindorf, C. P., 1895, The History of Gold Coast and Asante, Basle. Renaud, R. (ed.), 1972, De l’ethnocide, Paris, UGE, ‘10 × 18’.

178

Meillassoux

Renault, F., 1971, Lavigerie, l’esclavage africain et l’Europe, 1868–1892, Paris, E. de Boccard, 2 vols. Renault, F., 1972, L’abolition de l’esclavage au Sénégal, Paris, Soc. franç. d’Hist. d’OM. Renault, F., 1976, Libération d’esclaves et nouvelle servitude, Dakar, NEA. Retel-Laurentin, A., 1974, Infécondité en Afrique noire, Paris, Masson & Cie. Retel-Laurentin, A., 1979, Causes de l’infécondité dans la Volta noire, Paris, PUF, ‘INED, Trav. et Doc.’, no. 87. Rey, P.-P., 1971, Colonialisme, néo-colonialisme et transition au capitalisme, Paris, Maspero. Rey, P.-P., 1975, ‘L’esclavage lignager chez les Tsangui, les Punu et les Kuni du CongoBrazzaville’, in Meillassoux (ed.), 19756: 509–28. Rey, P.-P., 1976, Capitalisme négrier, Paris, Maspero. Rey, P.-P. (n.d.), Les concepts de l’anthropologie économique marxiste, thesis, Doct. Etat, Univ. R. Descartes. Rey-Hulman, D., 1975, ‘Les dépendants des maîtres tyokossi pendant la période coloniale’, in Meillassoux (ed.), 1975b: 297–320. Richard-Molard, J., 1948–9, ‘Démographie et structure des sociétés négopeul parmi les hommes libres et les “serfs” du Fouta Dialon (région de Labé)’, Rev. Géog. humaine et Ethnol, 1, 4: 45–51. Richards, W. A., 1980, ‘The Import of Firearms into West Africa in the 18th Century’, Jl of Aft. Hist., 21, 1: 43–60. Richardson, J., 1854, Narrative of a Mission to Central Africa, London, 2 vols. Riesman, P., 1974, Société et liberté chez les Peul Djelgôbé de Haute-Volta, Paris, Mouton. Rizvi, M. S. S. A., 1967, ‘ “Zenj”: its First Known Use in Arabic Literature’, Azania, 2: 200–1. Roberts, R., n.d., Production and Reproduction in Warrior States: Segu Bambara and Segu Tukulor, 22 p. mult. Roberts, R., 1978, The Maraka and the Economy of the Middle Niger Valley, Ph. D. Dept. of History, Un. of Toronto. Roberts, R., 1980a, Multiplier Effect in the Ecologically Specialized Trade of Pre-colonial West Africa, Centre for African Studies, Dalhousie Univ. Roberts, R., 19806, ‘Long-distance Trades and Production: Sinsanni in the 19th Cent.’, Jl of Afr. Hist., 21, 2: 169–88. Roberts, R., 1980c, The Emergence of a Grain Market in Bamako, 1883–1908’, Rev. canadienne des Et. afr. 14, 1: 55–81. Roberts, R., 1984, Warriors and Merchants, Dept of History, Stanford Un. Roberts, R. and Klein, M. A., 1980, ‘The Banamba Slave Exodus of 1905 and the Decline of Slavery in Western Sudan’, Jl of African Hist., 21, 3: 375–94. Robertson, C. C., 1983, ‘Post-Proclamation Slavery in Accra: A Female Affair?’, in Robertson and Klein (eds), 1983: 220–42.

The Anthropology of Slavery

179

Robertson, C. C. and Klein, M. A. (eds), 1983, Women and Slavery in Africa, Madison, Un. of Wisconsin Press. Robertson, G. A. (1819), Notes on Africa, London, Sherwood, Neely & Joves. Roche, D., 1973, Ordres et classes, Paris, Mouton. Rodinson, M., 1966, Islam et capitalisme, Paris, Seuil. Rodney, W., 1966, ‘African Slavery and Other Forms of Social Oppression on Upper Guinea Coast in the Contact of the Atlantic Slave Trade’, Jl Afr. Hist., 7: 431–43. Rodney, W., 1969, ‘Gold and Slaves on the Gold Coast’, Trans. Hist. Soc. Ghana, 10: 13–38. Ronen, D., 1971, ‘On the African Role in the Transatlantic Slave Trade in Dahomey’, Cah. Et. afr., 11, 1 (41): 5–13. Rouch, J., 1953, Contribution à l’histoire des Songhay, Dakar, IFAN. Roussier, P., 1935, L’établissement d’Issinie (1628–1702), Paris. Ruyle, E. E., 1973, ‘Slavery, Surplus, and Stratification of the Nothern Coast’, Cur. Anthropology, 14, 5, Dec. 1973: 603–32. Ryan, T. C. I., 1975, ‘The Economics of Human Sacrifice’, Afr. Econ. Hist. Rev., 2, 2, Fall 1975: 1–9. Ryders, A. F. C., 1969, Benin and the Europeans, 1485–1897, London. Saint-Père, J.-H., 1925, Les Sarakollé du Guidimakha, Paris, Larose. Salifou, A., 1971, Le Damagaram ou le sultanat de Zinder au XIXe siècle, Niamey, ‘Etudes nigériennes’, no. 27. Samb, A., 1980, ‘L’Islam et l’esclavage’, Notes africaines, Dakar, 168: 93–7. Samuel, M., 1977, Le prolétariat africain noir en France, Paris, Maspero. Sanneh, L. O., 1976a, ‘Slavery, Islam and the Jakhanke People of West Africa’, Africa, 46, 1: 80–97. Sanneh, L. O., 1976b, The Origin of Clericalism in West African Islam’, Jl of Afr. Hist., XVII, I: 49–72. Saraiva, A. J., 1967, ‘Le P. Antonio Vieira, s.j., et la question de l’esclavage des Noirs au XVIIe siècle’, Annales, Et. Soc. Civ., 22, Nov.–Dec: 1289–309. Sauvageot, S., 1965, Contribution à l’histoire du royaume bambara de Ségou—XVIIIe et XIXe siècles, Paris, Fac. des Lettres, Sorbonne. Schnakenbourg, C., 1980, La crise du système esclavagiste 1835–1847, Paris, L’Harmattan. Schoelcher, V., 1880, L’esclavage au Sénégal en 1880, Paris, Librairie Centrale des Publications Populaires. Schoelcher, V., 1948, Esclavage et colonisation, Paris, PUF. Servet, J.-M., 1981, Genèse des formes et pratiques monétaires, thesis, Doct. d’Etat et Sc. éco., Université Lyon II. Seydou, C., 1972, Sillâmaka et Poullôri, Paris, A. Colin. Shui Hu, 1974, Les 108 brigands du Liang Shan, Paris, Signes, 2 vols. Sidibe, M., 1935, ‘Tableau de la vie indigène au Soudan et dans la boucle du Niger (du XVIIIe à nos jours)’, L’Educ. afric, 24, 89: 3–26.

180

Meillassoux

Silberbauer, G. B. and Kuper, A. J., 1966, ‘Kalagari Masters and Bushmen Serfs: Some Observations’, Afr. Stud., 25, 4: 171–9. Siran, J.-L., 1980, ‘Emergence et dissolution des principautés guerrières Vouté (Cameroun)’, Jl Soc. des Afric, 50, 1: 25–58. Sissoko, K., 1975, ‘La prise de Dionkoloni’, in Dumestre and Kesteloot (eds), 1975. Skertchly, J. A., 1874, Dahomey as it is, London. Skinner, E. P., 1964, The Mossi of the Upper Volta, Berkeley, Stanford Un. Press. Smaldone, J. P., 1972, ‘Firearms in the Central Sudan: A Revolution’, Jl of African History, 13: 591–607. Smith, M. G., 1954, ‘Slavery and Emancipation in Two Societies’, Social and Econ. Stud, 3, 3–4: 239–90. Smith, M. G., 1967, ‘A Hausa Kingdom: Maradi under Dan Bashore, 1854–1875’, in Forde and Kaberry, 1967: 93–122. Smith, R., 1967, ‘Yoruba Armaments’, Jl Afr. Hist, 8: 96–8. Snelgrave, W., 1734, A New Account of Some Parts of Guinea and the Slave Trade, London, James, John and Paul Knapton. Sow, A. I. (ed.), 1971, Dictionnaire élémentaire fulfulde-français, Niamey, CRDTO. Spini, A. and S. (eds), 1984, ‘I Bozo del Niger: Insediaments e Architettura’, Storia della Citta, 25, 132 p. Stavenhagen, R., 1969, Les classes sociales dans les sociétés agraires, Paris Anthropos. Stein, R., 1978, ‘Measuring the French Slave Trade’, Jl of African History, XIX, 4: 515–22. Strobel, M., 1983, ‘Slavery and Reproductive Labor in Mombasa’, in Robertson and Klein, 1983: 111–29. Suret-Canale, J., 1964a, ‘Contexte et conséquences sociales de la traite africaine’, Présence africaine, no. 50: 142–3. Suret-Canale, J., 19646, L’Afrique noire, l’ère coloniale 1900–1945, Paris, Ed. Sociales. Suret-Canale, J., 1969, ‘Les origines ethniques des anciens captifs du Fouta Djalon’, Notes afr., 123: 91–2. Sy, M.-O., ‘Le Dahomey, gouvernement, administration’, Acta Ethnographica (Budapest), XII, 3–4: 333–65. Sy, M. O., 1965, ‘Le Dahomey, le coup d’Etat de 1818’, Folia Orientalia, VI: 205–38. Talhami, G. H., 1977, ‘The Zanj Rebellion Reconsidered’, Int. Jl of Afr. Hist. Studies, 10, 3: 443–61. Tandeter, E., 1981, ‘Forced and Free Labour in Late Colonial Potosi (Bolivia)’, Past and Present, 93: 98–136. Tardits, C., 1970, ‘Femmes à crédit’, in Pouillon and Maranda (eds), 1970: 382–90. Tardits, C., 1973, ‘Parenté et pouvoir politique chez les Bamoum’, L’Homme, XIII, 1–2: 37–49. Tardits, C., 1980a, Le royaume bamoum, Paris, A. Colin.

The Anthropology of Slavery

181

Tardits, C., 19806, ‘Roi divin’, Encyclopedia Universalis, Paris. Tautain, Dr L., 1884, ‘Notes sur les castes chez les Mandingues et chez les Bamanas’, Rev. d’Ethnographie, III: 343s. Tembera, T., 1978, ‘Biton Kulibali ka masala’, in Kesteloot (ed.), 1978. TEF: Tarikh el-Fettach [1964], Adrien Maisonneuve (see Kâti). Teixera da Motta, A., 1969, ‘Un document nouveau pour l’histoire des Peuls au Sénégal pendant les XV e et XVIe siècles’, Bol. Cult, da Guiné Portuguesa, 96. Teixera da Mota, A. and Mauny, R., 1978, ‘Livre de l’armement du navire Sāo Miguel de l’île de São Tomé au Bénin (1522)’, Bull. IFAN, 40, B, 1, 1978: 68–86. TEN (Tedzkiret en-Nisiān) [1966] (transi. O. Houdas), Paris, A. Maisonneuve. TES: Tarikh es-Sudan [1964], A. Maisonneuve (see Es-Sa’di). Terray, E., 1974, ‘Long-Distance Exchange and the Formation of the State’, Economy and Society, 3, 3: 315–45. Terray, E., 1975a, ‘Class and Class Consciousness in the Abron Kingdom of Gyaman’, in Bloch (ed.), 1975: 85–136. Terray, E., 1975b, ‘La captivité dans le royaume abron du Gyaman’, in Meillassoux (ed.), 1975b. Terray, E., 1976, ‘Contribution à une étude sur l’armée asante’, Cah. Et. afr., 61–62, XVI (1–2): 297–356. Terray, E., 1982a, ‘Nature et fonction de la guerre dans le monde akan’, in Bazin and Terray (eds), 1982: 375–422. Terray, E., 1982b, ‘Réflexions sur la formation du prix des esclaves à l’intérieur de l’Afrique de l’Ouest précoloniale’, Journal des Africanistes, 52, 1–2: 119–44. Terray, E., 1983, ‘Gold Production, Slave Labour and State Intervention in Pre-colonial Akan Societies’, Research in Economic Anthropology, 5: 95–129. Thapar, R., 1980, ‘State Formation in Early India’, Int. Soc. Sci. J., XXX, I, 4: 655–69. Thapar, R., 1984, From Lineage to State, Bombay, OUP. Thilmans, G., 1976, ‘La relation de François de Paris (1682–1683)’, B. IFAN, B, 38, 1, Jan. 1976: 1–51. Thilmans, G., Descamps, C. and Khayat, B., 1980, Protohistoire du Sénégal, I: Les sites mégalithiques, Dakar, IFAN, Mem. no. 91. Thorner, A., 1982, ‘Semi-Feudalism or Capitalism’, in ‘Caste et classe en Asie du Sud’, special issue, Purusārtha, 6: 19–72. Thornton, J., 1979, The Kingdom of Kongo in the Era of Civil Wars, 1641–1718, Ph.D. dissertation, Berkeley, Univ. of California. Tiendrebeogo, 1963, ‘Histoire traditionnelle des Mossi de Ouagadougou’, Jl Soc. Afric., 33, 1963: 11. Toledano, E. R., 1981, ‘Slave Dealers, Women, Pregnancy and Abortion’, Slavery and Abolition, 2, 1: 53–68.

182

Meillassoux

Traore Ray Autra, M., 1979, ‘Les unités de mesure dans l’ancienne société mandingue de Guinée’, Notes afr., 163: 63–5. Triaud, J.-L., 1981, ‘L’Islam en Afrique de l’Ouest’, Colloque de la Société pour l’avancement des études islamiques, Paris, 27–28 Mar. 1981. Trimingham, J. S., 1962, A History of Islam in West Africa, London, OUP. Triulzi, A., 1981, Sait, Gold and Legitimacy, Istituto Un. Orientale. Tuden, A., 1958, ‘Ila Slavery’, Hum. Prob. Brit. Cent. Afr., 24: 68–78. Tymowski, M., 1974a, ‘La ville et la campagne au Soudan occidental du XIVe au XVIe siècle’, Acta Poloniae Historica, 29: 51–79. Tymowski, M., 1974b, Le développement et la régression chez les peuples de la boucle du Niger à l’époque précoloniale, Warsaw, Wydawnictwa Universytetu Warszawskiego. Unesco (ed.), 1979, The African Slave-Trade from the 15th to the 19th Century, Paris, Unesco. Valensi, L., 1967, ‘Esclaves chrétiens et esclaves noirs à Tunis au XVIIIe siècle’, Annales (ESC), 22: 1267–88. Van Dantzig, A., 1975, ‘Effects of the Atlantic Slave on Some West African Societies’, Revue franç, de l’Hist. d’OM, 62, 226–7; 252–69. Vansina, J., 1961, ‘De la tradition orale’, Annales, no. 36, Tervuren. Vansina, J., 1975, ‘Kuba Chronology Revisited’, Paideuma, 21: 134–50. Vercoutter, J., 1976, ‘L’iconographie du Noir dans l’Egypte ancienne des origines à la XXV e dynastie’, in Bugner, L., 1976, I: 33–88. Verlinden, C., 1955, L’esclavage dans l’Europe médiévale, Bruges. Verlinden, Ch., 1966, ‘Esclavage noir en France méridionale et courants de traite en Afrique’, Annales du Midi, 78, 77–78; 335–43. Vidal-Naquet, P., 1965, ‘Economie et société dans la Grèce ancienne: l’œuvre de Moses I. Finley’, Archives européennes de Sociologie, 6: 111–48. Vidal-Naquet, P., 1967, ‘Les esclaves grecs étaient-ils une classe?’, Raison présente, 6: 103–12. Villamarin, J. A. and J. E., 1975, Indian Labor in Mainland Colonial Spanish America, Newark, Un. of Delaware, Latin America Studies Program, Mon. no. 1. Vissière, I. and J.-L., 1982, La traite des Noirs du siècle des Lumières, Paris, A.-M. Métailié. Walckenaer, C.-A., 1842, Collection des relations de voyages par mer et par terre en différentes parties de l’Afrique depuis 1400 jusque’ à nos jours, Paris, 21 vols. Wansbrough, J., 1970, ‘Africa and the Arab Geographs’, in Dalby, D. (ed.), 1970: 89–101. Watson, J. L. (ed.), 1980, Asian and African Systems of Slavery, Oxford, Blackwell. Weil, P. M., 1970, ‘Mandinka Fertility, Islam, and Integration in a Plural Society’, American Anthropological Ass., San Diego, 19–22 November. White, L. A., 1969, The Science of Culture, New York, Farrar, Straus & Giroux. Wilks, L, 1967, ‘Ashanti Government’, in Forde and Kaberry, 1967: 206–38.

The Anthropology of Slavery

183

Wilks, I., 1971, ‘Asante Policy towards the Hausa Trade in the 19th Century’, in Meillassoux (ed.), 1971: 124–44. Williams, E., 1944, Capitalism and Slavery, NY, Capricorn Books. Wismes, A. de, 1973, La vie quotidienne dans les ports aux XVIIe et XVIIIe siècles, Paris, Hachette. Wright, M., 1983, ‘Bwanikwa: Consciousness and Protest among Slave Women in Central Africa’, in Robertson and Klein, 1983: 246–67. Zeys, E., 1900, ‘Esclavage et guerre sainte. Consultation adressée aux gens du Touat par un érudit nègre, cadi de Timboctou au XVIIe siècle’, Bull. Réunion Et. algér., 2: 125–151; 166–89.

Slavery: A Question of Definition Suzanne Miers Slavery is arguably the most misused word in the English language. It has become a metaphor for extreme inequality, for subordination, deprivation and discrimination. It is bandied about in all manner of contexts. Thus we have the ‘classic’ or ‘chattel slave’, the Marxian ‘wage slave’, the ‘sex slave’ and, in the late twentieth century, the ‘contemporary slave’. Scholars, government officials, colonial civil servants, explorers, missionaries, nationalists, League of Nations and United Nations anti-slavery committees, Human Rights activists, and lawyers charged with drawing up national laws and international conventions, have all wrestled with the difficulty of giving a precise meaning to the term.

The Academic Debate

Scholars have disagreed over the attributes of slavery. Were slaves primarily property, and if so, how was property to be defined?1 Did slaves have to be saleable, or otherwise transferable? What rights did owners have over their slaves that they did not have over the free members of a kin group, family or community? Would it be more accurate to define slaves as persons under the complete control of, and utterly dependent upon a master, a mistress, a kin group or some other organization? If so, did that control have to be lifelong and/or hereditary, or could slavery be a temporary condition? Was the slave always acquired through an act of violence? If so, what constituted violence? Source: Miers, Suzanne, “Slavery: A Question of Definition,” Slavery & Abolition 24(2) (2003): 1–16. © Taylor & Francis Ltd, reprinted by permission of Taylor & Francis Ltd, www.tandfonline.com. 1  Only a few of the relevant works dealing with the subject can be cited here. See for instance Igor Kopytoff and Suzanne Miers, ‘African “Slavery” as an Institution of Marginality’, in Suzanne Miers and Igor Kopytoff (eds.), Slavery in Africa: Historical and Anthropological Perspectives (Madison: University of Wisconsin Press 1977), pp. 3–81; Igor Kopytoff, ‘Commentary One (on Paul Lovejoy)’ in Michael Craton (ed.), Roots and Branches: Current Directions in Slave Studies (Toronto: Pergamon Press, 1979), pp. 62–77; Igor Kopytoff, ‘Slavery: Introduction’, Annual Review of Anthropology, II (1982), pp. 207–30; Paul Lovejoy, Transformations in Slavery: A History of Slavery in Africa (Cambridge: Cambridge University Press, 1983); Claude Meillassoux, The Anthropology of Slavery (Chicago: University of Chicago Press, 1991); Orlando Patterson. Slavery and Social Death: a Comparative Study (Cambridge, Mass. and London: 1982). © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_008

Slavery: A Question Of Definition

185

If people were sold against their will was this sufficient by itself to define them as slaves? Were slaves necessarily ‘socially dead’—that is non-persons without social existence or status except by virtue of their owners—or could they be recognized as people with rights of their own? There have been debates on the purpose of slavery. Was it primarily a way of mobilizing labour? If so, was it cost effective? Was it a step towards the incorporation of ‘outsiders’ and their descendants, as full members of a society? Alternatively, was it a method of incorporating outsiders into a particular social formation while excluding them permanently from full membership of a kinship or other group? Was it a device for elites to acquire and retain power and prestige? The questions are endless and the approaches are not mutually exclusive. Slavery as the term is commonly used can fulfil all these purposes, as the chapters here make very clear. No definition of slavery can be separated from the definition of its antithesis—freedom. This is as difficult to define as slavery, since freedom has meant different things to different peoples and even to the same people at different times in their history. The main question is whether different societies have a different concept of freedom or whether there are certain rights without which people in any society would not consider themselves to be free. In the western world at the outset of the twenty-first century a free person is generally considered to be an autonomous individual—someone with the economic and social rights to move freely, to choose his (or her)2 occupation, to keep the proceeds of his labour, to determine his own lifestyle, to choose his spouse, to control his own children and so forth. Free persons are protected from physical abuse and have a range of political rights, which include freedom of speech, of religion, of association, and freedom from arbitrary arrest, embodied in the concept of the ‘rule of law’. Communist writers and statesmen have argued that political rights are secondary to the social and economic rights to sustenance, employment and equal opportunity. Asian statesmen and intellectuals have challenged some of the western human rights agenda on the grounds that primacy should be given to social and political stability and the well-being of the community.3 Complete freedom obviously exists only in theory. In practice free people, even when freedom is defined in the western sense, do not have complete

2  I use the masculine term henceforth only to avoid the more cumbersome his/her, but most slaves were probably female. 3  See Orlando Patterson, Freedom in the Making of Western Culture (New York: Basic Books 1991); David Kelly and Anthony Reid (eds.), Asian Freedoms: The Idea of Freedom in East and Southeast Asia (Cambridge: Cambridge University Press, 1998).

186

Miers

control over their lives, nor do they have equal social standing or material wealth. Their choices are limited by economic, political and social constraints. In certain societies, they are also limited by the demands of religion and caste, and/or the demands of family or kin group, which in some cases had the power to sell or transfer their own members, just as fathers could sell their children in ancient Rome and in pre-communist China. In many societies, also, women do not have equal rights with men. Parents or guardians everywhere have certain rights over their children. Yet in spite of these constraints women, children, and other full members of families, kin groups and societies consider themselves to be free. The term, therefore, may be taken to designate the norm in any society, and hence is culturally defined, but within certain limits. Thus, while different societies emphasize different essential features of freedom depending on their history, economic development, politics and culture, some forms of curtailment of personal liberties are universally considered to be forms of slavery. Thus chattel slavery is usually seen as the antithesis of freedom. A chattel slave is normally defined as someone under the complete domination of an owner who has powers of life and death over him or her,4 can sell and transfer him at will and has full control over his daily and domestic life including his progeny. Moreover, his status is hereditary. However, as will be seen from the studies discussed below, chattel slavery took many forms, and was only one form of servitude among many.

Slavery and Unfree Labour in the Indian Ocean World

With some notable exceptions,5 most discussion on the nature of slavery and freedom has been in the context of the New World and Africa. The essays in this collection provide a rich source of information on the diversity of both slavery and freedom in the world of the Indian Ocean. They cover a huge geographical area, stretching from South Africa to East Asia and the Philippines. Since the authors use the term slavery to cover a great variety of institutions, they add to the complexity and difficulty of finding any universally acceptable definition of either term. 4  In some societies the owner had the right to punish the slave but not actually to kill him. 5  See for instance, James Watson (ed.), Asian and African Systems of Slavery (Oxford: Basil Blackwell, 1980); Anthony Reid (ed.), Slavery, Dependency and Bondage in Southeast Asia (New York: St Martin’s Press, 1983); Martin A. Klein (ed.), Breaking the Chains: Slavery, Bondage, and Emancipation in Modern Africa and Asia (Madison: University of Wisconsin Press, 1993).

Slavery: A Question Of Definition

187

A number of essays deal with slavery as practised by Europeans. Called by Edward Alpers the ‘Christian European model’, this varied from the plantation system of the Mascarenes and Seychelles discussed by Musleem Jumeer6 to the Cape Colony model described by Nigel Worden.7 In both cases, as in the Americas and the Caribbean, the slave was a chattel, with no rights, who existed only as the property of his or her master.8 In Mauritius, slavery, together with the apprenticeship system which succeeded it (which tied freed slaves to their owners for a term of years after their emancipation), and the indentured labour and other later forms of labour on the island were all part of a single ‘core system of labour exploitation’ for economic purposes.9 In theory there were important differences between slaves and indentured labourers. Theoretically the latter came voluntarily, although some may, as Richard Allen suggests, have been ‘praedial slaves’, or other unfree persons in southern India who had no choice in the matter. Once in Mauritius, however, they were all confined to the plantations, poorly fed, housed and clothed, and worked just as the slaves had been.10 Keya Dasgupta shows that indentured labourers on tea plantations in the Brahmapura Valley in Assam were also subjected to much the same work regime as the slaves on the plantations of the New World.11 Moreover, like slaves in the Americas, in both Mauritius and Assam, indentured labourers were uprooted from their homelands, settled in regions where they had no links with the local people, and kept isolated on the plantations. However, their condition differed from that of chattel slaves in that their families often accompanied them and laboured with them. They were also paid wages, although at a minimal rate. Most importantly when their contracts ended they were free to leave the plantations. Thus, although planters often found ways to keep them, they were legally free. Their servitude was thus temporary. The question for us is whether or not this was ‘a new system of slavery’ as has been suggested in relation to the Caribbean, or, as contended by recent researchers, this temporary servitude enabled labourers to escape the 6  See article by Alpers in this collection; Musleem Jumeer, ‘Slave Systems in Mauritius: Structure and Change, 18th–20th centuries’, paper presented at the SSAI (2000). 7  Nigel Worden, ‘Indian Ocean Slavery and its Demise in the Cape Colony’, in Gwyn Campbell (ed.), Abolition and Its Aftermath in Indian Ocean Africa and Asia (London and Portland, OR: Frank Cass, 2004), ch.2. 8  I use the word master for convenience although some women owned slaves. 9  To quote Jumeer. 10  Richard Allen’s article in this collection. 11  Keya Dasgupta, ‘Plantation Labour in the Brahmapura Valley: Regional Enclaves in a Colonial Context’, in Campbell (ed.), Abolition and Its Aftermath.

188

Miers

poverty of their homelands and raise their standards of living.12 Thus although some may have been acquired as slaves and for a while treated as slaves, the final outcome was an escape from slavery. In the Cape Colony, many slaves were bought and imported from all over the huge Indian Ocean catchment area, while other persons were enslaved by a different process. Neighbouring Khoi pastoralists were gradually reduced to servitude. After slavery lost its legal status, African children were captured in commando raids and theoretically indentured for a period of years. In practice, they rarely regained their independence. Kerry Ward shows that even Africans taken from captured slave ships by the British and landed at the Cape, although ostensibly free, actually became effectively slaves.13 All these victims, acquired in different ways were melded into what became a complex system of unfree labour, often called ‘virtual slavery’—a term used as loosely as slavery itself, when no more precise definition can be found, for a situation in which the victim has most but not all the attributes of a slave. Several chapters deal with Islamic slavery.14 These cover a wide geographical span from East Africa to Indonesia. Everywhere practice varied with local circumstances and variations in Muslim law. In Muslim slavery, as in the western world, slaves were chattels. Most were initially acquired by force. Some were bought. Others were victims of trickery. All were freely traded and transferred. What was notably different from the slavery of the western world, however, was the degree to which they were protected by Muslim law.15 When the 12  I am not discussing here the forced recruitment of Africans for the Indian Ocean plantations, which was certainly a disguised form of slave-trading. For two different views on the traffic in indentured labour see Hugh Tinker, A New System of Slavery: the Export of Indian Labour Oversees 1830–1920 (London: Hansib Publishing, 1993); Pieter Emmer, ‘“A Spirit of Independence” or Lack of Education for the Market? Freedmen and Asian Indentured Labourers in the Post Emancipation Caribbean, 1834–1917’, in Howard Temperley (ed.), After Slavery: Emancipation and its Discontents (London: Frank Cass 2000). 13  Kerry Ward, ‘Changing Patterns of Forced Migration at the Cape after the Collapse of the Dutch East India Company’, paper presented at the SSAI (2000). 14  See Omar Eno, ‘The Abolition of Slavery and the Aftermath Stigma: the Case of the Bantu/ Jareer People on the Benadir Coast of Southern Somalia’; Suzanne Miers, ‘Slavery and the Slave Trade in Saudi Arabia and the Arab States on the Persian Gulf 1921–1963’; Behnaz Mirzai, ‘The 1848 Abolitionist Farmân: A Step Towards Ending the Slave Trade in Iran’; Abdul Sheriff, ‘The Slave Trade and Its Fallout in the Persian Gulf; and W. G. ClarenceSmith, ‘Islam and the Abolition of the Slave-Trade and Slavery in the Indian Ocean’—all in Campbell (ed.), Abolition and Its Aftermath. 15  For a discussion of this see Bernard Lewis, Race and Slavery in the Middle East: An Historical Enquiry (New York and Oxford: Oxford University Press 1990).

Slavery: A Question Of Definition

189

law was observed, their treatment was good. They might expect to marry and have families of their own, and they had a good chance of being freed. There were also built in avenues of escape. For instance, a female slave who married her master had to be freed first. Concubines—slaves by definition—were freed or, at least were not saleable, once they had borne their masters’ children; and in most branches of Islam, their children were free. As in the West, slaves could be manumitted by their masters, or could ransom themselves, but in Islam manumission was a meritorious act, earning credit in the next world and was more widely practised. Manumitted slaves, however, were not free in the Western sense. They and their descendants became clients in perpetuity of their former owners and their progeny. Clients had to have their patrons’ permission to marry and suffered other restrictions on their liberty. It is clear that not all owners obeyed Muslim law. The actual well-being of slaves in the Islamic world, as elsewhere, depended largely on the position and the disposition of their owners. Many slaves, as Alpers shows, fled from the plantations on the east coast of Africa, preferring to take the risks of marronage to the cruelties of slavery. The hardships of slave life on these plantations are commented on by both Omar Eno and by Mohamed Kassim, writing on the Benadir coast and hinterland.16 Abdul Sheriff writing on the Persian Gulf region, and Suzanne Miers writing on Saudi Arabia and the Gulf States, note that a number of slaves took refuge with British officials or on British warships, risking recapture and severe punishment, in order to obtain manumission. Nevertheless the rules were adhered to in general to the point that there was a steady attrition in the numbers of slaves in the Muslim world, hence the demand remained high.17 As has often been noted, the chapters on Muslim slavery show that slaves could rise to high office and have power over the free. They were valued just because, without kin ties, if they were well treated, their loyalty to their owners could be counted on. Moreover, they often had considerable personal freedom. James Warren cites examples in Sulu where they were sent on trading expeditions and even on armed raiding parties.18 Some became richer than their owners through trade and other means. They even acquired slaves of their own, and could keep their wealth for their lifetime. In the Muslim world, slaves were not necessarily the lowest rank in society. Michel Boivin, writing 16  Eno, ‘The Abolition of Slavery’; Mohamed Kassim, ‘The Maroon Settlements of Southern Somalia’, paper presented at the SSAI (2000). 17  Miers, ‘Slavery and the Slave Trade’; Sheriff, ‘The Slave-Trade’—which see for a discussion of slave reproduction in the Persian Gulf. 18  Warren’s article in this collection.

190

Miers

on Sindh, notes that the castes whose task was to dispose of refuse suffered greater discrimination, although they were legally free.19 In the non-Muslim world of the Indian Ocean, slaves could also have more security and be better provided for than free poor people. Karine Delaye points out that in Indo-China they were hereditary chattels, with all the usual disadvantages, but they could hold most offices.20 Moreover, even ordinary slaves had certain recognized legal rights, including protection from physical or sexual abuse. Delaye suggests that these rights and the fact that they could be freed was a recognition of the slave’s humanity. Even in the western world the humanity of slaves could not be completely denied in practice, and the slave in Islamic society has been described as ‘property with a voice’.21 However, in Indo-China, as elsewhere, these rights were more theoretic than real, given that slaves would find it almost impossible to bring cases against their owners. In Indo-China some slaves were only called upon to work intermittently and, as in the Muslim world, they also might be better off than the free in terms of lifestyle, although they could never be their social equals. This was a common situation. Writing on the Merina empire in Madagascar, Gwyn Campbell emphasizes the fact that, at certain periods, even ordinary slaves were so much better off than the free that they refused freedom because it would have rendered them liable to conscription for corvée labour.22 This raises the question of whether slaves who preferred enslavement to freedom, more particularly, whether those whose ‘worldly success’ and standard of living was greater than that of the poor free population, should really be considered slaves. However, their legal status had not changed and their success was possible just because they were slaves and outside the normal kinship and political groupings of the society. In Sulu, where slaves were imported on a large scale, Warren states unequivocally that slavery was a method of incorporating strangers into society. However, Peter Boomgaard, writing on Indonesia, where slavery was widely practised both by the Dutch colonists and the indigenous people, takes issue with the proposition that the slave was always an outsider. He admits that this is often the case, but disputes the assumption on the grounds that people in 19  Michel Boivin, ‘La Condition servile dans le Sindh colonial—remarques préliminaires’, paper presented at the SSAI (2000). 20  Delaye’s article in this collection. 21  Alaine Hutson, ‘Enslavement and Manumission in Saudi Arabia, 1926–1938’, Critique 11, 1 (2002), pp. 49–70. 22  Gwyn Campbell, ‘Unfree Labour and the Significance of Abolition in Madagascar, c. 1825– 1949’, in Campbell (ed.), Abolition and Its Aftermath.

Slavery: A Question Of Definition

191

Indonesia often captured their neighbours. This turns on the definition of the outsider. In Africa, slaves might be members of the same society as their captors, but were nevertheless strangers to the kin group into which they were inducted, a process which might require ritual sanction and renaming. Members of the society sentenced to slavery for crime, or debtors reduced to slavery for non-payment of their debts became ‘non-persons’ by the very act of enslavement, and hence became outsiders, whose reintegration began when they found an owner. Indrani Chatterjee notes that in India slaves, usually children and girls acquired for various purposes including prostitution, were incorporated into their owner’s household under guise of adoption, concubinage or marriage. A common occurrence, noted here in relation to Indo-China, the BurmaThai border, Arabia, Persia, India, Indonesia and Sulu, was that people sold themselves or their children in times of famine or other disaster. Slavery in such cases was sometimes justified as a form of ‘poor relief’.23 This raises again the question of whether the method of acquisition determined slave status. Was a slave necessarily someone sold or otherwise acquired against their will; or was the defining characteristic how they were subsequently treated; or was it their legal status? Some, as we have seen, were well treated and even rose to high office. Women and children might be treated like full members of their adoptive kin group or families. Men might be indistinguishable in their daily lives from junior kinsmen. Concubines might live in such luxury that their relations joined them. Observers sometimes said they could not tell the slaves from the free.24 Nevertheless, the distinction between slave and free was clear to both owners and slaves and it remained so long after the legal end of slavery. A point not discussed in this collection, but it is well known in relation to Africa, that some slaves freed by Europeans, particularly in Muslim societies, would insist on paying ransom to their owners in order to be free in their own eyes and those of their former owners. Robin Maugham, who bought a slave in Timbuktu in 1958, long after slavery had been outlawed,25 was told by an informant that whereas ‘in his head’ the slave knew he was free, ‘in his heart’ he did not believe it.26 Moreover, servile origins were also often remembered

23  Boomgaard’s article in this collection; Indrani Chatterjee, ‘Abolition by Denial: The South Asian Example’, in Campbell (ed.), Abolition and Its Aftermath. 24  See e.g. Delaye’s article in this collection. 25  At that time the French were still in control. 26  Robin Maugham, The Slaves of Timbuktu (New York: Harper and Brothers, 1961), p. 164.

192

Miers

generations after slavery lost its legal status and led to discrimination when it came to questions of marriage and inheritance.27 Then there is the all-important question of whether slavery was necessarily a system of labour exploitation for economic motives or whether it had other equally important functions. It is no surprise to find the many examples in this collection of the uneconomic use of slaves as status symbols. In IndoChina rulers had hundreds of barely occupied slaves, and there was a whole category of ‘Pagoda slaves’, who served the temples. Boomgaard states that in Indonesia slaves were used in agricultural and urban production, but that both Europeans and Indonesians also kept them to enhance their prestige, even, sometimes, to the point of impoverishing themselves. Slaves were also, as he says, used for sacrifice and were thus a wasting asset. He notes also that the rulers of Bali sometimes had as many as 300 slave prostitutes.28 Andrew Turton notes that slaves were used for conspicuous consumption as well as productive labour in Bangkok.29 Slaves could obviously fulfil both roles and were also a political asset. Apart from providing leaders with loyal supporters, they could be used to produce the goods needed to buy other slaves. They could be given away to attract prospective followers. Concubines were used for this purpose in Sulu. The more adherents, including slaves, a potentate had, the greater his power, wealth and prestige. Then there is the question of whether slaves were always acquired by violence.30 The problem here is that many people enslaved themselves to escape starvation or to gain protection, or parents sold their children to improve their prospects or ensure their survival. The definition of violence, therefore, has to be stretched to include peaceful alienation from one’s natal kin. Chatterjee writing on India, notes that persons who accepted food in time of famine could incur an ‘obligation debt’ resulting in their becoming the slaves of their benefactors. In such cases, hospitality, rather than an act of benevolence, could be an act of aggression. Moreover, if a third party paid the debt, they were considered to have bought the debtor.31 Many of the studies here show that slavery often existed side by side with debt bondage. Legally this is distinct from slavery because the servitude ended with the repayment of the debt. However, this distinction was often more theoretic than real, as the debtors might never be able to repay the debt, which 27  This point is not discussed in this collection as most articles deal with an earlier period. 28  Boomgaard’s article in this collection. 29  Turton’s article in this collection. 30  As maintained by Claude Meillassoux, Anthropology of Slavery. 31  Chatterjee, ‘Abolition by Denial’.

Slavery: A Question Of Definition

193

could be indefinitely inflated. Often the debtor’s whole family laboured with him. His children might inherit the debt, or be forced to run into debt themselves in order to survive. Chatterjee describes how slavery and debt bondage were enmeshed in India. A man might be in debt but the debt was not called unless the debtor tried to leave. If a third party repaid the debt for him, they were considered to have ‘bought’ him. He was thus transferable.32 Debt bondage is an older institution than slavery. It also outlasted slavery as a legal status. Examples from Sindh, Sulu and Indo-China show that its incidence rose when slavery ended.33 Slavery also existed side by side with serfdom. This term has been used almost as loosely as ‘slavery’. A serf was usually conceived as different from a slave because he was in some way attached to the land. He was not saleable but could be transferred with the land. He could not leave and was bound to perform certain services for the landlord. His status was hereditary. However, his position varied in different societies. Boomgaard states that serfs in Java may have been bound to a master rather than to the land. He also describes a ‘more or less serf-like population’ employed making bricks, among other things. The point is that a serf was neither a slave nor a free person. In Sumatra he was described as ‘free as a chicken’ rather than ‘free as a bird’.34 As a number of authors point out, slavery was never a static institution. So far this discussion has been a-historical in that I have taken examples from different geographical regions and different centuries in order to illustrate the problem of definitions. Most of the studies mentioned here discuss the changes in the institutions they describe. For instance, Kim Bok Rae, writing on Korea, discusses the evolution of the nobi from slaves to serfs to independent producers. Campbell describes the evolution of slavery and forced labour in the Merina empire, showing that at times slavery was widespread and at others it contracted. Both these authors also deal with the changes in the ownership of slaves.35 In Imerina, for instance, at certain times there were poor owners, but changes in the economy squeezed them out, while the holdings of the elite escalated. However, this was not a steady progression for poor owners resurfaced later. The studies on Sulu and Korea furnish examples of the range of uses of slaves in a society at any one time.36 In both cases there were agricultural 32  Ibid. 33  Boivin, ‘La Condition servile’; articles by Warren and Delaye in this collection. 34  Boomgaard’s article in this collection. 35  Kim Bok Rae’s article in this collection; Campbell, ‘Unfree Labour’. 36  Articles by Warren and Kim Bok Rae in this collection.

194

Miers

slaves living in their own settlements, and bound to their owners in various ways. For instance, some were sharecroppers, some paid tribute, some gave military service. At the same time there were slaves living in or near their masters’ households under closer control. Slaves in most areas were used not just as field hands or domestics but also as concubines, as officials, as soldiers, as trading agents, as fishermen, as pearl divers, and in an endless range of other occupations. The point is that each occupation determined not only the lifestyle of the slaves but also the nature of their relationship with, and obligations to, their owners. The existence of forced labour together with slavery is also noted. Theoretically, forced labour differs from slavery because the labourers are free and are usually conscripted for a limited time. In practice, however, it may deteriorate into permanent bondage, as in Madagascar in the nineteenth century, and as Shigeru Sato believes it would have done in Indonesia had the Japanese not lost the Second World War.37 The victims of forced labour, like persons in debt bondage, have often been more exploited and worse off than slaves. In theory, the servitude of a debtor, a forced labourer and a contract worker is temporary. Theoretically they are free when the debt is repaid, the contract expires or the job is done. Similarly, the ‘comfort women’ employed by the Japanese as ‘sex slaves’ to serve their troops would regain their freedom when they grew old or ill.38 None of these practices are hereditary, nor are the victims ‘owned’. However, many of the victims never achieved the freedom that theoretically awaited them. They are often described as having been in ‘virtual slavery’—a term which emphasizes the difficulty of drawing a clear line between the slave and the free person. A term not used in these volumes, but used by the United Nations, was ‘collective’ slavery. This was applied to persons suffering extreme discrimination, such as Africans in South Africa in the days of apartheid. This term might well be applied here to the various groups in Japan discriminated against on account of their ethnicity, as described by Tracy Steele.39 Similarly, it might apply to the depressed castes in Sindh and elsewhere on the Indian sub-continent.

37  Campbell, ‘Unfree Labour’; Sato’s article in this collection. 38  Tracy Steele, ‘Race, Ethnicity, and Labor in Japan’, paper presented at the SSAI (2000). 39  Ibid.

Slavery: A Question Of Definition



195

The Politics of Defining Slavery

Finding a universal definition of slavery is all the more difficult because it has been manipulated for political, economic, and social reasons. Michael Salman, writing on the Philippines, demonstrates how both Filipino nationalists and American politicians and officials defined it to suit their political aims.40 The Filipinos, anxious to gain independence from the United States, maintained that colonialism was slavery, and distorted historical records to show that chattel slavery had never existed in the islands. American administrators and politicians, on the other hand, ignored the existence of slavery in Sulu until it suited them to use it to justify their war of conquest. They also played up the existence of slavery to delay granting independence to the islands. Similarly, anti-slavery ideology played an important role in justifying the European colonial conquests of the nineteenth and twentieth centuries, as Delaye shows in relation to Indo-China.41 However, when it came to actually governing conquered territories, the colonial powers delayed taking action to end slavery for fear of antagonizing the slave-owning elites, disrupting the economy, and being saddled with having to look after freed slaves. As Chatterjee shows, the British in India were particularly wary of antagonizing the people in frontier regions for fear that they would turn to neighbouring powers for support. In order to excuse inaction, the British redefined Indian slavery, resorting to euphemisms, describing it, for instance, as ‘mild serfdom’, or ‘a form of poor law’ or ‘unpaid service’. Non-western slavery was described as ‘benign’. Chatterjee points out that they even distinguished between permissible and other forms of slavetrading—allowing for instance, the sale of women and children for domestic purposes but not for prostitution, and maintaining that it was permissible to sell a child for ‘adoption’.42 Moreover, administrators and other observers were wont to call practices slavery without defining them and without any clear frame of reference, with the result that what one administrator might call slavery another might consider adoption or clientage, or even poor relief. Similarly explorers saw what

40  Michael Salman, ‘The Meaning of Slavery: The Genealogy of “an Insult to the American Government and to the Filipino People”’, in Campbell (ed.), Abolition and Its Aftermath. 41  Delaye’s article in this collection. For a discussion of this in relation to Africa see Suzanne Miers and Martin A. Klein, Slavery and Colonial Rule in Africa (London: Frank Cass, 1998); Suzanne Miers and Richard Roberts, The End of Slavery in Africa (Madison: University of Wisconsin Press, 1988). 42  Chatterjee, ‘Abolition by Denial’.

196

Miers

they wanted to see, as the essay on Indo-China makes clear.43 European administrators also had a culturally specific view of freedom—if a slave could leave his/her owner and could earn a living and keep the proceeds, by western standards he or she was free, whereas to the slave real freedom might require not just personal autonomy but complete social and, sometimes ritual, equality with the freeborn.44 It might even involve renouncing some freedoms. Thus when freed some slave women in Muslim societies elected to wear the veil and accept the constraints of purdah rather than retain the greater freedom of movement and the dress of slaves.45 Different societies also had different views of what constituted slavetrading. In China, for instance, the sale of children was only slave-trading if they were sold by someone other than their legal guardians, whereas to the British in nineteenth century Hong Kong all exchanges of person for money or exploitation were illegal.46 Research has been hampered because in many areas slavery is still a sensitive subject. In Africa, where much successful work has been carried out, the descendants of slaves are often reluctant to talk of their origins—evidence that there is still a stigma attached to slave descent. Owners may be equally reluctant to point out persons of such descent, and in some countries it would be illegal to do so. The desire to forget a slaving past has also impeded research. Chatterjee points out that in India, historians have virtually ignored the subject.47 In Arabia, 30 to 40 years since abolition, it seems it is still too delicate to investigate. Scholars are thus prisoners of their sources, which are often unreliable and sometimes almost nonexistent. They are also prisoners of their particular individual political or cultural outlook. Thus Kim Bok Rae points out the differences in the approach to slavery and serfdom between ‘colonialist’ and Marxist Korean scholars.48 Another problem is that, since slavery was never a static institution, all descriptions of it must be put into the context of 43  Delaye’s article in this collection. 44  For an interesting discussion of this point in relation to Africa see Carolyn A. Brown, ‘Testing the Boundaries of Marginality: Twentieth Century Slavery and Emancipation Struggles in Nkanu, Northern Igboland, 1920–1929’, in Journal of African History, 17, 37 (1996), pp. 51–80. 45  See for example Barbara M. Cooper, Marriage in Maradi: Gender and Culture in a Hausa Society in Niger, 1900–1964 (Portsmouth, NH: Heinemann 1997). 46  Elizabeth Sinn, ‘Chinese Patriarchy and the Protection of Women and Girls in 19th Century Hong Kong’, in Maria Jaschok and Suzanne Miers (eds.), Women and Chinese Patriarchy: Submission, Servitude and Escape (London and New Jersey: Zed Books Ltd., 1994), pp. 141–70. 47  Chatterjee, ‘Abolition by Denial’. 48  Kim Bok Rae’s article in this collection.

Slavery: A Question Of Definition

197

their time as well as considered together with the cultural and political outlook of the author and/or the informant.

The Need for a Definition of Slavery

Considering the great range of institutions and practices that have been labelled slavery and the fact they cannot all be encapsulated in one word with a concise definition, one must ask why we continue to use the term. The simple answer is that it gets attention. As Salman has shown, it evokes strong reactions in those accused of practising it, as well as in those who perceive themselves as victims. The reason being that it is irrevocably associated in the public mind with the extreme deprivation and inhumanity of chattel slavery, particularly in the western world. This is a tribute to the work of the British anti-slavery society and, to a lesser extent, its counterparts abroad. This society, now called Anti-Slavery International, together with its antecedents, has conducted unremitting propaganda against slavery for over 200 years. The simplest answer for the scholar is to do what the contributors to this collection have done—use the local terms and describe their meaning. Alternatively they can suggest their own definitions. The academic controversy over definitions has been more useful in showing up the difficulties of reaching a consensus than in pointing to a solution. The definition of slavery, however, is not confined to discussions between scholars. It has some very practical implications for the modern world. On one level it is still the subject of dispute. For instance the government of Sudan has been accused of fostering the capture, sale, and use for menial labour of southern Sudanese women and children, by northern Muslim pastoralists. It rejects the accusation that it is allowing slave raiding and trading and countenancing slave labour. It insists that since slavery is illegal, the victims of the raids are not being enslaved but are ‘abducted’ and used not as slaves but as ‘forced labour’.49 Slavery has been outlawed in Mauritania but nevertheless some ‘freed’ slaves, mostly women and children, are still forced to work for their owners, whose rights are upheld by shariah courts. Such abuses are commonly referred to as

49  ‘Is There Slavery in Sudan? Provisional Observations of a visit to Sudan by Anti-Slavery International Representatives (18–28 October 2000)’ and ‘Exchange of faxes between the Committee for the Elimination of the Abduction of Women and Children (30 August 2001) and Anti-Slavery International (12 October 2001)’.

198

Miers

‘vestiges’ of slavery. But to activists and to the victims they are manifestations of slavery itself.50 At the international level, the League of Nations’ Temporary Slavery Commission established in 1924 to investigate ‘slavery in all its forms’ had great trouble finding a definition that would include all the forms they wished to attack. They ended by defining slavery as the ‘status of a person over whom any or all of the powers of ownership’ were exercised. They listed among ‘all its forms’ chattel slavery, domestic slavery (slaves born in their owner’s household), serfdom, debt bondage, forced marriage, the adoption of children to exploit them, and forced labour. They knew the colonial powers would not agree to a charter to protect native labour nor would they end child labour or interfere with indigenous marriage customs. However, if these were called forms of slavery it would be more difficult for them to resist taking action, or at least signing a treaty binding them to take action. The commission also called for the negotiation of a convention against forced labour to prevent its deteriorating into slavery. When this was signed in 1930, it was aimed at the practices of the colonial powers. At the insistence of Britain, it was carefully worded to outlaw the French conscription of labour under guise of military service, as well as the forced crop-growing practised by colonial powers to stimulate export production. The supplementary forced labour convention of 1957 was designed to outlaw forced labour exacted as punishment for political opposition, as well as its use for economic purposes or a means of labour discipline. This was aimed at the communist countries, particularly Russia with its Gulags. Today it remains as an attack on the Chinese labour camps. The supplementary Slavery Convention of 1956 included all the practices named by the League Commission in the definition of slavery. The definition was extended in the late 1960s to include ‘the slavery-like practices of apartheid and colonialism’. This was added because some of the newly independent states and the communist bloc were anxious to attack colonialism and apartheid but not anxious to have an inquiry into slavery ‘in all its forms’. The new definition gained their support and in 1974 the United Nations Working Group on Slavery was established. It has met almost every year since 1975 to hear evidence from non-government organizations (NGOs) and United Nations bodies. As far as definitions go, it has muddied the waters still further. It has considered slavery, debt bondage, forced labour, child labour, trafficking in persons, prostitution, pornography, sex slavery, sweated labour, the 50  See for instance Boubacar Messaoud, ‘L’esclavage en Mauritanie: de 1’idéology du silence à la mise en question’, in Journal des Africanistes: l’ombre porté de l’esclavage: Avatars contemporains de l’oppression sociale, 70, 1/2 (2001), pp. 291–337.

Slavery: A Question Of Definition

199

exploitation of contract and migrant labour, and of illegal aliens, as well as forced marriage, adoption for exploitation, and the use of child soldiers. If these practices all bear some relation to slavery, it takes a stretch of the imagination to include some of the other practices brought before the group. These include female circumcision, the honour killing of Muslim women by their relations, marriage practices which discriminate against women, incest and the killing of people in order to sell their organs for transplants. And so it goes on—an ever-widening definition of slavery to accommodate whatever human rights violations or labour practices are under attack. In order to bring its title into line with its work, it is now called the Working Group on Contemporary Forms of Slavery. As in the past, calling a practice a form of slavery attracts media attention. It may also shame governments into taking action. Thus, as in the past, slavery was the wild card—the joker which could be played or not depending on circumstances, and the definition was changed to suit a variety of purposes. While, on the one hand, the group has been continually expanding the definition of slavery until it threatens to become a meaningless term, the United Nations and the International Labour Organization (ILO) have been negotiating conventions requiring the precise definitions needed in international law. These conventions set international standards for the protection of labour. Two papers presented at the Avignon conference in 2000, but not included here, dealt with this question. One by Jérôme Joubert made the point that politics are never far from any discussion of the definition of slavery.51 Admirable, he says, as these labour standards are, they are considered in the developing world to be part of an effort by the richer countries to protect themselves from competition from cheap labour in poor countries. The other paper by MarieHélène Advielle deals with setting standards which distinguish child slavery from child labour.52 Thus children are considered to be slaves if they are deprived of education and recreation, and if they are used in ways detrimental to their health or life threatening. The latter include prostitution, military service, riding as camel jockeys, working in fireworks factories, producing charcoal, and spending long hours tied to looms and other crippling occupations. Since child labour is essential to the survival of many poor families and produces much needed export goods, politics, economics and humanitarianism are interwoven in the negotiation of treaties setting these standards. 51  Jérôme Joubert, ‘Travail forcé, travail des enfants et commerce international: qui a peur des normes sociales?’, paper presented at the SSAI (2000). 52  Marie-Hélène Advielle, ‘Le travail des enfants en Inde. Une forme d’esclavage moderne’, paper presented at the SSAI (2000).

200

Miers

Slavery has been outlawed everywhere since 1970. However, ‘contemporary forms of slavery’ are not only widespread but in some cases are increasing. Forced prostitution of women, and of children of both sexes, is growing with the rise of the global sex industry, spurred on by the use of the Internet for trafficking, by pornography, and by linkages to organized crime and drug trafficking. The use of child soldiers, and the sexual slavery of girls forced to serve rebel armies, is increasing with the proliferation of small wars and the collapse of some states. Almost unnoticed has been the kidnapping of thousands of women and girls in China for sale as brides. There is also an active trade in children across boundaries in Africa. Children are sent from rural areas to work as exploited domestics or apprentices in urban areas in Africa, Haiti and elsewhere. Migrant labour, legal and illegal is exploited in sweatshops and even in private homes. Forced labour exists, particularly in Myanmar. The Chinese still sentence political dissenters and others to labour camps producing goods for export. Debt bondage is widespread and taking new forms.53 The many conventions against these abuses are almost unenforceable. This is partly because governments are too weak, too poor, or too corrupt to take action. But it is also because of the lack of clear definitions without which the courts cannot launch successful prosecutions. What is needed at the international level is a new definition of slavery. The old League definition based on ownership is inadequate to counter contemporary forms of slavery. ‘Slavery’ and ‘virtual slavery’ are now catch-all terms without precise definition. At the level of international relations they should either be scrapped or redefined. Abbreviation SSAI conference on ‘Slave Systems in Asia and the Indian Ocean: Their Structure and Change in the Nineteenth and Twentieth Centuries’, Université d’Avignon, 18–20 May, 2000.

53  For a brief discussion of contemporary forms of slavery see Suzanne Miers, ‘Contemporary Forms of Slavery’, in Richard Roberts and Philip Zachuernuk (eds.), Canadian Journal of African Studies (Special Issue: ‘On Slavery and Islam in African History: A Tribute to Martin Klein’), 34, 3 (2000), pp. 714–47. See also the Reports of the Working Group on Contemporary Forms of Slavery, and the publications of Anti-Slavery International, and Human Rights Watch.

History as a Problem of Slaving Joseph Miller The demonstrated—but relatively timeless, and lamented more than explained—ubiquity of slaving throughout the world’s history, historicized, becomes a strategy readily explicable in terms of the marginality of the slavers suggested in the opening chapter in this book. This chapter extends the discussion of the epistemology of history to problematize the conventional, still significantly Eurocentric and structural approach to the history of the world. Here I sketch a historical vision of slaving over some thirty to fifty thousand years, centering on strategies of slaving as both means and consequences of major innovations in the human narrative. The historical processes sketched here in terms of slaving form the basis for the following, more elaborated— though still hardly nuanced—chapters on Africa and the Americas. Readers knowledgeable in the often highly specialized regions and epochs in which professionals normally cluster, and into which here I intrude, however recklessly (but always respectfully), will surely—at least initially—have acute impressions of what I have omitted, or of hotly debated nuances of their fields that I have ignored. However, I hope that these first impressions will be modulated by awareness of the potential insight to be gained from contextualizing their expert knowledge in the broader patterns revealed by a global synthesis centered on strategies of slaving. Inspired by the words of the Malian sage, Tierno Bokar of Bandiagara, I hope for indulgence on a first reading: If you wish to know who I am, If you wish me to teach you what I know, Cease for the while to be what you are And forget what you know

Toward a Comprehensive Global Approach

Slavery around the world pulled off one of the more artful disappearing acts of the early twentieth century. Not that slavers ceased to slave, though the widespread slaving violence of the late nineteenth century did decline with S ource: Miller, Joseph, “History as a Problem of Slaving,” in Joseph Miller, The Problem of Slavery as History: A Global Approach, New Haven: Yale University Press, 2009, 37–73.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_009

202

Miller

imposition of a rough pax coloniana; instead slaving was marginalized in public discourse and not yet embraced as a subject suitably remote for contemplation as history. In the imperialist campaigns of the generation before World War I, European governments had sent armies through Africa and Asia under high-flying banners of eradicating there the abominably inhumane slaving practices of which they had only recently purged themselves, closer to home. Then in the following generation of interwar colonial politics a deep silence fell over the subject. Professional ethnographic descriptions of the largely unknown residents of European `colonies in Africa, South Asia, the Indonesian archipelago, and the Sino-East-Asian region concentrated on rendering rational, and hence intelligible, the strange ways of the “people without history”1 for whom European governments had taken responsibility for nudging into their own visions of the twentieth century. Similar wishful attentions to order and social harmony pervaded understandings of the Native Americas. Slavery, retrograde in the extreme according to the progressive “civilizing” standards of the time, was not part of those pictures, and the reforming European heirs to the anti-slavery components of imperial conquests shifted their attention toward the forced labor policies to which new, struggling, underfinanced colonial regimes not infrequently resorted.2 The colonial regimes themselves had other secrets to keep with regard to slaving. Their minimal resources and personnel, and their innocence of the unprecedented scale of the responsibility they had theoretically assumed for hundreds of millions of people around the world, in practice left them significantly dependent on the collaboration of local authorities in their colonies. Extreme weakness on the ground was the reality of colonial occupation that the British, in particular, dignified as policy, the doctrine they celebrated as

1  Eric Wolf, Europe and the Peoples without History (Berkeley: University of California Press, 1982). 2  Suzanne Miers is the leading authority on slavery in the twentieth century, including AntiSlavery International, the League of Nations, and the colonial politics of slavery. Among her major works are Britain and the Ending of the Slave Trade (New York: Africana, 1975); with Richard Roberts, eds., The End of Slavery in Africa (Madison: University of Wisconsin Press, 1988); with Maria Jaschok, eds., Women and Chinese Patriarchy: Submission, Servitude, and Escape (Hong Kong: Hong Kong University Press, 1994); with Martin Klein, eds., “Slavery and Colonial Rule in Africa,” Slavery and Abolition 19, 2 (1998), special issue; and Slavery in the Twentieth Century: The Evolution of a Global Pattern (Walnut Creek, Calif.: Altamira Press, 2003). Also see Amalia Ribi, “Humanitarian Imperialism. The Politics of Anti-Slavery Activism in the Interwar Years” (D. Phil. thesis, University of Oxford, 2008).

History As A Problem Of Slaving

203

“indirect rule.”3 The problem of slavery inherent in this quiet concession to political reality was that the local authorities on whom colonial officials relied had gained the positions that gave them value to colonizers largely because they had profited, as slaveholders, from the violence of the late nineteenth century that the colonizers had allegedly colonized to terminate. The less said in public about this quiet compromise, the better, and so the classical generation of ethnographers in the 1930s and 1940s found other rural exotica in the colonies to engage the interests of their government sponsors in political control and of social theorists in the harmoniously integrated societies they believed they had discovered. Historical interest in slavery increased in the 1950s and 1960s as a generation of scholars who had not known the subject as politics turned their attentions to a subject sufficiently remote to contemplate. This interest in slavery coincided with growth of academic attention to the remote parts of the world, then emerging from the shadows of colonial propaganda into the light of national independence. But, for reasons having as much to do with the interested scholars’ perspectives from the Americas and Europe, historians found slavery all but invisible there. Whatever echoes of slavery they found bore so little resemblance to stereotypical cotton plantations in the Mississippi Delta of the 1850s that the practice seemed infrequent at best (or worst). They found it mild in contrast to the hard labor that defined the harshness of slavery in the Cotton South and on the sugar plantations of the eighteenth-century Caribbean. The mitigating absences they attributed to it nearly always included the racial aspect entangled with the American institution as they thought they knew it. Slavery in Africa seemed to admit of assimilation of the enslaved rather than exhibiting the inherited permanent exclusions of institutionalized slavery in the Americas. Legal formulations of the institution were absent or, in the case of Islam, less comprehensive than seemed appropriate. It was particularly difficult to identify economic exploitation reminiscent of American plantations. These efforts to seek slavery around the world only in the familiar terms of abolitionist characterizations of American plantations highlight the selective logic employed in this self-fulfilling quest: one can always find (at least some bits of) whatever one is looking for, particularly when one looks intently enough. But always at the cost of finding only that. In the case of seeking slavery construed as a legal institution applied for economic purposes of commodity production, not finding what one thought one was looking for came 3  For an intricate study of slavery in Africa in these terms, Paul E. Lovejoy and Jan S. Hogendorn, Slow Death for Slavery: The Course of Abolition in Northern Nigeria, 1897–1936 (New York: Cambridge University Press, 1993).

204

Miller

to the evident relief of neo-abolitionist historians, embarrassed by their own racialized heritage and eager to redeem the “people without history” throughout the world from any taint of practices—like slavery—that progressive historiography had consigned to the dustbin of the past. Bringing the “natives” into the historical world—where they belonged—meant endowing them with progress too, and progress did not involve slavery. For Africa, as the following chapter elaborates, historians faced the added challenge of rescuing the ancestors of the people wrongly enslaved in the Americas from the stigma of having been involved in the institution from which they had suffered so much. A prominent part of their neo-abolitionist heritage from the earlier politics of abolition had been eighteenth-century European slavers’ ardent defense of their business practices as relieving their captives of their sufferings in Africa as slaves who would be sacrificed by savage kings far more brutal than the charitably Christian planters in the West Indies. A historian raising the prospect of slavery in Africa risked appearing to shift responsibility for the tragedy, or crime against humanity, away from the European participants in the process. Mid-twentieth-century politics of race associated people in Africa with people of African descent in the Americas. And, at the inauguration of research into the regional histories of the world, the empirical limits—the lack of research that Davis had noted, in another context—provided little basis for even the most cautious and respectable members of the historians’ guild to recognize when they were projecting their own pasts, and the politics of the present, or inversions of both, into the voids of real knowledge that they were entering. The historian should consider the historicized alternative in defining slaving strategies in terms of objectives other than producing commodities and should approach enslavement in terms of the experiences of the enslaved—of wondering how people might have experienced the hardships and disabilities of betrayal, isolation, vulnerability, and humiliation—and then assess how enslaved people might have responded constructively in terms of those understandings. A half century later, innocence no longer imprisons Western scholars in their own perspectives, and scholars from these once-remote parts of the world are beginning to broach a subject of a continuing sensitivity no less intense for them than for modern Europeans and Americans.4 4  Three recent evocations of these politics are Apex A. Apeh and Chukwuma C. Opata, “Social Exclusion: An Aftermath of the Abolition of Slave Trade in Northern Igboland, Nigeria,” William and Mary Quarterly 66, 4 (2009), 941–58; Sandra E. Greene, “Modern Trokosi and the 1807 Abolition in Ghana: Connecting Past and Present,” William and Mary Quarterly 66, 4 (2009), 959–74; and Ella Keren, “The Transatlantic Slave Trade in Ghanaian Academic

History As A Problem Of Slaving

205

However, the enabling studies of slavery extended into other parts of the world over the last four decades built on evidence grounded in the structural anthropological concepts—lineages, states—available at the time. This sociologically framed scholarship has remained only incipiently historical, in the rigorous sense defined in the first essay in this book. Alternative inspiration, or inspired alterity, has been slowest to contribute to understanding slavery in the New World, including North America. Here (from where I am writing, in Virginia), in spite of recent masterworks,5 historians’ personal positioning in the midst of slavery’s modern legacies have made many of them confident that they understood what they were looking for, even in earlier times, well before the classic antebellum formulation of slavery as an institution. Alternative perspectives have also been slow to permeate work on slavery in the ancient Mediterranean, because so many historians identified with it so strongly that they thought they knew what to make of the very limited sources at their disposal.6 Elsewhere, Africanists and specialists in the Muslim world (often led by historians working on Islamic Africa) started with a clearer sense of how little they knew, although they compensated for the initially fragmentary evidence at hand by attempting to fit whatever they had into the sociological framework of an institution rather than seeing the array of particular historical

Historiography: History, Memory, and Power,” William and Mary Quarterly 66, 4 (2009), 975–99. 5  Philip D. Morgan, Slave Counterpoint: Black Culture in the Eighteenth-Century Chesapeake and Low Country (Williamsburg: Omohundro Institute for Early American History and Culture, 1998); Walter Johnson, Soul by Soul: Life Inside the Antebellum Slave Market (Cambridge: Harvard University Press, 1999); Ira Berlin, Generations of Captivity: A History of AfricanAmerican Slaves (Cambridge: Belknap Press, 2003). No similar syntheses yet exist for the Caribbean region, the Hispanic Americas, or Brazil. However, see Herbert S. Klein and Ben Vinson, African Slavery in Latin America and the Caribbean, 2d ed. (New York: Oxford University Press, 2007). 6  The recent controversies surrounding a very technical acknowledgment of the multiple backgrounds of classic Greek culture—including Egypt, construed by the publisher as African in a racialized sense—indicate the continuing extent of the sensitivities Finley highlighted in Ancient Slavery and Modern Ideology; see Martin Bernal, Black Athena: The Afroasiatic Roots of Classical Civilization, 3 vols. (New Brunswick: Rutgers University Press, 1987), and ensuring discussion in, for example, Mary R. Lefkowitz and Guy MacLean Rogers, eds., Black Athena Revisited (Chapel Hill: University of North Carolina Press, 1996); and David Chioni Moore, ed., Black Athena Writes Back: Martin Bernal Responds to His Critics (Durham: Duke University Press, 2001). These controversies continue.

206

Miller

strategies or practices.7 Asia, particularly maritime southeastern Asia, came into focus next, though again only partially on terms of its own, with occasional extensions to China and Korea, where world historians had long, and comfortably, interpreted empires and civilizations by their resemblances to the modern West’s congratulatory image of itself, and thereby marginalizing slaving yet again.8 Ethnohistorians of Native North America followed the lead of the Africanists into positioning slavery in small communities structured in metaphors of kinship.9 Non-Muslim India lingered even longer in the obscurity of exoticizing colonial conceptions of caste and of labor bonded by means other than slaving, though South Asia is now becoming a fascinating example both of the unsuspected strategies of slaving employed there and of the ideological processes of colonial rule that obscured them from ethnographers and historians.10 There is, of course, a salutary lesson here for historians: they 7  For my characterization of African history generally in these sociological terms, see Joseph C. Miller, “History and Africa/Africa and History,” American Historical Review 104, 1 (1999), 1–32; reprinted in Gad Heuman and James Walvin, eds., The Slavery Reader (London: Routledge, 2003), 707–38. An innovative attempt to extend the conventional framing of slavery as a legal institution to Africa was Paul E. Lovejoy, ed., The Ideology of Slavery in Africa (Beverly Hills: Sage Publications, 1981); I discuss further details of the literature on Africa in the following chapter. 8  Or slavery had been emphasized primarily in terms derived from highly structural and abstract Marxist theorization of an ancient “slave mode of production,” systematically overriding culture and meaning as well as sometimes straining evidence. For pioneering synthetic works, see James L. Watson, ed., Asian and African Systems of Slavery (Berkeley: University of California Press, 1980); Anthony Reid, ed., Slavery, Bondage, and Dependency in Southeast Asia (St. Lucia: University of Queensland Press, 1983); Martin L. Klein, ed., Breaking the Chains: Slavery, Bondage, and Emancipation in Modern Africa and Asia (Madison: University of Wisconsin Press, 1993). More recently, Gwyn Campbell, ed., “The Structure of Slavery in Indian Ocean Africa and Asia,” and Edward A. Alpers, Gwyn Campbell, and Michael Salman, eds., “Slavery and Resistance in Africa and Asia,” Slavery and Abolition 25, 2 (2004), both special issues. 9  Among recent notable examples, Claudio Saunt, A New Order for Indians: Creeks and Seminoles in the Deep South, 1733–1816 (Cambridge: Cambridge University Press, 1999); James F. Brooks, Captives and Cousins: Slavery, Kinship, and Community in the Southwest Borderlands (Chapel Hill: University of North Carolina Press, 2002); Alan Gallay, The Indian Slave Trade: The Rise of the English Empire in the American South, 1670–1717 (New Haven: Yale University Press, 2003). 10  In the European historiography of world slavery, the guilty heritage of colonial rule in the tropical world plays an obscuring role comparable to the American problem of the contemporary politics of race. For India, Indrani Chatterjee, Gender, Slavery and Law in Colonial India (New Delhi: Oxford University Press, 1999), and Indrani Chatterjee and Richard Maxwell Eaton, eds., Slavery and South Asian History (Bloomington: Indiana

History As A Problem Of Slaving

207

should be careful about presuming what they are looking for, because they are likely to find it. A new and growing scholarly cottage industry is now finding copious evidence of slaving in all the prosperous regions of the Renaissance Mediterranean: Italian cities, parts of the Iberian Peninsula, the Balearic and other islands. Among these discoveries are the complementing seizures of hostages, most of them enslaved, by the Christian and Muslim adversaries confronting one another through the religious tempests then roiling that conflict-torn sea.11 Everywhere, including Europe, rereadings of anthropological accounts, reexamination of archives and other seemingly familiar sources, and new research have now established an empirical base sufficient to support a serviceably contextualized, hence humanly motivated rather than model-driven, history of slaving on a globally inclusive, and perhaps even analytically comprehending, scale.12

In the Beginning

Having suggested how treating slavery as an institution, peculiar or not, has obscured the global historical processes in which marginal interests thrived by slaving, having offered the epistemological parameters of a rigorously University Press, 2006). For comparisons of slavery with the debt bondage so often attributed to Asia, see Alain Testart, L’esclave, la dette et le pouvoir: études de sociologie comparative (Paris: Errance, 2001). 11  For the last, John O. Hunwick and Eve Troutt Powell, The African Diaspora in the Mediterranean Muslim World (Princeton: Markus Wiener, 2002); John Wright, “Morocco: The Last Great Slave Market?” Journal of North African Studies 7, 3 (2002), 53–66; Linda Colley, Captives (New York: Pantheon Books, 2002); Robert C. Davis, Christian Slaves, Muslim Masters: White Slavery in the Mediterranean, the Barbary Coast, and Italy, 1500– 1800 (New York: Palgrave/Macmillan, 2003); John L. Wright, The Trans-Saharan Slave Trade (London: Routledge, 2006). 12  For a reasonably wide ranging, up-to-date array of the known practices of slaving, see Paul Finkelman and Joseph C. Miller, eds., Macmillan Encyclopedia of World Slavery, 2 vols. (New York: Simon and Schuster, 1998), and (more briefly) Martin A. Klein, Historical Dictionary of Slavery and Abolition (Lanham, Md.: Scarecrow Press, 2002). Two other recent encyclopedias are designed on the same premise but achieve less coverage: Junius P. Rodriguez, ed., The Historical Encyclopaedia of World Slavery, 2 vols. (Santa Barbara: ABC-Clio, 1997), and Seymour Drescher and Stanley L. Engerman, eds., A Historical Guide to World Slavery (New York: Oxford University Press, 1998). One eagerly anticipates David Eltis, Stanley Engerman, Paul Cartledge, and Keith Bradley, eds., Cambridge World History of World Slavery, 4 vols. (New York: Cambridge University Press, 2011 and forthcoming).

208

Miller

historical approach to the problem, and having alluded to the range of the recent research that makes it possible here to attempt such an integration on a global scale, I turn to some highly selected, but I hope also highly suggestive, aspects of a world history of slaving.13 On these global scales, slaving arguably contributed to the very beginnings of human history, tens of thousands of years before times recorded in any conventional document-based sense. An approximate date for when history in this inclusive sense may be said to have begun is roughly forty thousand to twenty thousand years ago. I propose that date partly for purposes of debate, but also with a degree of consideration that leaves me confident that the remote era I suggest is at least a professionally responsible estimate. It was the threshold of both humanity and history, marked by abilities to strategize collectively, in contexts of shared meaning. The marker of such strategically effective collectivity is functionally syntactical language: that is, language sufficiently grammatical that speakers can wield it flexibly enough to succeed in confronting novel, unexpected historical contexts. Someone, somewhere, perhaps many competing bands, gradually elaborated syntactical, hence creative, language from cranial capacities that had earlier enabled efficacious, but rote, signaling. In complex ways language enabled the collective creativity and intentionality that distinguish fully human beings from even the most alert of other animals. That remote epoch is also an analytically considered starting point for slaving. With language, early humans learned to strategize together, giving them the unprecedented advantage of being able to rely on knowledge accumulated and then shared and on actions coordinated, and to pass along the successful strategies that they worked out as learned cultures. Slaving was arguably one successful strategy that these earliest peoples hit upon as a means of creating and maintaining the commitment to mutual loyalties that gave such humans the edge in confronting other contemporary, but less inventive, hominids, as well as the changing environmental and other contextual challenges that have brought us to where we are today. That is, humanity, history, and slaving were inherently linked; they were fundamentally mutually constituted. Slaving has also proved exquisitely adaptively efficacious, including subsequent resort to slaving by key initiators in new ways in new contexts during the last twenty thousand years. This deep imbrication ought to be obvious, given the ubiquity of the practice, even today, though our progressivist conviction in our own slavery-freed perfectibility has tended to obscure its elementally human temptation. 13  A fuller narrative survey is in preparation, for the Cambridge University Press, based on the epistemological position developed here.

History As A Problem Of Slaving

209

Historical linguistics provides the primary evidence for my dating of the beginning of efficaciously historical strategizing.14 Research in this field dates the broad origins of the modern language families of the world—that is, the languages that generations of heirs to talkative ancestors found useful enough to maintain them—to approximately the time depth I propose. The last generations of clever hominids, including media favorites like the bulky but slow and mumbling Neanderthals in western Europe, had been adept imitators, but their limited imaginations also narrowed the range of innovations they could copy to random, unintended accidents. Noticing accidents to copy them was a further very low-percentage process and produced cumulative change only through slow processes of evolution. History, defined as humans acting purposefully in collective, or social, contexts out of shared systems of meaning, in contrast follows the subsequent accelerating sequence of collaborative strategies that strategizing humans developed to try to preserve what their ancestors had achieved.

14  Archaeological and other evidence of much earlier (~100,000 bp) hominid collaboration, memory and sentiment, and purposive manufacture of complex tools and even machines (e.g., boats) may indicate an earlier phase of less creative communication (beyond signaling) that may have preceded the modern language families and the adaptability they enabled. For a recent summary, Patrick Manning, “Homo sapiens Populates the Earth: A Provisional Synthesis, Privileging Linguistic Evidence,” Journal of World History 17, 1 (2006), 115–58. However, for my purposes, the key distinction is between (earlier) occasional, even increasingly recurrent, strategizing in reaction to familiar challenges and fully historical creative initiatives eventually institutionalized through slaving. A historical inference from context that might explain such a global appearance would derive from climatic change so dramatic that only groups who developed, inter alia, new, presumably syntactical linguistic techniques of collaboration survived it, thus rendering obsolete all others’ more rudimentary systems of communication. The sequence of key shifts—always incremental, since the law of Ockham’s razor applies—would run from purely (and all but perpetually) instinctive animalistic behavior to individual awareness among hominids and eventually collective human imitation of advantageous accidents, to teaching and thus building reserves of advantageous reactions, which became the threshold of creative collective strategizing for advantage. This step-by-step sequence conceivably converts advantageous evolutionary physical capacity to historical strategizing and success. Awareness of individual heritages, arguably apparent much earlier in archaeological records of burials and other commemorative practices, was eventually extended to collective awareness of a future. Historical consciousness is thus social and derives from orientation toward a shared future rather than toward the past from which it draws its discourse; the past, also imagined as shared, is only fodder for a futuristic ethos.

210

Miller

Focusing more specifically on change as the incremental, inadvertent outcomes that constitute the processes characteristic of history in an epistemological sense, this essentially conservative historical motivation,15 of course, reverses the future-oriented vector of intention that underlies the vision of history as progressive held by people like us who seem, or claim to be, anything but sentimental about their pasts. The progressive ethos of this conventional modern western idea of history confuses the hope that springs from the key human ability to imagine with the lumbering, still basically imitative realities of how humans in fact think and act. In the aggregate, people are risk-averse, wary of the unknown, skeptical about novelty. They generally respond to dilemmas arising from circumstances that they could not influence directly (that is, historical contexts) by changing other aspects of their lives that they sense as of secondary value in the moment in order to conserve other values, regarded as core, that circumstances seem to threaten. An epistemologically historical vision, based on actors’ experiences rather than sociological modeling, therefore perceives change as unintended in its most significant aspects. Long-term changes follow inadvertently from reactions to what initiators are trying to preserve. They are outcomes of intensifications or extensions secondary to initial short-term responses—means to immediate ends, often sacrifices made in pursuit of another losing cause. Even calculated innovations made to escape intolerable circumstances often failed, because all strategies are contextualized in the competitive circumstances of scarcity in which all human endeavors wallow. Adaptations therefore created their own contradictions; anyone’s gain someone else always perceived as their loss, and the losers reacted. This dialectical, ironic, accidental, and historical sort of change tolerates, even welcomes, the lessons in humility that it teaches, even while acknowledging the nobility, virtue, and good intentions of individual actors in the past, by the standards of their times.16 In history, all are responsible, but no one is in charge. With this ironic quality of historical change in mind, we return to understanding those very early humans, and their strategies, in their times, without anticipating what we may now think about what others did later. My emphasis falls on suggesting how slaving may plausibly have contributed, within the generically human ability to communicate and strategize, to fashioning the 15  And restoration, not progress, is the underlying ontology of many historical epistemologies in Africa and elsewhere. It is arguably more common among ourselves than the ideology of progress acknowledges. 16  As distinct from the culture heroes lionized in the memories of claimed heirs. Such semimythologized images are not historical but the products of the historical processes of creating them.

History As A Problem Of Slaving

211

intensely communal ethos of collective (historical) action that they created. What I am calling a communal ethos was—and is—an enveloping moral commitment to the collectivity, a commitment sufficiently intense, naturalized, that it produced the consistent, dependable degrees of collaboration that enabled small groups to sustain themselves, against enormous odds. Elders and parents instilled such loyalties in their children, passing along what then became the sociability and culture that define humans and also history: they think and act routinely in terms of collectivity. This prioritization of the group, in an enduring sense, over the transient individuals in it at any given moment, succeeded widely, though not universally, in early human history, so that its beneficiaries around the world preserved it, and it was observed ethnographically in the first half of the twentieth century. For historians working outside the domain of the modern West’s heavily ideological celebration of modern individualism, it is commonplace, and normatively human. Community is the heart of the historicized problematization of slaving that this essay develops. But community was a historical creation, not a given. Back at the beginning I am positing, someone invented a communal ethos intense enough, and thus sufficiently efficacious, that its heirs literally used it to populate the world. My historical candidates for the credit can be no one but those earliest humans whose historicized context placed them in positions of recovering from late hominids’ lingering, instinctual solitude. They somehow had to forge communal loyalties of sufficient emotional intensity to motivate reliably collective action spontaneously, even—or particularly—under the kind of duress that might otherwise have provoked panicked reversions to “each hominid for himself ... or herself.” Knowing, as we do now, how difficult it is to change what we think of as basic “personality traits,” we can imagine the mental trauma—or shock therapy—necessary to effect so fundamental a transition as the inauguration of history itself. In emergencies, abominations concentrate the mind. Drawing on these arguably universal qualities of the human mind,17 the historian can imagine a kind of twelve-step program of humanization. Someone had to induce—more than impose, since coercion could not have been viable as a continuing strategy to control individuals then only just learning to rely consciously on one another—collaboration as an efficacious premise for historical action. To create such conditioned trust in others it was necessary to move relatively amoral creatures operating primarily with only hominid 17  Without implying monothetic determinism of the sort that attempts to force an artificial choice between nature and nurture, both pervasively present in general, though one or the other may prevail momentarily in given circumstances, that is, historically. The false dichotomy prevails in numerous academic discourses about language, mind, reality, and so on—all singular, hence abstracted, rather than historicized and multiplistic.

212

Miller

predation—and with associations with one another beyond family groups limited to instinctual, self-interested, incidental, and instrumental moments (as all other animals have continued to behave)—toward not only functional but also sentimental, future-oriented mutual dependency. The basic premise is the historical one: community itself was not obvious to people who had never experienced it, or easy for them to invent. A considered historical logic like this contrasts with the mechanistic teleology of characterizing very ancient moves of this sort as inevitable, as abstractly somehow necessary, as some sort of advance, or as inherently obvious progress. Human collaboration is usually acknowledged as an abstract contrast with preceding hominid independence, described as a transition or transformation from one abstract state to another. But for historians, this contrastive logic merely defines the difference; it does not explain how it happened, that is, shifting from abstractions to human historical actors, who did it, why, and how.18 The doubtlessly numerous incidents that opportunistic leaders seized on to induce enduring, functional groups included the strategy of amplifying contrasts between people within momentary assemblages of people, counted as insiders, with others whom they scorned as outsiders. Since these elemental contrasts would not then have been presumed, it is plausible to imagine that they appropriated moments of threat, experienced incidentally (or calculatingly interpreted) as shared. By dramatizing, memorializing fear, they instilled what they could of emotional,19 routinized, hence stabilized commitments to one another among individuals in what had been incipient, unstable, open, and still-ephemeral aggregates. And they could have done so, at least in part, by alienating individual enemy combatants seized in violent encounters with others as inherently dangerous to all within. Since ritual is a powerful collectivizing tactic, they then assembled everyone available as direct participants in, or as witnesses to, collective torture of the captive. The more gruesome the immolation, the more effective the process of forging internalized collective identities, extending even to no less intensely ritualized consumption of the physical remains.20 Such bloody, intimate violence concentrated the 18  “Who” in the sense of positionality, “why” in a motivated sense, and “how” again contextualized to assess historical actors’ limitations in terms of ability to conceptualize. 19  Beyond personally sentimental, evidently much older among hominids, as an extension from primary maternal-infant bonding. The sequence suggested here would apply the notion of historical change as incremental to the emergence of human consciousness itself. 20  See the very suggestive opening remarks on the modern Tupinambá (Amazonia, in Brazil) in Patterson, Freedom in the Making of Western Culture, 13–15.

History As A Problem Of Slaving

213

[collective] mind. Ritualized immolation of others would have constituted the first appropriation of alienated outsiders for internal purposes, or for purposes of internalizing the essential mutuality of human history. In this sense, slaving contributed to the beginnings of history itself. The most direct evidence of strategies of appropriating outsiders for collective, or collectivizing, purposes comes from scattered recent ethnographic reports of small foraging modern communities living in low population densities approximating conditions of long ago—the Brazilian Amazon and the Maori in New Zealand, and in the background of what we know about other Native American practices, not least in pre-Columbian Mesoamerica, where Mayan and Aztec warlords seem to have elaborated ancestral techniques to consolidate an agonistic sort of personal authority over large concentrated populations, otherwise strangers.21 The conventional sense of slavery as an institution hardly accommodates sanguinary ritualized destruction of captured enemies, since they are not held as property and produce nothing of commercial value, but such appropriations of human resources external to given historical contexts, in this case the formative community, for internal purposes readily fits my historical definition of slaving as strategy. The historical effects, the changes it facilitates, are political. The historian also notes that the dialectical qualities of history’s epistemology also flow from the intensity of human relatedness; every human action generates an equal, and likely opposing, reaction. Only an ethos of community holds these in bounds. From the beginning, then, one may define “slaving” as a strategy and practice of agents defined other than as masters, thus acting within specifiable contexts, of claiming absolute power over outsiders to pursue specific purposes internally. At that earliest stage of fully human collaboration, of course, these strategies were as emotional as they were rational, since they were more collective than individual; individuals may think, cognitively, but crowds, as such, react without reflecting. At that initial stage—or subsequently in moments intensely felt to be critically threatening—slaving was imminent and cathartic rather than coolly deliberated in terms of the longer-term consequences 21  And resurrected in subsequent moments of crisis in collectivities in Africa; see following chapter. Also Orlando Patterson, Rituals of Blood: Consequences of Slavery in Two American Centuries (New York: Civitas, 1999). I am indebted to my former colleague Kris Lane at the College of William and Mary for reference to formal sociological consideration of these processes, in Walter Burkett, René Gerard, and Jonathan Z. Smith, Violent Origins: Ritual Killing and Cultural Formation (Stanford: Stanford University Press, 1987). The point is a mainstay of the broader sort of military history, e.g., Robert L. O’Connell, Ride of the Second Horseman: The Birth and Death of War (New York: Oxford University Press, 1995).

214

Miller

evident now to the historian. Historical motivations—but not causation— need always to be considered in terms of varying combinations of these and other, multiple levels of impetus to act. The real-time chaos of actual human lives, the multiplicity and contingency of historical contexts, our lack of direct information about them (because the actors were themselves not articulately aware), make these nonrational impulses central for historians because they are powerfully generative of meanings, motivators of action, and hence change. Speaking historically, rationality slips easily into rationalization, which generates ideologies, and those primarily retrospectively, and so calculatedly obscuring significant aspects of the moments of their creation. A rational focus on cognitively accessible, and thereby only short-range, objectives necessarily makes the eventual significant consequences unintended by the initiators of historical action. The apparent rationality of history lies primarily in the cognitive coherence that historians attempt to impose on the past, more than in the past human behavior that they chart. Coherence is much easier to model sociologically than motivations are to infer historically.

Contradictions of Collectivity and Slaving

On the supramillennial timescale of these beginnings, the epistemological point is that a hypothesis of this sort offers a historical dynamic for how slaving originated with humanity itself, or how humanity employed slaving to develop its distinguishing ethos of sociability. The strategy of using outsiders contributed to effecting the major cumulative change22 of the remote era at the dawn of history. These collective strategies of immolatory slaving would have prevailed for at least ten thousand years, as successful communities defined themselves to reliable degrees of coherence and continuity by building other strategies of maintaining collaboration in less anguished moments, primarily 22  Consider this phrasing as a less teleological, less evaluative phrasing of the modern and Europe-centered idea that Davis considers as progress. As history, these are moments of clarification, or at least ideological definition, on which subsequent generations drew— always for the entirely different purposes of their own times and places. Africanists, who must work with sources framed in terms of orally maintained traditions of ongoing continuity, have converted the static overtones of this phrase into the thoroughly dynamic and creative and purposive processes that generate them by reference to “the (re-)invention of tradition.” A paradigm-shaping initial statement was that of Eric Hobsbawm and Terence Ranger, eds., The Invention of Tradition (Cambridge: Cambridge University Press, 1983); for a subsequent and elegant elaboration, see Jan Vansina, Paths in the Rainforests: Toward a History of Political Tradition in Equatorial Africa (London: James Currey, 1990).

History As A Problem Of Slaving

215

as ranked interlocking obligations, or debt. Successful strategies of creation are not the strategies of continuation. In the instance at hand, distinguishing languages and cultural displays defined communities’ ongoing terms of engagements with others as conceptual contrasts, largely complementing one another. Groups defined themselves around the shared environmental niches in which they lived and exchanged specialized products from their distinctive resources with neighbors. Within these environmentally anchored communities, reproducing lineages exchanged daughters or nieces as wives. Beyond this intensifying range of familiarity, more remote others gradually became human reservoirs of enslaveable outsiders. But, historically speaking, success always breeds its own discontents. The next significant appropriation of these outsiders for the purposes of insiders emerged dialectically and thus again historically enough as a response to the very successes of consolidating the ethos of community. Its ideology of selfless sharing subordinated the individual self to the interests of the group. By the collective standards of the communal ethos, success lay in larger numbers, achieved both by bearing and claiming heirs through time and also by attracting and assimilating others into groups large and coherent enough to pass their distinguishing cultural heritages down through the generations.23 The creators of the early ethical (and eventually also ethnic) communities thereby extended truly ancient sentiments of parental and family attachment to political groupings manipulable historically, that is, in response to changing circumstances. They converted a primordial universal constant of human sentiment into a contingent historical strategy of forming political collectivities. Historical change, as slaving strategies repeatedly confirm, often results from incremental extensions of existing habitual behavior. Humans are inherently conservative, seeking to preserve what in the recent past has seemed valuable or efficacious or even comfortable. But attempting to extend existing circumstances against an ebbing tide of time requires effort—in this case population pressures resulting from the efficacy of sustained collaboration. Intensification is necessary because the circumstances that once favored routine are already eroding. Intensification of focus on immediate needs forces concessions in other, less critical aspects of life, producing radically unintended consequences in the longer term of history, where change is a long-term process composed of the outcomes of short-term struggles to preserve positions against the encroachments of others with access to outside resources, prominently including slaves. 23  In chapter 3 I develop the contradictions internal to the communal ethos.

216

Miller

Eventually, something like ten thousand years or so ago, consolidation of effective, thoroughly socialized (and hence humanized) communities confronted the winners of the first round of human history with the next macrohistorical challenge: coordinating the contributions of the larger numbers of people thus assembled. These larger collectivities, increasingly organized by a manageably few elders or represented to others by chiefs, had to devise ways to call for the personal sacrifices of the many, at least nominally on behalf of the whole.24 Not everyone in larger, more structured groups—by definition— experienced the fullness of the personal attachments and commitments that characterize these small face-to-face communities or possessed the patience or the vision to take much satisfaction in persisting in giving without getting much, or very immediately, in return. Ideology of reciprocity aside, that is to say, the earlier hominid isolation that I’ve implicitly characterized as the challenge that the first humans had to overcome thrived on in the new historical context of strong group loyalties as individualism. Africans, among others, understand this additive, or accumulative, quality of history and celebrate it; ongoing presents are strained extensions of the past. In contrast, modern linear notions of time force us to pretend that we can build the edifice of the future without a solid foundation in what we have been. Since, in a historical epistemology, it is lived contexts that generate meanings, the ideological (or cultural) emphasis on the community had changed individuation from the hominid norm to a human anomaly, if not an abomination. Slaves in the communal ethos, for example, are defined by their exclusion from recognized standing in communities defined by relationality, whether as family or extended kin or clan or any of an infinite variety of other associational collectivities that people around the world created ad hoc, from many historical exigencies of many moments. The extreme vulnerability of the enslaved ensues upon abandonment or betrayal by, or capture from, their own communities of birth, where they otherwise belong ineradicably. Comparable individuation sensed within the community on the parts of their own members was condemned as witchcraft and extirpated in ritual performances of unified rejection reminiscent of the physical immolation to which captured enemies had been subjected long before. In the communal ethos individualism was correspondingly muted or converted to making personally distinguishing contributions to the group. Individuals exercising degrees of personal authority within such groups merely channeled specific, often ad hoc aspects of an abstract, collective domain 24  More detailed references to the key literature on these processes in Africa will follow in the succeeding essay.

History As A Problem Of Slaving

217

of potential power, whether imagined as ancestors, nature spirits, or deceased predecessors, all ultimately accessed through the authority of the group. The personal presence of influential individuals was closer to modern notions of earned respect than to our vested, single-vectored notion of power. Chiefs, in a historically precise sense, did not exercise an exclusive personal power within the group but represented it as a whole to outsiders; their authority was oriented externally on behalf of the collective, not exercised individually and internally. Within this collective context individuals who claimed personal power overtly—and few managed to do so until relatively recently25—did so in significant part through fears of them as witches, since the only way they could have stood apart in the ways they seemed to do must have been through secret access to powers external to the group and invisible to and inaccessible by others within it. Operating surreptitiously, they were dangerously uncontrollable. Nonetheless, within larger, complex groups hierarchy and personal rivalries inevitably emerged. Struggles for centrality to groups created to promote human reproduction often revolved around competition for women as means of reproducing children as heirs—that is, to propagate the future of the collectivity by creating potential successors outnumbering the descendants of rivals and prepared to enshrine their fathers and grandfathers26 in the ancestral sphere of spiritualized collective authority. Particularly after the development of agricultural strategies capable of sustaining large settled communities (and continuing into the recent past in times and places where growing populations had not put demographic pressure on available land), internal ideologies of legitimacy accorded respect to those who acquired enough experience and who had waited for genealogical seniority to envelop them in descendants and other human relationships of obligation within the group. Consolidation of control by seniority and siring of children by these elderly few disadvantaged no less ambitious and much more impatient junior males, or at least delayed their progress toward marrying the women who would become mothers of children recognized as members of the group, propelling them along the internal road to eventual prominence. Since greater population densities had made land scarce and valuable, marginalization developed along the dimension of residence and rights to using it; precedence generally went to first occupants, and later arrivals were tolerated only in formally subordinated ways. Younger men, guests present only on contingent terms, and other marginal residents were thus dependent for access to the primary valued local resources and had nowhere to look for the dependents who spelled 25  See chapter 3 for the definition of personal power in Africa, using slaves. 26  Or, in matrilineal contexts, uncles (and not females, except rarely).

218

Miller

success (respect, honor) other than outside the group. Outsiders, people from other communities or from the communities of others conceptually vulnerable to capture and reincorporation as slaves became reservoirs of personnel, mostly female, where marginal men could obtain dependents and bring them into the community, under their personal control, by means other than the reproductive ones validated culturally, guarded collectively, and hence beyond their reach. That is to say, once large communities or other more categorically differentiated local groupings intensified internal competition for personnel, the acquiring of isolated, vulnerable outsiders became the strategy by which the internally marginalized in communities nominally structured by local descent pursued ambitions as individuals against the restrictions of the community. Slaving of this individuated sort, which challenged the communal ethos rather than maintaining it, would have prevailed in more populous areas of the world from the consolidation of (female) labor-intensive hoe cultivation—and all the social complexity entailed in managing its dependencies—into the twentieth century. Elements of its practice survive in more than a few communities even today. As a result, rich ethnographic evidence illuminates the dynamics of slaving in contexts of hoe cultivation throughout the world.27 In them, male outsiders are appropriated opportunistically, usually as captives taken in wars, and appreciated at least as much for their sheer subordinated presence, often in humiliating assignments to women’s work, as for any material thing they might produce. Or they may be employed in utterly excessive and superfluous displays to manifest or perform the prestige of patrons who can afford to indulge in such competition and whose prominence depends on such displays. The value even of male slaves rests not only on the personal obeisance they display, on their conspicuous availability for the most whimsical of demands rather than routine services necessary to the group, but also on the reserves of distinctive knowledge and skills from elsewhere that they might add to a 27  And much of Patterson’s Slavery and Social Death insightfully details this evidence in terms of correlations with isolated elements of the circumstances that historians would integrate to build the sort of thick (Geertzian) description that they appropriate as meaning-generating context. Clifford Geertz, “Thick Description: Toward an Interpretive Theory of Culture,” in The Interpretation of Cultures: Selected Essays (New York: Basic Books, 1973), 3–30. In the context of this essay, Geertz thus historicized a rather abstracted ethnography (and anthropology) by contextualizing its objects of study. However, such ethnography remains less than historicized since it prioritizes context only as structural framing rather than becoming fully historical by following the dynamics of change that such contextualized (thereby motivated/meaningful) actions always generate in human (social) contexts.

History As A Problem Of Slaving

219

local community’s ability to tap its resident talent to deal with unpredictable historical contingencies.28 However, everywhere in the world before the sixteenth-century in the Americas, women and children were the great majorities of those enslaved.29 They joined in performing the literally vital responsibilities of reproducing and maintaining the community assigned to females, most of them accorded internal status as wives, or, in the relational context of the communal ethos, eventually as mothers of the children they bore for it. The respectable wives controlled by the older men were protected by accessible and attentive kin of their own in neighboring communities, through guarantees affirmed in formal marriages that were more treaties between the groups of kin of the married pair than the isolating coupling of individuals established by marriages in the modern West. But isolated females acquired by slaving lacked the protections of these ongoing local relationships. They were accordingly vulnerable to arbitrary disposition or to abuse by the men who had claimed them. Their captors might marry them and claim for themselves the children they might bear, or they might distribute enslaved females as wives to others within the community, thereby making clients of the favored husbands or currying the favor of influential senior men.30 Both of these strategies allowed younger or marginalized individuals to simulate descendants and dependencies otherwise reserved to insiders and elders. The spuriousness of using dependent outsiders to simulate legitimate methods of attaining rank limited the attractions of this strategy mostly to men of lesser standing—that is, the marginality of slavers generally as they turned up in historical contexts of reproducing communities. Yet such individuated ambition, as with all initiatives considered in historical contexts, provoked dialectical responses. Slaving to advance individually by assembling outsiders undermined communal solidarity and so challenged

28  Awareness of and thus motivation from what I am treating as context is overwhelming in this communal ethos and contrasts strikingly with the tendency of our individualistic ethos to encourage thinking that utterly disregards context (a. k. a. objectified reality external to the observer and abstracted sociological modeling), often reaching narcissistic proportions and sometimes, in degrees that endanger others, as megalomania. These are mental diseases of civilization, to the (considerable) extent that we equate civilized with individualism and its cognate, personal ambition. 29  Campbell, Miers, and Miller, eds., Woman and Slavery. 30  Miller, “Displaced, Disoriented, Dispersed, and Domiciled,” and many studies on Africa and on Native North Americans.

220

Miller

elders, the guardians of the communal ethos.31 Hence communities tended to regain control of outsiders acquired by slaving by absorbing them into their own collectively recognized universes of multiple dependencies, including juniority, clientage, and marriage.32 The recognized internal status of wife (but not the protection of kin recognized by legitimate marriages) accorded women acquired by slaving left them less directly and personally manipulable than their captors or purchasers might have intended. Communities, in effect, translated the female outsiders introduced as an external resource dangerously empowering to marginal individuals within the group into a collective asset by according them the internal status of wife. Women acquired through slaving gained additional standing, in the form of more and different relationships, when they achieved motherhood. The anthropological literature and most quasi-historical treatments based on it characterize this strategy from the sociological perspective of the abstract structure of the group, into which the female newcomer appears to be assimilated. The familiar ethnographic examples are the natives of North America, who made captive women wives33 or adopted men they captured as sons,34 but the parallel with African strategies has been obscured by structural analyses focused instead on assimilation to the group.35 The historical strategy of integrating women acquired through 31  The dialectics presented here in historical terms parallel those phrased in neo-Marxist terms by Claude Meillassoux in The Anthropology of Slavery: The Womb of Iron and Gold, trans. Alide Dasnois, foreword by Paul E. Lovejoy (Chicago: University of Chicago Press, 1991) (originally Anthropologie de l’esclavage: le ventre de fer et d’argent [Paris: Presses Universitaires de France, 1986]). Fuller discussion follows in the third chapter in this book. 32  In unilineal, kin-structured environments, numerous specific relationships of respect and responsibility are complementary rather than competitive. Differentiation there according to sex-based metaphors (not directly keyed to biology) is not equivalent to gender, which is a prejudicially differentiating construct characteristic of modern civic societies; for further discussion of this historicizing distinction from the theorized abstraction usually projected, erroneously, into earlier, remote contexts, see Miller, “Introduction,” in Campbell, Miers, and Miller, eds., Woman and Slavery, vol. 1. 33  It is important, as a historian, not to enclose these designations within quotes, since the status given vulnerable newcomers, at least once they had earned the places then accorded them, was real to the people who granted it. Identities were relational and adaptable, not immutable, isolated, abstracted, or individual. One may qualify such status with quotation marks only by invoking external (hence historically irrelevant) and abstract (again, therefore irrelevant) and singular definitions of “wife” or “son.” 34  For example, Theda Perdue, Slavery and the Evolution of Cherokee Society, 1540–1866 (Knoxville: University of Tennessee Press, 1979), chaps. 1, 2. 35  For further discussion and references to relevant literature, see chapter 3.

History As A Problem Of Slaving

221

slaving is a political process of domesticating human resources from outside the group to defuse challenges to its internal rankings, to defeat marginal members’ attempts to advance personally in terms of the very ethos they had used slaving to evade. Without elaborating further specifics, I make the epistemological point here to suggest how history explains slaving not by mechanistic association of isolated variables (as behavioral social-science comparisons of cases would attempt to correlate them) but rather in positioning particular slavers as motivated and enabled in changing specific, complex, multifaceted, and contradictory contexts—in this case, small reproducing communities defined by nominal kinship. The slavers tend to be motivated by and to strategize from positions of marginality; young warriors sent out to roam among strangers are removed from the disabilities of juniority in the presence of elders but gain access to the opportunity to kidnap young women from far enough away to bring them back to their home communities in positions of personal dependence and social isolation. Putting the historical dynamic of slaving in these strategic terms, it is almost a truism to observe that the vulnerabilities of the enslaved, and therefore their utility to their captors, derived from the structural quality that Patterson, Moses Finley,36 and many other students of slavery have stressed as definitional: the slaves’ origins outside the social contexts into which they ended up being introduced, confined, claimed, and deployed against their claimants’ rivals, often also their sometime sponsors. As the products of the politics of slaving contextualized in the complex ethos of community, rather than simply subjected to the theoretical domination of a master, outsiders were new resources adding to the strength of challengers otherwise too weak to compete locally for the resources available within the relevant historical context—in the contexts elaborated here, that of communities structured by descent. Slavers, by seizing or buying outsiders and incorporating them as slaves, evaded the zero-sum, self-defeating calculus of attempting to mount challenges from within.37 36  Patterson, Slavery and Social Death, elaborating Moses I. Finley, “Slavery,” International Encyclopedia of the Social Sciences, ed. David L. Sills (New York: Macmillan and Free Press, 1968), 14: 307–13. 37  A further general point about history: success produces aggregates that are bulky, in which too many people in them are too invested to retain significant flexibility. They are therefore relatively stable and consequently vulnerable to changing contexts. Change originates at the margins, in the regions of instability. Whatever the specific limitations of the famous frontier thesis of American history, it was solidly based epistemologically, as history. For a productive discussion of open frontiers as the locales of change in Africa and a further comment on the ironies of tradition, see Igor Kopytoff, ed., The African Frontier:

222

Miller

The Problematic Politics of Slaving in the Era of Agrarian Empires

If one gulps down millennia at a time, the historical dynamics of slaving become more apparent.38 Particular, usually marginal, interests slaved to effect changes in the historical communities in which they lived, sometimes momentously, since the outsiders were new resources. I am phrasing these dialectical processes carefully to avoid any implication of attributing causation to slaving, much less to the abstraction of slavery. Causation is alien to history’s distinctive epistemology.39 The recurrent efficacy of slaving for particular, usually marginal interests implies no deterministic contribution to the much broader processes leading to greater social complexity, involving many strategies and resources, all with their own corollary complications and contradictions, many of them—technological, ideological—internal to communities and consequently not introducing human resources from outside, as slaving critically did. Slaving focused on introducing specifically human resources, directly and privately controlled, to the exclusion of collective interests, however the controlling collectivity was formulated—from face-to-face communities of the sort that it had helped to create at the dawn of history to the integration of such communities under military warlords across temperate Asia five thousand years ago, and on to the public interests of modern civic nations. In this way slaving, again, contributed to creating the ancient age of military empires two thousand years or so before the current era among large agrarian populations along the Asian latitude running from the Yangtze River in China west to the Mediterranean and northeastern Africa. There, alluvial floodplains The Reproduction of Traditional African Societies (Bloomington: Indiana University Press, 1987). 38  A more nuanced sketch will appear in the monograph-length development of these strategies in preparation. For first-stage elaborations of more integrated versions of the general narrative, see my “Slavery,” for the on-line edition of the Encylopaedia Britannica (as yet unpublished, to my knowledge). Elements of one eventual component appear in “The Historical Contexts of Slavery in Europe,” in Per O. Hernaes and Tore Iversen, eds., Slavery Across Time and Space: Studies in Slavery in Medieval Europe and Africa (Trondheim: Dept. of History, Norwegian University of Science and Technology, 2002), 1–57. Other aspects of the general processes have emerged in the course of coediting the two volumes on Women and Slavery. See also “Slaving as Historical Process: Examples from the Ancient Mediterranean and the Modern Atlantic,” in Enrico Dal Lago and Constantina Katsari, eds., Slave Systems: Ancient and Modern (Cambridge: Cambridge University Press, 2007), 70–102. 39  See chapter 1.

History As A Problem Of Slaving

223

improved with irrigation channels, ox-plough agriculture, and wheeled transportation to move the vast quantities of grain produced with these technologies also became the focus of investments of local labor. The resulting agricultural production was sufficient to support the significant militarization that defined the empires of the era, specially trained professional troops and heavy weaponry that replaced ad hoc militias fighting only occasionally, on the strength of personal valor, and for the communities in which they lived. Militarization, as a strategy, mobilized communities on a permanent basis to fight for others’ interests. Elaborate hierarchies of priests performed the rituals of state cults to deify and legitimate the power of rulers who were, in origin at least, warlords. Approached from the point of view of the underlying agrarian communities, these seemingly familiar ancient empires were externalities, superimposed on the underlying productive population and engaged dialectically with them over control of the local resources and the products of their own efforts.40 Like the integration of community itself through slaving tens of thousands of years earlier, at least in (crucial?) part, someone had to build these ancient military and, eventually, literate civilizations out of contexts of earlier, smaller, but resilient local communities. Warlords had to impose military control over, rather than induce loyalty within, these existing communities of agrarian peasants, living in peasant-adapted versions of the intensely collective ethos at the base of all human history. Local communities recognized no ongoing need for external, and especially not extremely expensive, military overrule, any more than early humans had anticipated the eventual potential and personal constraints of what had become the communal ethos. Families, villages, and districts commanded everyone’s primary day-to-day attention; authority in these communities was multiple, in the sense that people recognized different scales of responsibility in the differing scales of their lives as they participated experientially in these collectivities. The relatively autonomous power of remote, singular military authorities (notionally figures we might term rulers, but only by not quibbling over historical technicalities of how they managed to rule) to exploit ruthlessly rested on their ability to set themselves outside—and ranked as above—the reciprocities of local communities. The resulting tendency of the successors to the conquering generals who had created such warrior regimes to deify themselves, thus transforming their 40  For provocative hypotheses regarding the composite structures of local communities underlying later militarized regimes, see Roderick J. McIntosh, “Clustered Cities and Alternative Courses to Authority in Prehistory,” Journal of East Asian Archaeology 1, 1 (1999), 1–24, and, more elaborately theorized, id., Ancient Middle Niger: Urbanism and the Self-Organizing Landscape (New York: Cambridge University Press, 2005).

224

Miller

political weakness into an ideological strength, achieved a kind of institutionalized extension of individuation, or precisely what the communal ethos condemned as predatory greed itself. That is to say, from the perspective of the communal ethos, rulers of this imperial (not to say imperious) sort were the epitomes of outsider, though in positions of irresponsible strength rather than the vulnerability of the enslaved, the other outsiders on the scene. Successors to the military conquerors who characteristically established dynasties, but were in practice rarely able to live up to the exalted standards set by or attributed to the founders, exaggerated all the more in memorializing them, went on to celebrate themselves through monumental construction or military plundering abroad. They limited their demands on local peasants to levels that they could provide, on an occasional basis, without destroying the local residents’ ongoing devotion to agriculture that sustained their regimes. They therefore accumulated material wealth on the grandiose scales that we, modern individualists and materialists that we are, now admire as civilization. These military rulers had few local allies on whom they could depend or on whose ethic of loyalty they could count. They also slaved, to surround themselves personally with other outsiders, similarly unencumbered by competing obligations to any in the local communities that they thus sought to transcend. Such militarized political power rested inherently on slaving.41 The strategy of political integration by military plundering and slaving that characterizes imperial rule is very expensive and ultimately self-defeating. The elaborated infrastructure, ideological and monumental as well as military, built by slaving strains the ability of local populations to support them, and increasingly so over time. Otherwise, military rulers would not have to use the force at their command locally to extract taxes and other revenues. Military successes intensified the zero-sum competition for local resources. Until nineteenth-century fossil fuel technologies enabled economic growth through radically increased local productivity,42 the human resources drawn off to fight on remote battlefields or assault distant cities drained agricultural economic sectors at home. Militarized regimes therefore tended to cover their growing local deficits by plundering, though also at very high costs, more and more remote 41  In implied contradistinction to chiefs in less militarized places like Africa, who maneuvered internally by coordinating the component interests of their chiefdoms through councils and other collaborative and participatory, i. e., integrative, devices. 42  But merely displacing the external resources that support local concentrations of wealth and power to the raw materials that became the focus of the strategies of modern imperial systems.

History As A Problem Of Slaving

225

regions for the wealth they need, including captives to supplement strained local supplies of labor.43 In ancient Mesopotamia in particular, local military elites came to tap distant populations susceptible to capture, removal, and reintegration as dominated newcomers, and rulers accordingly filled their palaces with attendants acquired through slaving. Slaves also supported the priests of the ideological cults that deified rulers. Enslaved retainers became critically differentiating supports for temples and large royal and aristocratic households filled with domestics, eunuchs, soldiers, concubines, children, and highly trained, widely experienced, powerful personal agents, emissaries, and enforcers. In the longterm historical process of imperial conquests and consolidation, military rulers slaved on imperial scales to support themselves independently of the ethos of communal reciprocities in which individuals had previously (or elsewhere) recruited women through slaving to support personal challenges to the ethos. They had no need to conceal enslaved women as wives or to adopt children but rather flaunted them to display the dominance they asserted, dominance not only of the enslaved but also of the local communities whom they claimed to protect as subjects. Slaving became a significant strategy by which military elites consolidated the infrastructural and then the ideological strength they claimed.44 Early Egypt, whose roots were deeper in time than those of Mesopotamia, peripheral to the commercialized crossroads of southwestern Asia, and less militarized in its basic political economy before the rather late definition of the pharaoh,45 may have relied on slaves correspondingly less. To the extent these rulers turned to slaving to maintain institutionalized military power and to implement a cover of divine legitimacy, they correspondingly structured the populations of outsiders they enslaved ideologically as eligible for, or rather innately suited to, such predation. In the communal ethos, demonization of enemies, both strangers and internal deviants, had taken forms reported in ethnographic descriptions of them as witchcraft. But as larger military consolidation and commercial networks became dependent 43  I will not pause here to take up Gerda Lerner’s famous thesis in The Creation of Patriarchy (New York: Oxford University Press, 1986), in which she emphasized the capture of women by imperial adventurers of the sort sketched here; see also Gerda Lerner, “Women and Slavery,” Slavery and Abolition 4, 3 (1983), 173–98. For discussion, see Miller, “Displaced, Disoriented, Dispersed, and Domiciled.” 44  And that we usually attribute anachronistically to founders, without examining the strategies by which they acquired what their successors enjoyed. Seeing through this teleological fog is always complicated by calculated celebration in subsequent memorialization of the founders as regime-legitimating heroes. 45  New Kingdom, ca. 15th c. bce.

226

Miller

on slaving in the Aegean Peninsula, the Greeks wrapped a pan-peninsular identity around their congeries of highly diverse, local polities and ethnologically reduced those they excluded to “enslaveable barbarians,” although they continued to acknowledge, if only ironically, the ethnically random quality of earlier opportunistic slaving there. This historical dynamic suggests that the dehumanizing qualities attributed to the enslaved intensify with the systematic dependence of the slavers on capturing them; racialization, the modern formulation of this recurrent tendency, is not a peculiarity of Western culture but rather the most recent in a long sequence of other discourses of difference, each reflecting its own times.46 Slaving accordingly staffed large aristocratic households in the cities of the Old World, and later also in more of the New World than acknowledged by the econocentric plantation paradigm of American slavery. The sensitive aspect of this domestic slaving was not labor in some abstract sense, although many of the enslaved were surely worked hard to maintain the households in which they found themselves. Neither were they held exclusively for products they might contribute, though some were trained in artisan skills that could be deployed to engage the domestic household with the commercial sectors of the economy. Significantly, their values also lay in the incremental political advantages that their sheer presence brought to householders engaged in otherwise delicately balanced composites of competing and interlocking personal and family networks.47 A marginal advantage that one such household, enlarged with slaves, gained in the highly competitive politics of singular imperial rule, without incurring reciprocal obligations to others through marriages or other legal arrangements, could compound to a significant superiority. Slaving in such composite, personalistic polities—as nearly all such entities were throughout the world before the consolidation of monarchies and their modern derivatives, nation-states—was significantly political. 46  The implication here is that many symbolic idioms other than race could express despicable alterity, but that the derisive and dismissive aspects of alterity were concentrated on those rendered enslaveable by contingent historical circumstances. To rephrase this abstraction in terms of the familiar conundrum of modern Atlantic history touched on earlier (chapter 1), slaving preceded the derogation that in the Americas eventually become concentrated in a racial idiom. The incremental steps in this process in the modern Atlantic stretched out over two “long” (the seventeenth and eighteenth) centuries. Further footnoted aspects of the process of elaborating racial ideologies, ancillary to slaving itself, follow in chapter 4. 47  A thoughtful reflection on these intricacies in the instance of a recent counterpart is Sean Stilwell, Paradoxes of Power: The Kano “Mamluk” and Male Royal Slavery in the Sokoto Caliphate, 1804–1903 (London: Heinemann, 2004).

History As A Problem Of Slaving

227

The political dynamics of the military regimes of ancient Eurasia, usually known as empires, hinged on systematic slaving to assemble captives from outside the historical spheres of internal, zero-sum competition and to employ them in whatever capacities within that enabled the slavers to gain on more established local rivals. Before the modern era, when direct control of human effort counted for more than military or other technology or money, successful slavers could move from their original positions of marginality to pose direct challenges and even achieve dominance within. Absent modern technologies, for five millennia of the world’s history these political dynamics were exquisitely sensitive to slaving, tending in the long run to favor merchant interests, who at first prospered from supplying armies in the field and handling the logistics of disposing of the captives they took. Depending on how many slaves the marginal merchants managed to acquire—and these numbers were fundamentally contingent on the historical outcomes of competition in particular times and places and in theory entirely indeterminate—their resort to slaving could fail, leaving the old guard safely embedded, if they, in effect, bought out the slavers by acquiring the captives as dependents for their own households or as workers for estates and temples or as soldiers for their armies. On the other hand, if circumstances brought more captives into the hands of the merchant slavers than their established rivals could relieve them of, the captured outsiders accumulated in the hands of the upstarts, empowering them. In the former case, the human resources of the society or polity or economy increased by means of slaving but resulted in no cumulative historical change. In the latter, it was the merchant-slavers who ended up eroding the control of their former aristocratic sponsors. The slaves’ contributions served the purposes of the most innovative actors and sometimes thinkers of their times.48

Monarchs and the Problem of Merchants

In broad terms, during the millennia characterized by horse-based military control of the populations resident in the fertile river valleys of Afro-Eurasia (that is, ca. 3000 bce–1000 ce), the unprecedented power of military and associated priestly establishments allowed them to extend violent conquests to the

48  “Innovative” and “cumulative” are other more historical analogs of what, in a teleological mode, is classed as progress.

228

Miller

limits of their logistical capacities,49 thereby straining infrastructures of transportation and communication built to consolidate power of a personalistic sort, locally. Conquest is always cheaper than consolidation of control. Empires are therefore historical processes rather than sociological structures. To fill the logistical gap opened by imperial success, merchants extended their local operations to become camp-followers in the field. They organized to transcend the growing distances involved and to overcome the anonymity of transactions with otherwise unaccountable strangers by developing commercial methods involving currencies and credit as well as systems of writing antecedent to literacy. But their credit-based commercial investments had greater inherent expansive potential than military plunder as well as greater sustainability, since their inventories and legal obligations (that is, their debts) endured through time and even increased in value in ways that armies and subjects did not. As a result, military rulers found themselves struggling to control the merchants whom their conquering exemplars had earlier enabled as commercial slavers. The classical empires of world history—from ancient Mesopotamia through China, northern India, Greece, Rome, and the military regimes of the Islamic world before the sixteenth century—cycled repeatedly through this historical dialectic of military expansion succeeded by commercial integration and an uneasy mutual interdependence and rivalry. The slaving on which military conquerors depended created political problems for their successors as rulers. It enabled merchants and other ambitious challengers to acquire outsiders and to sequester them in personal retinues removed from the rulers’ growing demands on local populations for taxes, corvée, and other services. By the middle of the last millennium before the current era, merchants in the commercialized areas of southwestern Asia adjacent to the Mediterranean Sea had accumulated assets in quantities that enabled them to create dependence among local farmers through indebtedness, enforced by the power of militarized regimes. To set these commercial strategies within the processes of the military successes of the preceding millennium, Sumer, the Assyrians, and later imitators, initially marginal merchants—always suspect and mundane hangers-on around the established “virtuous” claimants to deity, sanctity, and valor—had thrived on the plundering and captive taking of their military sponsors, not least through slaving. But they had thrived on and beyond the margins of the prevailing land-based military empires of the era, and they were more collaborative than competitive because 49  The historical dynamic emphasized in Paul M. Kennedy, The Rise and Fall of the Great Powers: Economic Change and Military Conflict from 1500 to 2000 (New York: Vintage Books, 1987), applies throughout the history of the militarized world.

History As A Problem Of Slaving

229

they exploited contacts in remote locations and did not need the services of the resident population of farmers or access to the lands they cultivated. Being excluded from investing in production, merchants concentrated their wealth in cash, consumable commodities, infrastructure, and urban households, both the real estate itself and the people they assembled, including slaves, within the featureless exterior walls of these compounds. Building incrementally on this wealth, they then used their cash liquidity to intrude on domestic elites’ access to local populations, extending credit to peasants to relieve pressures put on subjects otherwise desperately beholden to and belabored by rulers or state priests. Merchants loaned grain they had hoarded from good harvests to peasant families to cover short-term deficits in times of famine, whether induced by failed harvests, foreign invasion, or imperial taxation. This balanced arrangement—in which merchants, enriched by their logistical roles in overextended conquests, invested their wealth in the local populations stressed by the same logistical challenges—prevailed in many other parts of Asia for two millennia, from about 500 bce until about 1500 ce. Throughout much of agrarian non-Muslim Asia, particularly the Indian subcontinent, military rulers seem to have reached such accommodations with local landlords and with resident merchants. These balanced arrangements allowed merchants a significant but not overwhelming interest in agricultural production, and so they invested further in the form of cottage artisan production. Although these arrangements have often been classed with slavery as an institutionalized form of unfree labor, as historical strategies they are virtually opposing. As a historical strategy, debt bondage is one component of multiple, shared interests in distinct aspects of the lives and labors of resident agrarian communities, not the exclusionary control over the entirety of an isolated outsider that distinguishes slaving.50 But in the eastern Mediterranean in the last millennium before the current era, a dialectical and thus quintessentially historical elaboration of monarchy as a distinctive strategy of political authority countered the successes of mercantile slaving under the military/imperial umbrellas of the preceding millennium in southwestern Asia. My characterization of political systems as 50  For comments on unfree labor in the context of the historiography of modern slaving for commercial production, see chapter 4. I will integrate Asian regimes into this schema of slaving in Slaving in World History (New York: Cambridge University Press, in preparation). In terms of the alternative outcomes of competing military and merchant interests, the relevant distinction is between the cavalry-based inner Asian military regimes and the more mercantile regimes of the Indian Ocean rim and the southeastern Asian archipelago.

230

Miller

contextualized historical strategies gives a more differentiated, precise, and processual sense to the term “monarchy” than the general literature in political sociology, which focuses on power in an abstracted sense and as a result conflates widely varying strategies of exercising it. The distinction between abstracted power and contextualized strategies of creating and maintaining it parallels the contrast between slavery (as an abstract institution) and slaving that runs through this consideration of the problem of slavery as history. In this historicized sense, the term “monarchy” refers to a singular, direct, and personalistic ideology of power, strongly correlated, if not exclusively associated, with monotheistic theologies. Monarchy consolidated the personalism of the patron-client politics of imperial composites to a single, universally shared, albeit remote and idealized patron. It contrasts with the ideological and practical remoteness of military regimes oriented toward the gods and plundering foreign lands. It thus turned external authority in the personage of the chief in a communal environment, representing the group as a whole to outsiders, to a benevolent patron of the individuals within a polity thus rendered unitary. The paradigmatic example is found in the strictly monotheistic Hebrew monarchy in Jerusalem in the last millennium before the current era. The historical question is how the context of mercantile slaving at that time and in that region challenged, and in the end contributed to, the consolidation of monarchy’s distinctively singular political ethic. The compromise between merchant creditors and their landed, military, and religious competitors was a fragile one. Merchants claimed the agricultural product of indebted peasant farmers and elaborated complex strategies of inducing default on these contractual obligations. Under such financial pressure, members of local farming families could recover only by conceding persons to their creditors. The more successful merchant creditors, operating essentially as loan sharks, grew wealthy enough to use the proprietorial provisions of emerging commercial law51 to force the debtors they had created to sell themselves, individually or, more often, female dependents in their families 51  Which gave legal—that is, collective, or state—sanction to claims of individuals, not necessarily known to one another, as distinct from the prevailing communal ethos of consensus among familiars; the later distinction between personal property transferable at the will of an individual and the real wealth held in some degree of communal (later, state) trust derives from this transition from collective to private interest. Commercialized transactions represented only small parts of the lives of all but the most committed to cash and credit; writing here in a determinedly historical vein, I contextualize mercantile strategies as particular to a certain small segment of historical moments otherwise significantly conditioned by the communal ethos.

History As A Problem Of Slaving

231

to relieve their starving relatives both of potential mouths to feed and of further liability for such debt by the rules of collective responsibility. Notions of personal private property afforded the opportunity to establish exclusive personal claims to such defaulters. They also occasioned yet another aspect of modern slaving that has become a trope of slavery seen as institutionalized. However, the historian must carefully delimit the relevance of the concept of commercialized, fungible personal property to these earlier times and other places. Traces of proprietorial interests in local debtors, as well as commercial transactions involving outsiders as slaves, appeared as early as several millennia ago in legal documents from Mesopotamia, but to construe these isolated commercial transactions or transactional moments as defining slavery at that early time would indulge the fallacy of origins in an extreme degree. It would treat an incidental aspect of the practice as defining. In fact, or rather, as contextualized historically, such state-guaranteed proprietorial rights over persons applied only to moments of definitive transfers of interest in individuals, local debtors as well as outsiders, between contracting strangers. But many other kinds of less individuating transfers of personal responsibility, allegiance, and dependency among groups predominated in moving personnel among families in the Roman sense of large households in ongoing balanced relationships with one another. Property in persons, in the modern commercial sense, was present, but it was far from pervasive, and it was significant primarily as a mercantile strategy to intrude politically on the prevalent, if also tenuous, claims of military rulers to legitimacy in communal societies by extracting debtors from them.52 Such commercialized transactions in people could be regulated by public law, or more often by judicial enforcement of private contracts, within given political domains.53 By the rule of historical change as incremental, the quite specialized documentation from this era represents only mercantile extensions of the range of other means of transferring people, all of them below the radar of the public laws that constitute our evidence. We can hypothesize three outcomes of the general contest between merchants and militarists for control—via credit or coercion—without pretending that these three exhaust the possibilities of outcomes that historicized 52  The implicit distinction here is between slaving as a strategy of recruiting outsiders, by purchase or by force, and other strategies of acquiring more limited rights over insiders through debt, including the practices in Africa termed pawning. Hostages were equivalent transfers of personnel across the boundary between insiders and outsiders. 53  In this case, given the form of the very limited evidence available, as defined by its public laws rather than by the communities in which the great majority of people in fact lived.

232

Miller

research on ancient Mediterranean slaving might reveal. The political sensitivity of successful slaving was greater when it added or removed mostly men to or from local populations. Young females whom indebted peasant families conceded in satisfaction of what they owed to urban creditors, once moved into mercantile households, did not fundamentally alter the delicate political balance between merchants and military rulers. However, during the middle centuries of the last millennium before the current era merchants in the eastern Mediterranean region seem to have indebted productive farmers to degrees that led to seizures and enslavement of enough of them that they diverted agricultural populations from the services and taxes claimed by military rulers. When merchants began to use debt to intrude on the productive, tax-paying population of males on whom rulers depended, the reaction of the home authorities, particularly in urban settings, was expectably restrictive. Military authority loomed in the background of these mercantile intrusions on landed or other noncommercial interests from Syria to Palestine to parts of Greece. Rulers there repeatedly countered growing mercantile indebting and enslaving of peasants by asserting more direct, benevolent authority of their own. They thereby created a monarchical style of authority that, by reaching individuals personally as subjects protected by a good king, countered the similarly personal credit of the merchants. Monarchical authority thereby forgave commercial debt or limited its accumulation in the hands of merchants by intruding on the up-to-then primary loyalties of peasants to their own communities, whether seen as families or villages or tribes. To counter merchant commercial initiatives, rulers limited or forgave the farmers’ debts to creditors.54 These rulers, later and appropriately famed also as lawgivers, thus invented monarchy as a historical political strategy to push merchants to the margins of highly personalized polities, at no cost to themselves. The first half of the last millennium before the current era became an era of kings55 in lands around the eastern Mediterranean, and direct, unmediated subjugation to monarchs, wrapped appealingly in the cloak of protectors and grantors of beneficent liberties, became the peasants’ price of protection against the intrusive commercial strategies of mercantile slavers.

54  For development of this argument, see Miller, “Slaving as Historical Process.” 55  Distinguished from emperors or theocracies in the ideological identification of the ruler with the welfare of people they claimed to rule, often in paternal metaphors extended from those of the communities of kinship on which they intruded. What distinguished kings from warrior-emperors was not the degree of power they exercised, but their paternalistic relation to the society they controlled by means of that power, as contrasted with the primary orientation of remote emperors toward the gods.

History As A Problem Of Slaving

233

If one sets classical Greece against this background of Asian military empires and emergent kingdoms, the city of Athens (from whence comes a large but not necessarily representative portion of the information we have on slaving in that era) represented a triumph of mercantile interests. Athens in the fifth- and fourth centuries bce provides the implicitly paradigmatic example of political consolidation around mercantile prosperity. In effect, military rulers had yielded political recognition and eventually participation to wealthy merchant challengers of the sort elsewhere, with greater expanses of territory and populations to plunder, kept relatively marginalized. In gaining participation in the polity, merchants converted it ideologically from a militarized externality imposed on peasant domestic communities and one excluding merchants to an inclusive (if only imagined) collaboration among the commercially affluent heads of large urban households filled with dependents, including slaves. It became the polis at the etymological root of what we have overgeneralized as politics of any sort. Athens might be called a democracy of slavers. Great patrons assembled wives, clients, and outsiders of other sorts in large domestic aggregates of people. Affluence consisted largely in personal honor performed in significant part by surrounding oneself with dependent personnel in private households displayed in new public spaces. Dependents in superfluous numbers, including skilled and elaborately dressed slaves, became the currency of claims to individuated honor, deserving of respect among peers. Such participatory accords among the honorable, democratic in the Greek modeling of the strategy and republican at Rome, characteristically included the heads of the constitutive urban or urbanized households, familiar in the Greek oeikos and the Roman familia. Since these heads of household competed among themselves within the polity and also against the political collectivities in which they participated in a classic zero-sum game, the competitors could make absolute gains most easily by enhancing local cores of kin and clients with outsiders taken in as slaves. Merchants, who by then ranged widely throughout the Mediterranean, had direct access to the personnel and skills of the entire region. Concentrated back at home in Athens and in other commercial cities, they became the swarms of household slaves and other slaves employed in artisanal workshops that have puzzled historians of Greek democracy, and who caught David Brion Davis’s eye for irony in terms of the progress they have been claimed to represent.56 56  Davis, Slavery and Human Progress; also Patterson, Freedom, for another interpretation anchoring the invention of the ideal of civic freedom itself in the personal enslavement of the era, initially of women. The point, much debated, goes back to Moses Finley, “Was Greek Civilization Based on Slave Labour?” Historia 8, 2 (1959), 145–64.

234

Miller

With no less schematic brevity, I characterize the position of Rome in this array of the politics of slaving.57 In that considerably more martial polity, by the time of the military conquests in the last centuries before the current era that built the empire, the contest for domestic control became a standoff between an incipient integrative corporate polity and the independent-minded, large military-aristocratic households that composed it. The earliest days of the city of Rome had a commercial character that evidently introduced most captives through trade; the resulting commercial elements of the city’s laws, themselves of a public standing characteristic of the world of trade, therefore defined the slave in that limited context in proprietorial terms by their famous analogy with domestic animals, as a fungible thing (res). However, the historian does not reify a formulation that was in fact a legal analogy, specifying not the inherent nature of the enslaved but rather the commercial terms under which transactions in such persons might be settled. The Roman example adapted older proprietorial formulations of the transactional moments of outsiders brought in to live most of their lives as members of domestic households, well out of the public sphere of commercial law, and also beyond the range of much of the documentation of the era. In the last century or two before the current era the wide-ranging and overwhelming armed force of Rome’s military legions provided captives in massive numbers, sufficient to inflate to imperious scales the domestic retinues of the commanders of the armies. They invested the booty of their conquests also in rural lands for agricultural and livestock production for the city and staffed them in part with slaves, in a radically new application of military slaving to agricultural production.58 The strategy of military conquests also allowed the generals at the heads of the legions to convert the urban republic to an imperial system to a great extent built through slaving. So central was slaving 57  As distinguished from sketching a history of slaving through a millennium of Roman history, which deserves another scholar and much greater detail. 58  I am aware of the agitated debate among specialists as to the prominence of slaves among the working populations of the rural estates of the Roman aristocracy. I assert their presence, though not necessarily their dominance, on the basis of the general flows of resident rural Italians into the Roman legions and the city of Rome and on my reservations about the evidence cited essentially to prove their absence, that is, a negative. This alleged evidence consists primarily of the lack of archaeological remains of barrack-like housing resembling some modern, mostly Caribbean sugar plantations, i. e., the stereotyped modeling of slavery as an institution. I find much more plausible the village-like settlements of families that characterized nineteenth-century western Africa, per Meillassoux, Anthropology of Slavery, and various studies in Paul E. Lovejoy, Slavery, Commerce and Production in the Sokoto Caliphate of West Africa (Trenton: Africa World Press, 2005).

History As A Problem Of Slaving

235

to the Roman imperial state that emperors assembled the great imperial administrative household, the familia caesaris, around slaves.59 The values of personal valor and inherent or inherited honor excluded the merchants, whose slaving might otherwise, as it had done elsewhere, have created wealth of proportions threatening to the power of the generals.60 Even more schematically, but no less historical in its political dynamic, the slave-staffed Roman imperial palace combined these very old patterns of military authority, ideologically external to the underlying communal ethos of republican families, with an innovative civic inclusiveness of citizenship for the men who had left their home communities to join the legions to fight for interests they could effectively construe as their own.61 The civic terms of citizenship substituted an aggregate public approximation for the personal communities these men abandoned. The new citizens of Rome proved to be a sometimes unruly crowd, suggesting the problems, in spite of bread and circuses, of attempting to create a sense of political community among detached individuals numbering in the hundreds of thousands.62 The famous public immolations of slaves—gladiators and others—in the coliseums of Roman cities resurrected and expanded to colossal scales the very ancient strategies of consolidating senses of community in much smaller, more intimate contexts by immolation. Republican politics, no less than their democratic counterparts in Greece, thus made slaving a critical strategy both in gaining on one’s rivals within the civic sphere without having to confront them directly and in balancing popular armies and citizenship against an abstract state. In monarchical regimes, in contrast, kings consolidated the defining singularity of their 59  P. R. C. Weaver, Familia Caesaris: A Social Study of the Emperor’s Freedmen and Slaves (New York: Cambridge University Press, 1972). 60  For a slightly more extensive application of this logic to Rome, see Miller, “Historical Contexts of Slavery in Europe” also Miller, “Slaving as Historical Process.” 61  Extending insights I owe to the late and lamented Thomas Wiedemann, “The Arming of Slaves in Classical and Islamic Societies” (unpublished paper, conference entitled “The Arming of Slaves from the Classical Era to the American Civil War,” Gilder Lehrman Center, Yale University, November 16–18, 2000); unfortunately Professor Wiedemann did not live to review the paper for the eventual published proceedings of this occasion; Christopher Leslie Brown and Philip D. Morgan, eds., Arming Slaves: From Classical Times to the Modern Age (New Haven: Yale University Press, 2006). 62  A deliberated anticipation of similar issues of integrating abstract imagined political communities in the modern era, and a suggestion that military conscription has served often as an effective technique of political integration. Africanists know the efficacy of colonial conscription in the two twentieth-century world wars for the eventual course of nationalism in Africa in the 1950s.

236

Miller

authority by limiting the ability of lesser figures in their realms, especially merchants, to build personal autonomy around retinues of outsiders over whom they could assert exclusionary authority. Monarchy and slaving, incompatible in principle, were in practice therefore always in tension. Legal definitions of the standing of the people who entered the households of both the Roman civic state and the later Christian empire as slaves, and who eventually emerged as freedmen, have framed most of the discussion of Roman slavery. Laws are the working documentation of historians seeking slavery in Roman times. A historically contextualized treatment of these legal materials, usually discussed in the arcane and abstract domain of jurisprudence, would distinguish the dramatically changing political contexts that produced them over a time span of more than a millennium and in contexts ranging from a small republican polity to a populous city thriving on remote military conquests and maritime commercial integration, to a military composite empire, to an eventual monotheistic monarchy in Christian Constantinople. The earliest of these laws regulated the commercial terms through which merchants conducted slaving; late republican legislation attempted to manage the hordes of sometimes unruly captives dispersed throughout Italy; the bulk of the later imperial decrees, including monarchical-styled codifications, were essentially laws of manumission, not laws of slavery comparable to those constituting the modern institution. Unlike modern slave laws, which established masters’ control over humans valuable as property, Roman law attempted to balance the competing claims of former masters and a more enveloping monarchicalstyled regime to ex-slaves, freed in growing numbers as heads of households filled with slaves during the era of expansion later tried to consolidate diminishing domains by formally manumitting dependents to a semiautonomous standing as clients. The political question throughout revolved around the exclusivity of the slavers’ claims to the enslaved, or formerly enslaved, relative to the competing claims of the sequence of overarching political domains through centuries of changes.63 This variety of historical contexts and strategies demonstrates the futility of attempting to assimilate them all into a single, timeless Roman slavery (structured) as an implicitly static institution.

63  As there is no recent synthesis of slaving in Rome, pending the appearance of Keith Bradley and Paul Cartledge, eds., The Cambridge World History of Slavery, vol. 1: The Ancient Mediterranean World (New York: Cambridge University Press, 2011), and the existing surveys hardly begin to historicize on the scale proposed here, I have no direct authority to cite for an interpretation deriving primarily from the epistemological considerations at the heart of this chapter. Further specifics will follow in Slaving in World History.

History As A Problem Of Slaving



237

Slaving Contained: Politics in the Millennium of Monotheisms

After the late seventh century of the current era an uneasy, shifting balance between military rulers and mercantile challengers marked the greater Mediterranean region, most of it Muslim for the millennium following the disintegration of what had arguably been a political regime consolidated through centuries of military seizures of captives but that, like its predecessors, reached its logistical limits. The late eastern phase of Roman political consolidation, centered after the fourth century ce at Byzantium on the margin of southwestern Asia, recovered from the disintegrative tendencies of a military regime running out of conquests by tapping both the monarchical ideology of Christian monotheism and by enslaving the human resources of the accessible Black Sea region. In the Muslim millennium, the politics of slaving played out in the context of a monotheistic culture, within which warlords—often ruling as sultans, secular figures without claims to religious authority derived from the Prophet (Muhammed, ca. 570–632 ce)—used slaving both to sponsor and to compete with commercial interests.64 The overwhelming military conquests of the seventh and eighth centuries expanded Islam, the domain of peace, within which no believing Muslim could be enslaved, and put enslaveable populations of unbelievers, or “kaffirs,” far beyond the range of ordinary raiding. Islamic warlords—like their predecessors in similar positions for millennia—cultivated mercantile sources to provide slaves to sustain regimes always vulnerable to challenge from Shi’ite heirs to the family of the Prophet, a formidable clerical establishment, populist Sufi sheykhs, and other claimants to the singular and therefore disputed legacy of Muhammad. Military rulers, powerful but merely in secular terms and thus marginal in the world of Islam, therefore surrounded themselves with eunuchs and other young males captured abroad, trained them militarily, and used them as palace guards to exercise, through them, a highly personalistic power against a strong clerical establishment always capable of rallying devout populations of peasants against excessive secular demands. Merchants themselves were forced outward also by Islam’s prohibition of lending at interest, which inhibited the opportunity of investing in local 64  The schematic quality of this chapter allows no space to comment on such nuances as the distinction between sultanic figures and the early caliphs (and later claimants to the title), who attempted to combine the religious authority of the Prophet with the military power on which they, as successors capable of only a shadowy resemblance to the inimitably singular voice of the Prophet himself, enshrined in scripture, had to rely in practice.

238

Miller

peasant debt. By the fourteenth century, the thriving commercial sectors of the vast Islamic oecumene reached from the Strait of Gibraltar in the west to the Philippines in the Far East. Secular Muslim rulers positioned their slave soldiers in sensitive capacities, which varied with the contexts in which they found themselves but almost never included basic agricultural labor.65 Men of slave origin often were drawn into the highest ranks of the sultan’s advisers, emissaries, and generals because their isolation from local politics and networks of patronage made them reliably loyal. The famous militarily trained mamluk slave palace guards of prosperous Fatimid and Ayyubid Egypt (tenth century through the thirteenth) and in military polities elsewhere in Asia and Africa were also loyal protectors of incumbent rulers against the threat of relatives ambitious to succeed them. Successors continued the practice, since as a dynastic line they needed slave palace guards to protect their military rule from the competing legal and moral authority of Islamic clerics as well as from potential challenges originating in commercial circles, not to mention brothers or other family challengers. Slaving bolstered secular power in a Muslim world otherwise submissive to the word of God. At the same time, merchants throughout the Islamic world consolidated their positions by staffing their domestic households with slaves, mostly politically less sensitive females.66 The recurrent historical dynamics of these processes of slaving included the contradictory consequences of initial military successes capable of producing slaves directly by conquest and capture, without mediation by merchants. When conquerors reached their logistical or geographical limits, they turned to purchasing the slaves at the hearts of their regimes from traders, in this manner enriching the very competitors who came to threaten them. The standoff between the contending merchants and militarists continued because merchants in the Islamic world operated primarily across uninhabited deserts and empty oceans beyond or between the populous agrarian zones of primary concern to territorial rulers and the clerical establishment. Slaving brought returns enough for all, so long as the slaves 65  The famous exception, the tens of thousands of eastern African Zanj assembled in the eighth and ninth centuries to drain the salt flats of the Persian Gulf, proved the rule: they revolted in numbers that required mobilization of much of the military forces of the central Islamic heartland and several decades to quell. On the Zanj, Alexandre Popovic, La révolte des esclaves en Iraq au III e/IX e siècle (Paris: Geuthner, 1976), translated by Leon King as The Revolt of African Slaves in Iraq in the IIId–IXth Century (Princeton: Markus Wiener, 1998). 66  For the domestic politics of gender, Miller, “Displaced, Disoriented, Dispersed, and Domiciled.”

History As A Problem Of Slaving

239

came from enslaveable kaffir populations in Asian hill country, islands of the southeast Asian archipelago, and sub-Saharan Africa, and merchants invested their profits in politically benign ways, that is, in domestic servants at home and commercial operations abroad. The political risks of buying slaves from local merchants also inclined military rulers to favor foreign commercial sources as suppliers of the captives they needed. In the Muslim world, the politics of slaving thus favored the Christian and Jewish communities, tolerated legally as dhimmi, co-heirs to the monotheistic Abrahamic tradition and available to perform commercial services difficult to reconcile with the communitarian base of Islamic law. However, they were also excluded from acquiring politically sensitive interests in the great Muslim majority and from owning Muslim slaves, theoretically any captive acquired by a Muslim, who was required to convert her to Islam. The extrapolitical grand collectivity of Islam thus everywhere harbored non-Muslim communities, and its cities were intricate ethnoreligious composites that reconciled military rule with local commercial accumulation, particularly of women and girls as slaves. In the twelfth and thirteenth centuries, Christian merchants from northern Italy responded to the commercial opportunities at the height of Islamic prosperity in the eastern Mediterranean, then centered accessibly in Fatimid Egypt.67 Starting from economic positions quite marginal to Cairo and other thriving cities of the eastern and southern Muslim Mediterranean, and lacking comparable opportunities in Christian Europe, Venetians in particular recognized the opportunity in slaving for captives to sell in wealthy urban Muslim markets. In the fourteenth and fifteenth centuries, among their many other mercantile and banking strategies, they profited from selling (and eventually etymologically defining) ethnic Slavs from the Black Sea region to the Islamic world, in return for the valuable Indian Ocean spices available there. In an incrementally historical initiative, they extended their sales of Slavic captives to other cities and merchant households in the Christian Mediterranean. In a further historical, dialectical extension of this strategy, by the second half of the fifteenth century competitors from Genoa and elsewhere in the western Mediterranean supplemented these inflows of captive Slavs from the east with ventures out into the Atlantic and eventually enslaved Africans. The succeeding sixteenth-century Atlantic extension of this deep-running sequence framed slaving in the Atlantic in an unprecedentedly unrestrained context of commercial initiative. Merchants, particularly ones foreign to the 67  The massive research of Charles Verlinden remains classic, especially in this context, L’esclavage dans l’Europe médiévale. Vol. 2: Italie—Colonies italiennes du Levant—Levant latin—Empire byzantin (Ghent: Rijksuniversiteit te Gent, 1977).

240

Miller

sponsoring, but in Mediterranean terms very marginal, Portuguese monarchy clinging to the fringe of the Iberian Peninsula, invested in producing sugar on previously uninhabited or depopulated islands in the eastern Atlantic. Although repeating the deep historical pattern of merchants operating most successfully outside the populated territorial domains of Church and monarchy, Italian and German financial investors, for the first time on significant scales, extended uses of these new supplies of slaves from far western Africa from providing domestic servants for the urban wealthy to staffing agricultural production, for commercial purposes.68 The commodity of interest was sugar. Elsewhere in Christian Europe, slaving had figured only marginally in earlier Italian cultivation of sugar cane on Mediterranean islands and shore, not least because the region was too poor to compete for captives with Muslim commercial markets, too Christian to enslave local populations, and increasingly too politically inclusive in monarchical styles that granted personal liberties to subjects of patron-kings to turn to outsiders to support royal power or to tolerate potential local opponents’ seizures of people as slaves whom kings needed as subjects.69 However, the early Atlantic was a political and 68  Joseph C. Miller, “O Atlântico escravista: açúcar, escravos, e engenhos,” Afro-Ásia (Centro de Estudos áfro-orientais, FFCH-UFBA—Bahia, Brazil) 19–20 (1997), 9–36 (translation of “Africa and the Atlantic in World History: Working Out the Lethal Combination of Africans, Sugar, and Plantations in the Atlantic Basin” [unpublished, 1991–93]); earlier version also as “A dimensão histórica da África no Atlântico: açucar, escravos, e plantações,” in A dimensão atlântica da África (II Reunião Internacional de História de África, Rio de Janeiro, October 30–November 1, 1996) (São Paulo: CEA-USP/SDG-Marinha/CAPES, 1997), 21–40. The limited role of slavery in the marginal sugar industry of the later medieval Mediterranean features more fully also in Stuart B. Schwartz, ed., Tropical Babylons: Sugar and the Making of the Atlantic World, 1450–1680 (Chapel Hill: University of North Carolina Press, 2004). 69  For recent perspectives on why slaving might have declined within western Europe, see David Eltis, The Rise of African Slavery in the Americas: The English in Comparative Perspective (New York: Cambridge University Press, 1999), and Blackburn, Making of New World Slavery. Both authors rightly focus on contingency but fall back on structuralist logic by citing absences of conditions that they thereby presumed normative or necessary according to theorized models. The novel emphasis in my approach is the historical contextualization of monarchy as politically incompatible with internal slaving; I will return in chapter 4 to the implications of Atlantic slaving, seemingly safely displaced across an entire ocean, for western European monarchies. My historical phrasing of these arguments works positively from contexts, in this instance the remote but significant attractions of Muslim commercial dynamism and the emergence of monarchical forms of military authority no more compatible with private retinues of slaves than the monarchies of the ancient Mediterranean. The religious ethic that both Blackburn and Eltis

History As A Problem Of Slaving

241

legal vacuum, without competition from the continental landed and ecclesiastical interests that had inhibited mercantile elaboration of slaving in Asian and Mediterranean regions since the last millennium before the current era, It thus provided a propitious launching pad for two larger elaborations of this newly commercialized process of slaving, one in Africa, and the other— arguably no more pervasive—in the Americas.

A Note on the Enslaved

An inclusively historical approach to slaving necessarily takes into account the perceived worlds of the enslaved.70 Conventional attempts to identify the strategies and initiatives of the enslaved center on outcomes more plausible to modern historians than to the individuals enslaved. In accenting, as these scholars do, slaves’ public agency as resistance, or as sustaining communities separate from “the worlds the slaveholders made,” or as devotees of iconic remnants of their African backgrounds in the European cultures of the Americas, as they do, they presume a number of modern, essentially North American perceptions, motivations, and strategies not present in the Old World or, for that matter, absent also in most of the Americas, most of the time. Resistance does not stand out in the narrative developed in this chapter, or in Africa and in the Americas.

cite as minimizing opportunities to enslave within Christian Europe contradicts Davis’s established emphasis on the compatibility of European urban household slaving with Christian ethics and surprisingly, in arguments with strong economic underpinnings, resorts to ethics as somehow causative. In my alternative historical calculus, Christianity was the monotheistic analog of the ideological and political inclusivity of secular monarchy. Both monarchy and monotheism were homologously singular and direct in the access they gave kings (or God) to individual subjects (or believers). They were thereby restrictive of the exclusionary access to personnel that slaving gave to competitors— merchants or militarists. Slavery remained politically acceptable only in transoceanic and thus legally ambiguous spaces; for the legal point, Lauren Benton, “Legal Spaces of Empire: Piracy and the Origins of Ocean Regionalism,” Comparative Studies in Society and History 47, 4 (2005), 700–24, and extended in A Search for Sovereignty: Law and Geography in European Empires 1400–1900 (Cambridge: Cambridge University Press, 2010). 70   Joseph C. Miller, “Retention, Re-Invention, and Remembering: Restoring Identities Through Enslavement in Africa and Under Slavery in Brazil,” in José C. Curto and Paul E. Lovejoy, eds., Enslaving Connections: Changing Cultures of Africa and Brazil during the Era of Slavery (Amherst, N.Y.: Prometheus/Humanity Books, 2003), 81–121.

242

Miller

The enslaved in the Old World were few in number compared to these in the modern Americas or in increasing portions of eighteenth- and nineteenthcentury Africa. Unlike stereotypical gangs of men cutting cane in Jamaica, they were mostly women and children, who lived dispersed in domestic households, not interacting significantly with others who had also arrived through slaving. Their stories as active historical agents accordingly center on their attempts to make places for themselves within villages, communities of kin, and large urban households or in whatever face-to-face communities they might find themselves. Creating and struggling to maintain places for themselves, to belong somewhere, on terms more committed than the most contingent, were their logical responses to the disabling initial isolation of enslavement. The uprootedness or deracination usually emphasized as a proxy for this experience refers to an absence. It is essentially a negative concept derived from the abstract perspective of the modern historian, in relation to what the enslaved had lost, not to what they experienced as slaves. The experience that overwhelmed the enslaved instead was profound existential isolation.71 It is the slave’s experience of being utterly and helplessly alone that historians must emphasize to historicize enslavement for the enslaved in the experiential contexts that motivated their actions. Though absolute lack of autonomous social connections left the enslaved vulnerable to whatever mistreatments they suffered, these were not defining; rather, what motivated them was avoiding the consequences of the underlying isolation. More important as a motivation for people who had been uprooted was their resulting vulnerability to and fear of further arbitrary transfers and hence to having to find yet other new places 71  As one may read in the autobiographies of the survivors, including the famed Equiano: Vincent Carretta, ed., Olaudah Equiano: The Interesting Narrative and Other Writings (New York: Penguin Books, 1995), and Alexander X. Byrd, “Eboe, Country, Nation, and Gustavus Vassa’s Interesting Narrative,” William and Mary Quarterly 63, 1 (2006), 123–48. Also, particularly, Jerome S. Handler, “Life Histories of Enslaved Africans in Barbados,” Slavery and Abolition 19 (1998), 129–41. Also E. Ann McDougall, “A Sense of Self: The Life of Fatma Barka (North/West Africa),” Canadian Journal of African Studies/Revue Canadienne d’Études Africaines 32, 2 (1998), 398–412, and Indrani Chatterjee, “Colouring Subalternity: Slaves, Concubines and Social Orphans under the East India Company,” Subaltern Studies 10 (1999), 49–97, and id., “A Slave’s Quest for Selfhood in Eighteenth-Century Hindustan,” Indian Economic and Social History Review 37, 1 (2000), 53–86. Sandra E. Greene’s biographies of people enslaved along the nineteenth-century Gold Coast (western Africa) elegantly emphasize precisely this existential quality; West African Narratives of Slavery: Texts from Late Nineteenth- and Early Twentieth-Century Ghana (Bloomington: Indiana University Press, 2011). Recent literary studies of eighteenth-century captivity narratives promise an indirect route into arguably analogous experience.

History As A Problem Of Slaving

243

for themselves, starting over again, and yet again, living with the never-ending contingency of whatever relationships they might try to form, in the psychological and metaphysical senses of relationality. Even voluntary resettlement, as mobile individuals in modern times know all too well, is not easy even when it facilitates personal opportunity. How much harder were involuntary separation and relocation under the terrifying uncertainties of enslavement? In Old World processes of slaving, the slaves’ strategies of belonging or being retained were generally private affairs that the mostly women enslaved pursued within domestic entities, villages, and households. The historically interesting variability among their options arose from the differing cultural contexts in which they conducted such maneuverings and the intense rivalries they evoked, in doing so, from insiders to the communities in which they lived, what one might call the “legitimist interests.”72 The slaves in Greek and Roman theater are not political protestors but vibrant violators and exploiters of the intimacies of family life. The civically inclusive Roman imperial state was not a major exception, since asserting a public presence was a primary issue less for the enslaved than for the state itself. Slave revolts, the conventional but ahistorical emphasis that comes out of focusing narrowly on the master-slave dyad, lacked ideologies of freedom in a civic sense of state-guaranteed individual rights; even in the New World before the nineteenth century, rebel slaves had no recourse that might incline them to question slavery itself.

The Politics of Historicized Slaving

In sketching here these general dynamics of four or five millennia of slaving as historical strategies, I am not attempting to diminish the many insights of the skilled historians who see slavery as an institution of personal domination and economic or sexual exploitation. Masters dominated and abused the people they enslaved; some of them profited at their slaves’ expense; and adult males raped young girls. But the same men also aggrandized themselves, gained from, and abused others around them, others who were not enslaved. For many, if not most, of the enslaved, their extreme vulnerability to maltreatment upon arrival resolved into vital, if usually humble and still vulnerable, positions in households and in workshops, in guarding insecure and therefore belligerent sultans in their palaces, and much more rarely on the small family homesteads 72  As Carl N. Degler did, many years ago, with regard to the U. S. South, Neither Black nor White: Slavery and Race Relations in Brazil and the United States (New York: Macmillan, 1971).

244

Miller

and villages at the base of agricultural production. The captive men rowing galleys in commercial ports or diving for pearls or engaging in mining or sacrificed in gruesome public spectacles or disposed to other lethal tasks suffered all the horrors associated with slavery: extreme deprivation, whipped domination, and morbid social invisibility. But they died physically, slaughtered in imperial and commercial immolatory equivalents, including mining and diving as sacrifices to extreme wealth, of the rituals that had sustained the coherence of the earliest human communities. However, these tormented men were the publicly visible exceptions to strategies of slaving that centered on females and children, “displaced, disoriented, dispersed, and domiciled” within the urban households of the wealthy and powerful. Not even the boys and men in these households were exposed to a public, institutionalized slavery of the modern kind. The Roman jurisprudence known (again, as so often, by modern analogy) as laws of slavery in fact consisted of laws of manumission, expressing the struggles of a waning imperial state to establish its own claims to taxes and conscription over former slaves that lords and patrons of powerful households had manumitted (into public space) to earn money for the support of their households. This contest illustrates that what distinguished the enslaved from other dependents was their political potential, the exclusivity of the control that their captors or buyers could claim over them, as outsiders, in the otherwise balanced composite politics of the communal ethos, of military imperial regimes, and merchant republics. Slaving thrived where politics were personalistic and plural, as in all of the historical contexts just named. Only when political authority became singular did monarchs and monotheism tend to exclude slaving from realms seen as unitary; in the pervasive political domain they claimed, slaving was a strategy that would enable the slavers to stand apart, leading back to political pluralism or to composite politics. This politicized and therefore more historical approach to slaving embeds the abstracted master—slave dyad on which the modern historiography of slavery turns in the competitive contexts that both motivated and enabled a succession of interests marginal to prevailing practices—at first incipiently social late hominids, then ambitious individuals in strongly collective local agricultural communities, then military and associated priestly rulers external to the underlying agrarian populations on whom they depended, and eventually commercial and financial interests long kept marginal to military and ecclesiastical authorities at home. All attempted to gain personally by importing human resources from outside the contexts of internal politics where they competed at a disadvantage.

History As A Problem Of Slaving

245

The strategies of slaving I have cited should be understood as defining these varying historical contexts only in the limited historical sense of becoming the politically efficacious, hence motivating of responses, uses of slaves amidst the great variety of strategies followed in every historical context. I am not attempting to characterize any of these broad periods as abstractly homogeneous in terms of the strategies of slaving I use to define each; slavers with many different interests, in many competing ways, and in all times and places appropriated isolated outsiders for purposes of their own, at the intended expense of the others around them. Similarly, the more established interests within these historical contexts had similarly diverse ways of engaging local communities, on varying terms of dependency.73 Such a comprehensive approach describes how things were; my historical selectivity is meant to account for how they became, who changed the rules of the game, and how, by slaving. I defend the historical validity of the characterizations of the epochs I have offered here as distinguished by significant strategic innovations. In focusing on history, I am analytically interested in the innovatory strategies efficacious in the contexts from which they emerged. The strategies I emphasize therefore constitute incremental extensions of received practice, of unanticipated efficacies derived from momentary conjunctures of circumstances. And all began in positions of marginality of as many sorts as there were coherent historical contexts (from the perspectives of those present, not of the teleological retrospective of the historian), stepped outside the local spheres in which they were playing losing hands to introduce outsiders through slaving. Under the elders, ancestors, and chiefs of the cultivating and reproducing communities of the world, younger men advanced their own positions within the communal ethos by bringing in women through slaving but wrapped their personal gains into complementing advantages for other members of their groups. Third-millennium bce military exigencies in the temperate latitudes of Asia promoted individual valor, and war leaders appropriated captives from campaigns in remote regions to build massive and massively expensive personal power and monumental presence at lessened costs to local agrarian populations. When the spoils of plundering ceased to support these costly regimes, successor military and priestly establishments turned to merchants, by then much more efficiently in touch with still more remote regions, to provide continuing streams of helplessly isolated and therefore dependably loyal outsiders to staff imperial overlays on local peasant populations, thus maintained 73  Discussion of these alternatives follows, in the contexts of the epistemological issues that they raise, in succeeding chapters.

246

Miller

for their vital agricultural production and demographic reproduction. Slaving not only supported heirs to pretensions inherited from the past as dynastic lines grew increasingly marginal to their own changing times but also perpetuated the local communities who marginalized them. In the longer term, warlords turned to merchants to obtain slaves to sustain military establishments running out of significant battles to fight—other than mutually destructive internal struggles among successors struggling to live up to exaggerated memories of their founding predecessors. This move led in dialectically historical fashion to unintended challenges from the merchants. Merchants, originally marginal to military raiding, took advantage of the warriors’ growing dependence on their commercial networks to amass wealth in currencies and credit at rates faster than militarists could integrate the spoils of military domination of land and people. In time, merchants, as slavers, threatened to build up retinues of outsiders under their personal control on scales sufficient to compete for the services and produce of the local agricultural populations with older landed, military, and priestly interests. By the last millennium before the current era militarists and merchants had reached regional variants of this long-term historical standoff in much of Asia and northern Africa, where it prevailed until the nineteenth century. In no instance had slaving attained the status of a legal, public institution, where the state presumed to intrude on communities of kin, peasant villages, great Mediterranean households, even the rural domains of Roman generals and senators, and the slaves of the ruling houses themselves. Rather, these private domains sheltered vulnerable and manipulable outsiders, who effectively sustained the composite structures of polities other than monarchies, whether formalized as democratic cities, republics, or empires. Patterson’s “social death” turns out to have strategic political implications, differing in diverse historical contexts. Nor was the process then highly commercialized, other than in moments of acquisition. Even in the most commercially developed parts of southwestern Asia and the Mediterranean, its limited proprietorial aspect was supplementary to other means of transferring dependents directly among households. In the lives of the enslaved, the isolation and anonymity of being transferred from hand to hand like a thing was momentary or at worst intermittent, often a form of punishment for betrayal of the trust expected of a member of the household. From the slavers’ perspective, slaving enabled whatever degree of individual autonomy they could withhold from overweening collective/political/military authority; in increasingly commercialized urban contexts, slaves displayed wealth, performed services, and sustained competitors in the public sector. To the limited extent that authorities external to these large, slave-staffed

History As A Problem Of Slaving

247

domestic households interfered in relationships between slaves and masters living within the high, featureless walls of their compounds, they did so only through the masters, who represented these extended family collectivities to the world beyond. The polity itself was either a republican composite of these great households or an imperial authority external to them; in neither instance did it claim the direct access to individuals that eventually defined monarchs and that modern civic nations later claimed as participatory citizenship. The exceptions to this rule were the rare monarchies of the Christian Mediterranean, and eventually maritime western Europe; I account for their distinctive aversion to slaves in a succeeding chapter. In contexts of so limited and alien a political sphere, personal freedom in the modern, civic, public sense had only occasional significance at best, and even then only for a wealthy few, in particular those with many slaves, who used them to assert themselves in the politics of the composite. The relevant standards of treating the enslaved were therefore ethical and personal, not abstractly legal; for respectable patrons these were matters of honor, and they applied to everyone in the domestic retinues of household heads, not alone to the enslaved. These obligations found expression therefore primarily in the reflections of the Christian and Islamic thinkers on the responsibilities of believers in a monotheistic God, in a vein that we now term religious in a sense that they would not have recognized, contrasting with the political. Slaves fell within the realm of personal, self-motivated, honorable responsibility that we regard as ethical. Reflections of the same ethical order and strategic effect in the secular literate context of ancient Greece to us appear philosophical. These were the quite distinct realms of earlier times to which David Brion Davis attributed continuity in the “legal and philosophic concept ... the idea of slavery” into the Atlantic era. But historicizing them, as I propose, indicates radical discontinuities of slaving in the Atlantic, and not in terms of race but in the unfettered commercialization and consequently expanding public sphere of law and contract in which the enslaved served, first, as collateral for the credit that entrepreneurs required to build enterprises producing agricultural commodities for sale. In effect, borrowers displaced the costs of servicing their debts from themselves and their families—as peasants in the classical eastern Mediterranean and elsewhere had long had to offer themselves in cases of default—onto enslaved strangers whom they could work to death to produce commercial products to repay what they had borrowed or, in the event of a need to liquidate, to sell the fungible human collateral itself. Slaving, positioned within the political contexts of the times and places of marginal interests who used it to promote themselves and understood dynamically as a political strategy, often forced more established rivals to fall back

248

Miller

also on slaving in self-defense, thereby motivating and enabling highly varied changes obscured by abstracted understandings of a singular institution of slavery. The sociological institution thus imagined freezes selected outcomes of the throbbing dynamics of human historical processes in time long enough to contemplate their logical coherence and nuanced implications. But institutions are also ideologies, and, when contemplated historically, ideologies are inherently deceptive, denials of change. They are defensive, often retrospective rationalizations created in belated attempts to preserve the failing stages of spontaneous, opportunistic processes that earlier had coalesced spontaneously at paces too fast to need, or to gain further momentum from, intricate analysis of selected elements of their logic. Ideologies are also retrospective public memory systematically elaborated to wrap the actual historical dynamics of the past in political priorities of the present. As history, these rationalized outcomes therefore explain nothing; they are rather what historians must explain.74 The historical epistemology I outline here is meant to tie some of the many and varied instances of Old World slaving together over a more comprehensive geographical range than the usual Athens-to-Alabama narrative, and to do so in terms of historicized dynamics of integrated incremental changes rather than in terms of static sociological comparisons of isolated cases, abstracted retrospective ideologies, or selective originary continuities.75

74  See note 6, chapter 1. 75  E.g., the role of durability played by so-called Roman laws of slavery in most synthetic work in this tradition; e.g. William D. Phillips, Jr., Slavery from Roman Times to the Early Transatlantic Trade (Minneapolis: University of Minnesota Press, 1985).

Historiography and Research Problems of Slavery and the Slave Trade in a Global-Historical Perspective* Michael Zeuske Slavery research has been dominated since c. 1970 by two cultures of historiography and memory: those of the USA and Brazil—though completely unbalanced from a European perspective, with some 80 per cent of publications and research originating in the USA against 10 per cent in Brazil, despite a very important research institute in Canada (The Harriet Tubman Institute).1 Brazilian global-historical research dominates the history of the South Atlantic and naturally enough that of the Brazilian internal market. In Brazil itself, besides slavery research on the Anglo-American space (centred on the USA), there exists the best, quantitatively most comprehensive and detailed research in the world into slavery, the slave trade, and the slave condition, as well as national post-emancipation research that includes local-historical studies. This too is only natural, given that Brazil was the largest slave society in the world in its time.2 So, in a consequent global perspective, Brazil was the most important slave and slave trade society, and it is today the country with the most important historiography. With reference to the above mentioned distortion, John W. Sweet in a recent article claims that there are myths in the history of the slave trade in a macro-historical perspective: the myth of the primordial importance of North America and the myth of the nineteenth century as a “century of S ource: Zeuske, Michael, “Historiography and Research Problems of Slavery and the Slave Trade in a Global-Historical Perspective,” International Review of Social History 57(1) (2012): 87–111. © Cambridge University Press, reproduced with permission. * I would like to thank David Fernbach for the excellent translation. 1  Michael Zeuske, “Umrisse einer postkolonialen Geschichte der Sklaven und der Sklaverei im Atlantik”, in idem (ed.), Sklaven und Sklaverei in den Welten des Atlantiks, 1400–1940. Umrisse, Anfänge, Akteure, Vergleichsfelder und Bibliografien (Münster [etc.], 2006), vol. 1 of M. Zeuske (ed.), Sklaverei und Postemanzipation. 2  Luiz Felipe de Alencastro, O Trato dos Viventes Formacão do Brasil no Atlantico Sul, seculos 16. e 17. (São Paulo, 2000); Rafael de Bivar Marquese, Feitores do corpo, missionários da mente. Senhores, letrados e o controle dos escravos nas Américas, 1660–1860 (São Paulo, 2004). A very good overview is provided by Herbert Klein and Francisco Vidal Luna, Slavery in Brazil (Cambridge, 2010), bibliography: pp. 321–352.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_010

250

Zeuske

abolition” since 1808 (in reality North America was, until 1850, a periphery of the Spanish Caribbean).3 The remaining historiography of slavery is divided diffusely between other national historiographies: the British (often combined with the US into the Anglo-American historiography of slavery), followed particularly by the Dutch, French, Portuguese, Spanish, Italian, the Baltic states (Denmark, Brandenburg, Sweden), Hamburg, Bremen, and Switzerland.4 The historiography of different internal slaveries in Africa confronts particular difficulties, while the Atlantic slave trade from Africa 1650–1800 is relatively well known.5 Difficulties with the subject in Europe and Latin America also arise time and again in Spain, which from 1493 to 1898 possessed the largest colonial empire in the Americas (still including Cuba and Puerto Rico after 1825), with the longest history of slavery. Spain as a colonial power had no areas for obtaining slaves in Africa until the end of the eighteenth century. The particular view of Spanish Catholicism was (and remains) marked by the “mildness” and “gentleness” (suavidad) of Catholic Spanish-American forms of slavery (slaves being brothers in Christ with their masters). But because many investments (such as urban modernization in Barcelona, Seville, and Madrid, and large firms in the transport and banking sectors), as well as private wealth were drawn from the smuggling of slaves and from slavery to and in Cuba and Puerto Rico in the nineteenth century, the subject remains even today extremely thorny, being in fact scarcely known outside specialized historiographies.6 3  John W. Sweet, “The Subject of the Slave Trade: Recent Currents in the Histories of the Atlantic, Great Britain, and Western Africa”, Early American Studies, 7 (2009), pp. 1–45; Eliga H. Gould, “Entangled Histories, Entangled Worlds: The English-Speaking Atlantic as a Spanish Periphery”, American Historical Review, 112 (2007), pp. 764–786. 4  See the ongoing bibliographies of Joseph C. von Miller, Slavery: A Worldwide Bibliography 1900–1982 (White Plains, NY, 1985); idem, Slavery and Slaving in World History: A Bibliography, 2 vols (Armonk, NY, 1999), now an annual supplementary volume to Slavery & Abolition; Jürgen Osterhammel, “Transferanalyse und Vergleich im Fernverhältnis”, in Hartmut Kaelble and Jürgen Schriewer (eds), Vergleich und Transfer Komparatistik in den Sozial-, Geschichtsund Kulturwissenschaften (Frankfurt [etc.], 2003), pp. 439–466, 459–462; these two historiographies relate very well to the author list in José C. Curto and Paul E. Lovejoy, Enslaving Connections: Changing Cultures of Africa and Brazil during the Era of Slavery (New York, 2004). 5  Ibrahima Thioub, “Regard critique sur les lectures africaines de l’esclavage et de la traite atlantique”, in Issiaka Mandé and Blandine Stefanson (eds), Les Historiens africains et la Mondialisation (Paris, 2005), pp. 271–292. 6  Manuel Lucena Salmoral, La esclavitud en la América española (Warsaw, 2002); idem, Regulatión de la esclavitud negra en las colonias de América Española (1503–1886): Documentos

Historiography and Research Problems of Slavery

251

The hegemony of Anglo-American historiographies threatens a canonization, which is further promoted by a theoretizing history of historiography in central Europe, fixed on the USA, which basically rejects empirical research. The greater part of recent encyclopaedias, textbooks, and atlases of slavery and the slave trade express this hegemony.7 New empirical material is kept at bay with the dual argument that empirical research is quite impossible to conduct in such global-historical macro-connections as global history, plantation slavery, and the Atlantic slave trade (Atlantization).8 Along with the second argument, implicitly drawn from “media science”, of an almost exclusively “perception history”, this leads to a habitus of postmodern epistemology that is hard to attack, a habitus of impossible access to historical reality (and of the ever stronger self-reflexivity of present-day societies). It is from the fund of this slavery that the official memorial culture of slavery draws its material; its scenarios and scripts almost necessarily produce a new historicism.9

para su studio (Madrid [etc.], 2005); José Andrés-Gallego, La esclavitud en la América española (Madrid, 2005). 7  A selection: Kwame Anthony Appiah and Henry Louis Gates, Jr (eds), Africana: The Encyclopedia of the African and African American Experience (New York, 1999); Seymour Drescher and Stanley L. Engerman (eds), A Historical Guide to World Slavery (New York [etc.], 1998); Toyin Falola and Amanda Warnock (eds), Encyclopedia of the Middle Passage (Westport, CT [etc.], 2007); Paul Finkelman and Joseph C. Miller (eds), Macmillan Encyclopaedia of World Slavery, 2 vols (New York, 1998); Robert L. Paquette and Mark M. Smith, The Oxford Handbook of Slavery in the Americas (Oxford, 2010); Junius Rodriguez (ed.), The Historical Encyclopedia of World Slavery, 2 vols (Santa Barbara, CA, 1997); idem, Chronology of World Slavery (Santa Barbara, CA, 1999); Marcel Dorigny and Bernard Gainot (eds), Atlas des esclavages. Traites, sociétés coloniales, abolitions de l’Antiquité à nos jours (Paris, 2006); David Eltis and David Richardson, Atlas of the Transatlantic Slave Trade, foreword by David Brion Davis, afterword by David W Blight (New Haven, CT [etc.], 2010); James Walvin, Atlas of Slavery (London [etc.], 2006). 8  On Atlantization, see M. Zeuske (with Javier Laviña), “Failures of Atlantization: First Slaveries in Venezuela and Nueva Granada”, in Dale Tomich and Michael Zeuske (eds), The Second Slavery: Mass Slavery, World-Economy, and Comparative Microhistories, Part II, special issue of Review: A Journal of the Fernand Braudel Center, 31 (2008), pp. 297–343. 9  Stephan Palmié, “Slavery, Historicism, and the Poverty of Memorialization”, in Susannah Radstone and Bill Schwartz (eds), Memory: Histories, Theories, Debates (New York, 2010), pp. 363–375.

252

Zeuske

Earlier Historiographies of Slavery

The starting points of the modern historiography of slavery are the earlier historiographies of religious orders and missions (including those of Sandoval, Labat, and Oldendorp), Cuban historiography (Saco, Ortiz, and Moreno Fraginals), British West Indian radical thinkers (C. L. R. James, Eric Williams, and Walter Rodney),10 and Brazilian cultural sociology (Gilberto Freyre).11 These studies developed against a broad background, particularly in the historiography of the Caribbean: Dominicans and Jesuits, but also the Bohemian Herrenhuters (Moravians), works by Caribbean planters and slaveholders such as Ligon, Long, Edwards, Moreau de Saint-Méry, and Arango, and a strongly metropolitan and often political-economy-oriented nineteenth-century historiography such as Merivale, Peytraud, and Nieboehr. Jesuit historiography, especially Tomas de Mercado and Alonso de Sandoval,12 had aimed at enquiring into the rightness of Catholic proselytism and the conditions for it. The Jesuit universal mission thought and acted globally. It is interesting that John Kelly Thornton,13 in the context of a newly Africa-centred world historiography, has drawn massively on Alonso de Sandoval, and more recently also on Oldendorp (Protestant mission). Starting in the nineteenth century, Cuban historiography was the first of the three major Atlantic-American traditions of slavery research, both chronologically and in a certain sense also in terms of its early quality. The slave culture 10  Francisco A. Scarano, “Slavery and Emancipation in Caribbean History”, in Barry W. Higman (ed.), General History of the Caribbean, 6 vols, VI: Methodology and Historiography of the Caribbean (London [etc.], 1999), pp. 233–282; Winston James, Holding Aloft the Banner of Ethiopia: Caribbean Radicalism in Early Twentieth-Century America (London, 1998). 11  Gilberto Freyre, Casa-grande & senzala (Rio de Janeiro, 1933), 51st edn: Casa-grande & senzala. Formação da família brasileira sob o regime da economia patriarcal, introduction by Fernando Henrique Cardoso, bibliography by Edson Nery da Fonseca, notes, reviews, and index by Gustavo Henrique Tuna (São Paulo, 2006); idem, The Masters and the Slaves: A Study in the Development of Brazilian Civilization; transi. from the Portuguese of the 4th and definitive Brazilian edn by Samuel Putnam (New York, 1946). 12  Tomas Mercado, Suma de tratos y contratos, edn with introduction by Nicolás SánchezAlbornoz; transcription by Graciela S. B. de Sánchez-Albornoz, 2 vols (Madrid, 1977); Alonso de Sandoval, Un tratado sobre la esclavitud, introduction, transcription and translation by Enriqueta Vila Vilar (Madrid, 1987). 13  John K. Thornton, Africa and the Africans in the Making of the Atlantic World, 1400–1880 (Cambridge, 1998); Margaret M. Olsen, “African Conversion as Jesuit Enterprise”, in idem, Slavery and Salvation in Colonial Cartagena de Indias (Gainesville, FL [etc.], 2004), pp. 60–91.

Historiography and Research Problems of Slavery

253

of what was in the early nineteenth century the most efficient and compact plantation economy also led to research into slavery, and an ethnology and anthropology of slaves: Francisco de Arango y Parreño (1765–1837), the Adam Smith of the plantation economy, as well as one of the founding fathers of the theory of American mass slavery and functional racism, influenced Alexander von Humboldt’s essay on Cuba.14 But there was more. The owner of the “Surinam” ingenio and Professor of Latin at Havana, Anselmo Suárez y Romero (1818–1878),15 conducted a literary slave ethnology; José Antonio Saco (1797–1879), from the township of Bayamo in the east of Cuba, a functional racist like Arango, published one of the first comprehensive world histories of slavery (from the beginnings to around 1850).16 Cuba, in particular the Cuba grande of the technologized plantation economy in the west of the island, was the most “modern” slavery of the nineteenth century, which is why the first visualization of a slave economy in terms of global history, according to the rules of modernity, appeared in the words of an actor of slavery (Cantero/ Laplante, Los Ingenios, a work from which almost all “authentic” illustrations in books about slavery are still drawn today, Mialhe, and Landaluze).17 The other strong visual culture of slavery was—not 14  Francisco Arango y Parreño, “Representación hecha a S.M. con motivo de la sublevación de los esclavos en los dominios de la Isla de Santo Domingo” (20 November 1791), in idem, Obras de D. Francisco de Arango y Parreño, 2 vols (Havana, 1952), I, pp. 111–112; Dale W. Tomich, “The ‘Second Slavery’: Bonded Labor and the Transformations of the NineteenthCentury World Economy”, in Francisco O. Ramírez (ed.), Rethinking the Nineteenth Century: Contradictions and Movement (New York, 1988), pp. 103–117; Dale W. Tomich, “The Wealth of the Empire: Francisco de Arango y Parreño, Political Economy, and the Second Slavery in Cuba”, Comparative Studies in Society and History, 1 (2003), pp. 4–28; idem, Through the Prism of Slavery: Labor, Capital, and World Economy (Boulder, CO [etc.], 2004). 15  Anselmo Suárez y Romero, Francisco. El ingenio o las delicias del campo (Las escenas pasan antes de 1838), prologue by Eduardo Castañeda (Havana, 1974); Martin Lienhard, “Afrokubanische Oralität und ihre Darstellung in ethnologischen und literarischen Texten”, in Ottmar Ette and Martin Franzbach (eds), Kuba heute. Politik Wirtschaft Kultur (Frankfurt, 2001), pp. 393–410. 16  José Antonio Saco, Historia de la esclavitud desde los tiempos más remotos hasta nuestros días, 3 vols, I and II (Paris, 1875), III (Barcelona, 1877/1878). 17  Justo G. Cantero, Los Ingenios. Colección de vistas de los principales ingenios de azúcar de la isla de Cuba. Edición de lujo. El texto redactado por Justo G. Cantero, gentil.hombre de la camara de S.M. y alferez real de Trinidad. Las laminas dibujadas del natural y litografiadas por Eduardo Laplante. Dedicado a la Real Junta de Fomento (Havana, 1857); facsimile: Luís Miguel García Mora and Antonio Santamaría García (eds), Los Ingenios. Colección de vistas de los principales ingenios de azúcar de la Isla de Cuba. El texto redactado por Cantero,

254

Zeuske

surprisingly—from Brazil (Rugendas, Debret, Thomas Ewbank, Maximilian Prinz zu Wied-Neuwied, Johann B. Spix, Carl F. Ph. Martius, Thomas Ender, Richard Bate, Carl Wilhelm von Theremin, and Eduard Hildebrandt). The first cultural history of slaves, arising from highly dubious beginnings in criminal ethnology at the start of the twentieth century (Cesare Lombroso), and the first “post-colonial” history of a society that owed its world-historical rise to slaves, the slave trade, and slavery, yet insults ex-slaves as all being “witches”, more or less, stems from the pen of Fernando Ortiz (1881–1969), the creator of the concept of transculturación that has become so important today in cultural history, post-colonialism, and subaltern studies.18 He was followed in the line of concealed transculturation by the essayist Miguel Barnet in 1966, with his prototype of testimonial literature that enjoyed worldwide success under the title Biografía de un Cimarrón—a white literature that speaks for blacks (a cultural technique of the actual-historical “transculturation” that existed also in the nineteenth century).19 Barnet’s “authentic” history was written against the broad background of the works of José Luciano Franco, Pedro Deschamps Chapeaux, and Juan Pérez de la Riva,20 as well as of the most important structural and social history of a concrete slavery society (El Ingenio, which basically only covered developments until Justo G. Con las láminas dibujadas del natural y litografiadas por Eduardo Laplante (Madrid, 2005); see Michael Zeuske, “Sklavenbilder: Visualisierungen, Texte und Vergleich im atlantischen Raum (19. Jahrhundert, Brasilien, Kuba und USA)”, Zeitenblicke, 7, no. 2, (1. 10. 2008), www. zeitenblicke.de/2008/2/zeuske; last accessed 28. 11. 2011. 18  Fernando Ortiz, Los negros brujos (apuntes para un estudio de etnología criminal). Carta prólogo del Dr C Lombroso (Madrid, 1906); idem, Hampa afro-cubana: Los negros esclavos. Estudio sociológico y de derecho public (Havana, 1916), new edn publ. as Los negros esclavos (Havana, 1976); idem, Contrapunteo cubano del tabaco y del azúcar (advertencia de sus contrastes agrarios, económicos, históricos y sociales, su etnografía y su transculturación), introduction by Bronislaw Malinowski (Havana, 1940); idem, “El fenómeno social de la transculturación y su importancia en Cuba”, Revista Bimestre Cubana, 66 (July–December 1940), pp. 273–278. 19  Miguel Barnet, Biografia de un cimarrón (Havana, 1966). 20  José Luciano Franco, Historia de la Revolución de Haití (Havana, 1966); idem, Los Palenques de los esclavos Cimarrones (Havana, 1973); idem, “Piratas, corsarios, filibusteros y contrabandistas, siglo XVIII y XIX”, in idem, Ensayos históricos (Havana, 1974), pp. 45–92; idem, Las Minas de Santiago del Prado y la rebelión de los Cobreros (Havana, 1975); idem, Las conspiraciones de 1810 y 1812 (Havana, 1977); idem, La diáspora africana en el Nuevo Mundo (Havana, 1978); idem, Comercio clandestino de esclavos (Havana, 1996), bibliography on pp. 283–285; idem, La presencia negra en el Nuevo Mundo (Havana, 1981); Pedro Chapeaux Deschamps, El negro en la economía habanera del siglo XIX (Havana, 1971); idem and Juan Pérez de la Riva, Contribución a la historia de gentes sin historia (Havana, 1974).

Historiography and Research Problems of Slavery

255

around 1860).21 The first volume was published in 1964 by Manuel Moreno Fraginals (1920–2001). Moreno Fraginals belonged to one of the schools of Cuban slavery research that, in contrast to Fernando Ortiz, was based on structures and demography (linked above all with the names of Ramiro Guerra y Sánchez, who also introduced the British-Caribbean and French-Caribbean concept of the “sugar revolution”, and Raúl Cepero Bonilla).22 In contrast to this, Barnet’s testimonial construction of the biography of a former slave and cimarrón, still readable today, gives the blacks a place in the national cycle of revolution, appreciating also their role in revolutionary leadership. On the whole, however, Cuban national history is still predominantly a “white” history,23 overlaid today by the one-sided dominance of the US and Brazilian historiographies of slavery.24 21  Manuel Moreno Fraginals, El Ingenio. Complejo económico social cubano del azúcar, 3 vols (Havana, 1978). 22  Ramiro Guerra y Sánchez, Azúcar y población en las Antillas (6th edn Havana, 1976, 1st edn Havana, 1927); Raúl Cepero Bonilla, Azúcar y abolición: apuntes para una historia crítica del abolicionismo (Havana, 1948); Barry W. Higman, “The Making of the Sugar Revolution”, in Alvin O. Thompson (ed.), In the Shadow of the Plantation: Caribbean History and Legacy (Kingston, 2002), pp. 40–71. 23  Alejandro de la Fuente, “Introduction”, in idem,”A Nation for All”: Race, Inequality, and Politics in Twentieth-Century Cuba (Chapel Hill, NC [etc.], 2001), pp. 1–19; Volker Mollin, “La problemática de la historiografía nacional cubana”, in idem, Guerra pequeña, guerra olvidada (Santiago de Cuba, 2002), pp. 52–72. 24  Gloria García Rodríguez, La esclavitud desde la esclavitud. La visión de los siervos (Havana, 1996); María del Carmen Barcia Zequeira, La otra familia. Parientes, redes y descendencia de los esclavos en Cuba (Havana, 2003); Gloria García Rodríguez, “Tecnología y abolición”, in José A. Piqueras (ed.), Azúcar y esclavitud en el final del trabajo forzado. Homenaje a M. Moreno Fraginals (Madrid [etc.], 2002), pp. 76–92; Gloria García Rodríguez, Conspiraciones y revueltas. La actividad política de los negros en Cuba (1790–1845) (Santiago de Cuba, 2003); idem, “Los cabildos de nación: organización, vicisitudes y tensiones internas (1789–1868)”, Del Caribe, 43 (2004), pp. 65–73; idem, “El despegue azucarero de Cuba: la versión de Arango y Parreño”, in Imilcy Balboa and José A. Piqueras (eds), La excepción americana. Cuba en el ocaso del imperio continental (Valencia, 2006), pp. 155–175; Mercedes García Rodríguez, “Ingenios habaneros del siglo XVIII”, in Consuelo Naranjo Orovio and Miguel Ángel Puig-Samper (eds), Las raíces históricas del pueblo cubano (Madrid, 1991), pp. 113–138; Mercedes García Rodríguez, “El monto de la trata hacia Cuba en siglo XVIII”, in Naranjo Orovio and Tomás Mallo Gutiérrez (eds), Cuba, la perla de la Antillas: Actas de las I Jornadas sobre “Cuba y su Historia”, pp. 297–312; Mercedes García Rodríguez, “La Compañía del Mar del Sur y el Asiento de esclavos de Cuba”, Santiago, 76 (1993), pp. 121–170; idem, Misticismo y capitales. La Compañía de Jesús en la economía habanera del siglo XVIII (Havana, 2000); idem, La aventura de fundar ingenios. La refacción

256

Zeuske

Not surprisingly, United States research into slavery society outside North America begins with Cuba, in the positivist book on the slave trade by Hubert H. S. Aimes (1907).25 Aimes also began, at a very early date, to do comparative work on slavery in the Americas.26 Other equally early works are those of Irene Aloha Wright,27 who, however, dealt with early colonial history. The works of Frederick Douglass28 and W. E. B. Dubois also appeared in the USA, as well as the oldest journal of “black history” (The Journal of Negro History, founded 1916). The 1940s, with the work of Fernando Ortiz and Melville J. Herskovits, saw the start of analysis of the African background of slavery and the “black family”, the “forgotten memories” of slaves as well as the “myth of the negro past”;29 somewhat later, Pierre Vergier began to study the connections of Bahia to Benin and the Gold Coast.30 But predominantly, even in the universities, the view of history remained marked by the racist works of Ulrich B. Phillips.31 This bore mainly on slavery in the US, absolutizing this as the “peculiar azucarera en Havana del siglo XVII (Havana, 2004); idem, Entre Haciendas y Plantaciones. Orígenes de la manufactura azucarera en Havana (Havana, 2007). 25  Hubert H. S. Aimes, A History of Slavery in Cuba (1511 to 1868) (New York, 1907; repr. New York, 1967). 26  Idem, “African Institutions in America”, Journal of American Folklore, 18 (January–March 1905), pp. 15–32. 27  Irene A. Wright, The Early History of Cuba, 1492–1586 (New York, 1916); idem, “El establecimiento de la industria azucarera en Cuba”, La Reforma social (April–June 1916), pp. 26–42. 28  Frederick Douglass, Narrative of the Life of Frederick Douglass, An American Slave, Written by Himself [1845], David W. Blight (ed.) (Boston, MA, 1993). 29  E. Franklin Frazier, “The Negro Family in Bahia, Brazil”, American Sociological Review, 7 (1942), pp. 465–478; Melville J. Herskovits, “The Negro in Bahia, Brazil: A Problem in Method”, American Sociological Review, 8 (1943), pp. 394–402; idem, The Myth of the Negro Past (New York, 1941; new edn Boston, MA, 1990); on this debate, see Sidney W. Mintz and Richard Price, The Birth of African-American Culture: An Anthropological Perspective (Boston, MA, 1992). 30  Nina S. Rodrigues, O Animismo Fetichista dos Negros Bahianos (Rio de Janeiro, 1932); Pierre Verger, Bahia and the West African Trade, 1549–1851 (Lagos, 1964); idem, Flux et reflux de la traite des nègres entre le golfe de Bénin et Bahia de Todos od Santos du XVII e au XIX e siècle (Paris, 1963); idem, Fluxo e refluxo do tráfico de escravos entre o golfo do Benin e a Bahia de Todos os Santos dos séculos XVII a XIX (São Paulo, 1987); idem, Os Libertos: sete caminhos na liberdade de escravos da Bahia no século XIX, Salvador (Bahia, 1992). 31  Ulrich B. Phillips, “The Economic Cost of Slaveholding in the Cotton Belt”, Political Science Quarterly, 20 (June 1905); pp. 257–275; idem, “Slave Crime in Virginia”, American Historical Review, 20 (January 1915), pp. 336–340; on Philipps’s position in the historiography of slavery before Time On the Cross, see Daniel C. Littlefield, “From Phillips to Genovese: The Historiography of American Slavery Before Time On The Cross”, in Wolfgang Binder (ed.), Slavery in the Americas (Erlangen, 1993), pp. 1–23.

Historiography and Research Problems of Slavery

257

institution”. Only with the extremely influential works of Frank Tannenbaum and Stanley Elkins32 did a new phase comparing the American and the global history of slavery set in (with the influential works of the Genoveses and Anglo-American historians),33 and with the myth of the “mildness of IberoAmerican slavery” (compared with its pathological severity in the USA) partly persisting through to today.34

New Perspectives on Slavery

Somewhat outside the mainstream, beginning also in the 1950s, were anthropological studies in the Caribbean, their main protagonists being Sidney W. Mintz and Eric Wolf, and researchers of slave resistance (like Richard Price and João Reis).35 Both of these conducted empirical research in the field. They studied problems of rural populations (initially in Puerto Rico)36 that lived in former slavery societies. The publications of Sidney Mintz, above all, produced a plethora of stimuli, methodological concepts, and theoretical reflections, which in a certain sense founded an anthropological history of slavery “from below”, from the perspective of slaves and former slaves as actors in a transnational history of the Caribbean, against the background of a pronounced historical translocality (without using this concept).37 Mintz’s work continues 32  Frank Tannenbaum, Slave and Citizen: The Negro in the Americas (New York, 1946; repr. Boston, MA, 1992); Stanley M. Elkins, Slavery: A Problem in American Institutional and Intellectual Life (Chicago, IL, 1959); on this debate, see Peter J. Parish, Slavery: History and Historians (New York, 1985). 33  Elizabeth Fox-Genovese and Eugene D. Genovese, Fruits of Merchant Capital: Slavery and Bourgeois Property in the Rise and Expansion of Capitalism (New York, 1983); for example, Stanley L. Engerman, “Slavery in World Perspectives”, in idem, Slavery, Emancipation & Freedom: Comparative Perspectives (Baton Rouge, LA, 2007), pp. 1–36; Seymour Drescher, Abolition: A History of Slavery and Antislavery (Cambridge, 2009). 34  Alejandro de la Fuente, “Slave Law and Claims-Making in Cuba: The Tannenbaum Debate Revisited”, Law and History Review, 22 (2004), pp. 339–369, www.press.ullinois .edu/journals/ lhr/DLF22_2.pdf (28 June 2004). 35  Richard Price (ed.), Maroon Societies: Rebel Slave Communities in the Americas (Baltimore, MD, 1979); João Reis, Rebelião escrava no Brasil A história do levante dos malê em 1835 (São Paulo, 2003). 36  Sidney Mintz, Worker in the Cane: A Puerto Rican Life History (New York [etc.], 1974; 1st edn 1960). 37  A selection: Sidney Mintz and Eric R. Wolf, “An Analysis of Ritual Co-Parenthood (Compadrazgo)”, Southwestern Journal of Anthropology, 6 (1950), pp. 341–365; Sidney Mintz, “The Caribbean as a Socio-Cultural Area”, Journal of World History, 9 (1966),

258

Zeuske

to play an important role today, in debates over slave agency, creolity, and centralisms, particularly between Americanists and Africanists (Mintz and Price versus Thornton und Lovejoy etc.).38 Since then, empirically and anthropologically oriented research has taken a mediating standpoint, opening on the one hand on to slave voices and the archaeology of the lives of slaves, the slave trade from an under-deck perspective, and the African diaspora along the line Africa–Atlantic–America, as well as the archaeology of the slave trade in West Africa, and on the other hand to a stronger visualization: (www.slaveryimages.org; Jerome S. Handler; Jane G. Landers; Christopher DeCorse etc.).39 There have also been studies of historiography and the history of ideas of slavery, starting, above all, with the works of Elsa W. Govea and David Brion Davis.40 Under these conditions, research, historical sciences, and historical social sciences in the USA responded to the challenges of Caribbean “postcolonialism before postcolonialism” (especially made by Fernando Ortiz and Eric pp. 912–937; idem, “Afro-Caribbeana: An Introduction”, in idem, Caribbean Transformations (Chicago, IL, 1974), pp. 1–42; Sidney Mintz and Sally Price (eds), Caribbean Contours (Baltimore, MD [etc.], 1985); Sidney Mintz, Die süße Macht. Kulturgeschichte des Zuckers (Frankfurt [etc.], 1986); Sidney Mintz and Richard Price, The Birth of African-American Culture: An Anthropological Perspective (Boston, MA, 1992); Sidney Mintz, Tasting Food, Tasting Freedom: Excursions into Eating, Power, and the Past (Boston, MA, 1996); Walton Look Lai, Indentured Labor, Caribbean Sugar: Chinese and Indian Migrants to the British West Indies, 1838–1918, with an introduction by Sidney W Mintz (Baltimore, MD [etc.], 1993); Sidney Mintz, “The Localization of Anthropological Practice: From Area Studies to Transnationalism”, Critique of Anthropology, 18 (1998), pp. 117–133; Eric R. Wolf, Peasant Wars in the Twentieth Century (New York, 1969); idem, Europe and the People without History (Los Angeles, CA, [etc.], 1982). 38  Richard Price, “The Miracle of Creolization”, in Kevin A. Yelvington (ed.), Afro-Atlantic Dialogues. Anthropology in the Diaspora (Santa Fe, NM [etc.], 2006), pp. 115–147. 39  Jerome S. Handler, “Life Histories of Enslaved Africans in Barbados”, Slavery & Abolition, 19 (1998), pp. 129–141; idem, “Survivors of the Middle Passage: Life Histories of Enslaved Africans in British America”, Slavery & Abolition 23 (2002), pp. 25–56; idem, “The Middle Passage and the Material Culture of Captive Africans”, Slavery & Abolition, 30 (2009), pp. 1–26; Christopher DeCorse, West Africa During the Atlantic Slave Trade: Archaeological Perspectives (London, 2001); idem, An Archaeology of Elmina: Africans and Europeans on the Gold Coast, 1400–1900 (Washington DC, 2001); Jane G. Landers and Barry Robinson (eds), Slaves, Subjects, and Subversives: Blacks in Colonial Latin America (Albuquerque, NM, 2006). 40  Elsa V. Goveia, A Study on the Historiography of the British West Indies to the End of the Nineteenth Century (Mexico City, 1956); David Brion Davis, The Problem of Slavery in Western Culture (Ithaca, NY, 1966).

Historiography and Research Problems of Slavery

259

Williams).41 As a background, the works of economic history and quantitative studies also need to be mentioned, these likewise having their origins in the late 1950s,42 and later culminating in Time on the Cross,43 in which it was shown from a standpoint of neoclassical economics that slaves in the USA worked more efficiently than free workers, and were relatively well provided for in terms of diet and medical care, without which it would be impossible to explain the growth of the slave population from around 400,000 in 1808 to something like 4 million in 1865, in the absence of the Atlantic slave trade (even though, in this timeframe, we do not really know how many slaves were smuggled into the USA from the Caribbean, and first of all from Cuba after 1820).44 The line of quantitative research, without Fogel’s neoclassical series of assumptions, continues in what is today the most important social-historical data base on the Atlantic slave trade, under the direction of David Eltis, Herbert Klein, and David Richardson: Trans-Atlantic Slave Trade Database; www .slavevoyages.org.45 The intensive “slavery debate”, particularly in the USA, which was initially comparative and then also quantifying and social-historical,46 has seen in the last forty years the introduction of computerized methods into the historical and social sciences. Together with the new approaches of Mintz and 41  Eric Williams, Capitalism and Slavery (London, 1964, original edn, 1944); Barbara L. Solow, “Caribbean Slavery and British Growth: The Eric Williams Hypothesis”, Journal of Developmental Economics, 17 (1985), pp. 99–115; idem, “Capitalism and Slavery in the Exceedingly Long Run”, in idem and Stanley L. Engerman (eds), British Capitalism and Caribbean Slavery: The Legacy of Eric Williams (Cambridge, 1987). 42  Alfred H. Conrad and John R. Meyer, “The Economics of Slavery in the Ante Bellum South”, Journal of Political Economy, 6 (1958), pp. 95–130. 43  Robert William Fogel and Stanley Engerman, Time on the Cross: The Economics of American Negro Slavery (New York [etc.], 1995, 1st edn 1974); Robert William Fogel, Without Consent or Contract: The Rise and Fall of American Slavery, 4 vols (New York, 1989); idem, The Slavery Debates, 1952–1990: A Retrospective (Baton Rouge, LA, 2003). 44  Leonardo Marques, “A participação norte-americana no tráfico transatlântico de escravos para os Estados Unidos, Cuba e Brasil”, Historia: Questões & Debates, 52 (January–July 2010), pp. 91–111. 45  David Eltis and David Richardson (eds), Extending the Frontiers: Essays on the New Transatlantic Slave Trade Database (New Haven, CT, 2008); Michael Zeuske, “Out of the Americas: Sklavenhändler und Hidden Atlantic im 19. Jahrhundert. Ein Forschungsprojekt am Historischen Seminar der Universität zu Köln”, AHF Jahrbuch der historischen Forschung in der Bundesrepublik Deutschland, (2009), pp. 37–57, www.ahf-muenchen.de/ Forschungs berichte/Jahrbuch2009/AHF_Jb2009_Zeuske.pdf. 46   Ulrike Schmieder, “War die iberoamerikanische Sklaverei mild?”, Zeitschrift für Weltgeschichte, 4 (2003), pp. 115–132.

260

Zeuske

others, research into slaves and slavery has seen a blossoming of comparative research and new approaches from the side of culturalist-oriented world and global historiography, as well as a series of new research orientations, such as “racism and slavery”,47 “transculturation, diaspora/migration history”,48 including trans-Atlantic construction of identities, ethnicities, and Atlantic creoles,49 “comparative history of slavery”,50 “history from below”, not only of slavery as an institution but first of all based on slave voices, life histories, and

47  Starting with Stanley L. Engerman and Eugene D. Genovese (eds), Race and Slavery in the Western Hemisphere: Quantitative Studies (Princeton, NJ, 1975). 48  Ronald Segal, The Black Diaspora (London [etc.], 1995); idem, Islam’s Black Slaves: The Other Diaspora (New York, 2001); Isidore Okpewho, Carole Boyce Davies, and Ali A. Mazrui, (eds), African Diaspora: African Origins and New World Identities (Bloomington, IN, 1999); David Eltis (ed.), Coerced and Free Migration: Global Perspectives (Stanford, CA, 2002); Dirk Hoerder, Cultures in Contact: World Migrations in the Second Millennium (Durham, NC, 2002); Harald Kleinschmidt, “Early Forms of Colonialism and Trans-Atlantic Slave Trade”, in idem, People on the Move: Attitudes Toward and Perception of Migration in Medieval and Modern Europe (Westport, CT, 2003), pp. 150–159; Marcus Rediker, The Slave Ship: A Human History (New York, 2007); Eric Robert Taylor, If We Must Die: Shipboard Insurrections in the Era of the Atlantic Slave Trade (Baton Rouge, LA, 2006); Stephanie E. Smallwood, Saltwater Slavery: A Middle Passage from Africa to American Diaspora (Cambridge, MA, 2007); Emma Christopher, Slave Ship Sailors and Their Captive Cargoes, 1730–1807 (Cambridge, 2006); idem, Cassandra Pybus, and Marcus Rediker (eds), Many Middle Passages: Forced Migration and the Making of the Modern World (Berkeley, CA, 2007); Beatrix Heintze and Achim von Oppen (eds), Angola on the Move: Transport Routes, Communications and History/Angola em Movimento. Vias de Transporte, Comunicação e Histórica (Frankfurt, 2008); Wim Klooster (ed.), Migration, Trade, and Slavery in an Expanding World: Essays in Honor of Pieter Emmer (Leiden, 2009); Patrick Manning, The African Diaspora: A History Through Culture (New York, 2009). 49  Thornton, Africa and the Africans; Paul E. Lovejoy (ed.), Identity in the Shadow of Slavery (London, 2000); Toyin Falola and Matt Childs (eds), The Yoruba Diaspora in the Atlantic World (Bloomington, IN, 2004); Linda Heywood (ed.), Central Africans and Cultural Transformations in the American Diaspora (Cambridge, 2002); Thornton Heywood, Central Africans, Atlantic Creoles, and the Foundations of the Americas, 1585–1660, (Cambridge, 2007). 50  Herbert S. Klein, Slavery in the Americas: A Comparative Study of Virginia and Cuba (Chicago, IL, 1967); Franklin W Knight, Slave Society in Cuba During the Nineteenth Century (Madison, WI, 1970); Michael Zeuske, “The Names of Slavery and Beyond: The Atlantic, the Americas and Cuba”, in Ulrike Schmieder, Katja Füllberg-Stolberg, and Michael Zeuske (eds), The End of Slavery in Africa and the Americas: A Comparative Approach (Münster [etc.], 2011), pp. 51–80; Enrico Dal Lago, “Comparative Slavery”, in Paquette and Smith, Oxford Handbook of Slavery in the Americas, pp. 664–684.

Historiography and Research Problems of Slavery

261

experiences, including “abolition and post-emancipation”,51 “social history of law and slavery”,52 “social history of medicine” (and sciences in general),53 as well as international history, microhistory, in particular Atlantic history, history of the seas and migrations,54 history of slaveries and creolizations in different spaces of the Indian Ocean,55 or “translocal/transnational” cultural history,56 with the variants of “new” imperial history,57 histories of diasporas, and translocal south-south history.58 More recently, the themes “religion and slavery” (if the slavers were only interested in the bodies of the enslaved, slaves in their agency transcended the individualities of the dead and the ubiquity of death),59 “women, children,

51  Rebecca J. Scott, Slave Emancipation in Cuba: The Transition to Free Labor, 1860–1899 (Princeton, NJ, 1985; repr. Pittsburgh, 2000); Michael Zeuske, Schwarze Karibik. Sklaven, Sklavereikulturen und Emanzipation (Zurich, 2004). 52  Lauren Benton, “Law in Diaspora: The Legal Regime of the Atlantic World”, in idem, Law and Colonial Cultures: Legal Regimes in World History, 1400–1900 (Cambridge, 2002), pp. 31–79. 53  Steven Palmer, “From the Plantation to the Academy”, in David Wright, Juanita Barros, and Steven Palmer (eds), Health and Medicine in the Circum-Caribbean, 1800–1968 (New York, 2009), pp. 53–75; Vincent Brown, The Reaper’s Garden: Death and Power in the World of Atlantic Slavery (Cambridge, 2008). 54  Rediker, The Slave Ship; Taylor, If We Must Die; Smallwood, Saltwater Slavery; Christopher, Slave Ship Sailors and Their Captive Cargoes; idem, Pybus, and Rediker, Many Middle Passages. 55  Ulbe Bosma and Remco Raben, Being “Dutch” in the Indies: A History of Creolisation and Empire, 1500–1920 (Athens, OH, 2008); Michael Mann, Sklaverei und Sklavenhandel im Indik, 16. bis 20. Jahrhundert (Leipzig, 2009) (Working Paper Series, No. 3), at www.unileipzig.de/ral/gchuman/documents/working_paper_series/RAL_WP_3_Mann_web.pdf (last accessed 31 October 2011); Richard B. Allen, “Suppressing a Nefarious Traffic: Britain and the Abolition of Slave Trading in India and the Western Indian Ocean, 1770–1830”, William & Mary Quarterly, 66 (2009), pp. 873–894; Tatiana Seijas, “The Portuguese Slave Trade to Spanish Manila, 1580–1640”, Itinerario: International Journal on the History of European Expansion and Global Interaction, 32 (2008), pp. 19–38. 56  Rebecca J. Scott, Degrees of Freedom: Louisiana and Cuba after Slavery (Cambridge, MA [etc.], 2005). 57  Jane Burbank and Frederick Cooper, Empires in World History: Power and the Politics of Difference (Princeton, NJ [etc.], 2010). 58  Susanne Freitag, “Translokalität als ein Zugang zur Geschichte globaler Verflechtungen”, http://geschichte-transnational.clio-online.net/forum/2005-06-001 (11 June 2005). 59  Brown, The Reaper’s Garden; Joel James Figarola, La muerte en Cuba (Havana, 1999); idem, Los sistemas mágico-religiosos cubanos: principios rectores (Caracas, 1999).

262

Zeuske

and slavery”,60 and “gender and slavery”61 have also begun to play major roles, the high point of research so far (resulting in new criteria of analysis for a genuinely global history of slavery from today’s perspective) being the two volumes of Women and Slavery and Children in Slavery. The appearance of Joseph Miller’s article of synthesis on this subject is also not accidental.62 A relatively new theme, in part influenced (but not solely) by the distortions of the perception history complained of above, is that of “slavery and memory”.63 In the USA and the Anglo-American language zone including Canada, ever more works are appearing that open to Atlantic and other historiographies, conducting histoire croisée, the history of workers (pirates, sailors, slaves), the transcultural roots of North America, and the importance of the Haitian revolution (1791– 1803) for the world history of slaves, thematizing freedom and the Atlantic.64 This has led in turn to new assessments of slaveries and the slave trade, as well as of transitional forms in Africa and in world or global history (coolies

60  Claire C. Robertson and Martin A. Klein (eds), Women and Slavery in Africa (Madison, WI, 1983); Roger Sawyer, Children Enslaved (London [etc.], 1988); Gwyn Campbell, Suzanne Miers, and Joseph C. Miller (eds), Women and Slavery, 2 vols (Athens, OH, 2007–2008); idem (eds), Children in Slavery through the Ages (Athens, OH [etc.], 2009). 61  Diana Paton, “Decency, Dependence, and the Lash: Gender and the British Debate over Slave Emancipation, 1830–1834”, Slavery & Abolition, 17 (1996), pp. 162–184; idem, “Punishment, Crime, and the Bodies of Slaves in Eighteenth-Century Jamaica”, Journal of Social History, 34 (2001), pp. 923–954; idem and Pamela Scully, “Introduction: Gender and Slave Emancipation in Comparative Perspective”, in Pamela Scully and Diana Paton (eds), Gender and Slave Emancipation in the Atlantic World (Durham, NC, 2005), pp. 1–34; Diana Paton, “Bibliographic Essay”, in Scully and Paton, Gender and Slave Emancipation, pp. 328–356; Ulrike Schmieder, Geschlecht und Ethnizität in Lateinamerika im Spiegel von Reiseberichten: Mexiko, Brasilien und Kuba 1780–1880 (Stuttgart, 2003). 62  Joseph C. Miller, “Domiciled and Dominated: Slavery as a History of Women”, in Campbell, Miers, and Miller, Women and Slavery, II: The Modern Atlantic, pp. 284–312. 63  Ira Berlin, “American Slavery in History and Memory and the Search for Social Justice”, Journal of American History, 90 (2004), pp. 1251–1268; Stefan Palmié, “Slavery, Historicism, and the Poverty of Memorialization”, in Susannah Radstone and Bill Schwartz (eds), Memory: Histories, Theories, Debates (New York, 2010), pp. 363–375; Ana Lucia Araujo (ed.), Paths of the Atlantic Slave Trade: Interactions, Identities, and Images (Amherst, MA, 2011). 64  Dale Tomich, Through the Prism of Slavery: Labor, Capital, and World Economy (Boulder, CO [etc.], 2004); Taylor, If We Must Die; Smallwood, Saltwater Slavery; Christopher, Slave Ship Sailors and Their Captive Cargoes; Tomich and Zeuske, The Second Slavery; Laurent Dubois, Laurent and Julius S. Scott (eds), Origins of the Black Atlantic (New York, 2010).

Historiography and Research Problems of Slavery

263

mostly from China and India)65 also of the eastern hemisphere (especially the Indian Ocean and India, as well as the Arab-Islamic zone).66 New impulses for world and global history come especially from Africanists and Brazilianists, including the Atlantic dimension and different cosmologies;67 from the new 65  On the historical context of the trade in Chinese coolie slaves to America, see Lin Yun, “Historical Context of Coolie Traffic to the Americas”, in idem, The Coolie Speaks: Chinese Indentured Laborers and African Slaves in Cuba (Philadelphia, PA, 2008), pp. 1–35; for the historiography, see Juan Jiménez Pastrana, Los chinos en la liberación de Cuba (Havana, 1963); idem, Los Chinos en la historia de Cuba: 1847–1930 (Havana, 1983); Juan Pérez de la Riva, “El viaje a Cuba de los culiés chinos”, in Chapeaux Deschamps and Pérez de la Riva, Contribución a la historia de gentes sin historia, pp. 191–213; Juan Pérez de la Riva, “El tráfico de culiés chinos”, in ibid., pp. 215–232; see also “Informe del señor D. Francisco Diago a la Real Junta de Fomento sobre el proyecto de inmigración china”, in ibid., pp. 219–223; Jiménez Pastrana, Los Chinos en la historia de Cuba: 1847–1930 (Havana, 1983); William Gervase Clarence-Smith, “The Portuguese Contribution to the Cuban Slave and Coolie Trades in the Nineteenth Century”, Slavery & Abolition, 5 (1984), pp. 25–33; Evelyn Hu-DeHart, “Chinese Coolie Labour in Cuba in the Nineteenth Century: Free Labour or Neo-Slavery?”, Slavery & Abolition, 14 (1993), pp. 67–86; The Cuba Commission Report, A Hidden History of the Chinese in Cuba: The Original English-Language Text, Johns Hopkins Studies in Atlantic History and Culture, introduction by Denise Helly (Baltimore, MD [etc.], 1993); José Baltar Rodríguez, Los chinos de Cuba. Apuntes etnográficos (Havana, 1997); Evelyn Hu-DeHart, “Chinese in Cuba”, in Lingchi Wang and Gungwu Wang (eds), The Chinese Diaspora (Singapore, 1998); Naranjo Orovio and Imilcy Balboa Navarro, “Colonos asiáticos para una economía en expansión: Cuba, 1847–1880”, Revista Mexicana del Caribe, 8 (1999), pp. 32–65; Nadia Fernández de Pinedo Echevarría, “Chinos y yucatecos”, in idem, Comercio exterior y fiscalidad: Cuba 1794–1860 (Bilbao, 2002), pp. 222–224; E. Hu-DeHart, “Opium and Social Control: Coolies on the Plantations of Peru and Cuba”, Journal of Chinese Overseas, 1 (2005), pp. 169–183; Eduardo Marrero Cruz, “Traficante de esclavos y chinos”, in Marrero Cruz, Julián de Zulutea y Amondo. Promotor del capitalismo en Cuba (Havana, 2006), pp. 46–79; Kathleen López, “Afro-Asian: Marriage, Godparentage, and Social Status in Late-Nineteenth Cuba”, Afro-Hispanic Review, 27 (2008), pp. 59–72; on coolie traffic from a macrohistorical perspective, see Christopher, Pybus, and Rediker, Many Middle Passages. 66  William Gervase Clarence-Smith (ed.), The Economics of the Indian Ocean Slave Trade in the Nineteenth Century (London, 1977); Gwyn Campbell (ed.), The Structure of Slavery in Indian Ocean Africa and Asia (London [etc.], 2004). 67   Paul E. Lovejoy and Robin Law, “The Changing Dimensions of African History: Reappropriating the Diaspora”, in Simon McGrath, Charles Jedrej, Kenneth King, and Jack Thompson (eds), Rethinking African History (Edinburgh, 1997), pp. 181–200; Thornton, Africa and the Africans, passim; Alencastro, O Trato dos Viventes; Curto and Lovejoy, Enslaving Connections, passim; Paul E. Lovejoy, Transformations in Slavery: A History of Slavery in Africa (Cambridge, 2000); idem (ed.), Slavery on the Frontiers of Islam (Princeton, NJ, 2004).

264

Zeuske

historiography of workers (Boutang, Linebaugh, Rediker, Van der Linden, Lucassen, and Brass);68 while especially from Indian Marxist historians there has been a new assessment of global labour relationships including forms of slavery existing through to today.69 Histories of slavery have been supplemented by a body of economic, political, and sociological studies of “labour compelled by force” (= slavery), or “free and unfree labour under conditions of globalization”, that arose in parallel with the new cultural paradigm of the “black Atlantic” (Paul Gilroy)70 In general, slaves were and are classed by their contemporaries in the lowest and most dishonourable rank of society, expressed above all in violence against bodies and control over the bodies of the enslaved. Even when they have assumed quite high positions as domestic slaves, soldiers, or luxury slaves, their position has always been extremely precarious, “radically uncertain”, as Brent D. Shaw formulated it.71 This is basically also what the American 68  Jan Lucassen, “Free and Unfree Labour before the Twentieth Century: A Brief Overview”, in Tom Brass, Marcel van der Linden, and Jan Lucassen (eds), Free and Unfree Labour (Amsterdam, 1993), pp. 7–18; Tom Brass and Marcel van der Linden (eds), Free and Unfree Labour: The Debate Continues (Bern [etc.], 1997); Yann Moulier-Boutang, De l’esclavage au salariat. Economie historique du salariat bride (Paris, 1998); Marcel van den Linden, “Zur Sozialgeschichte des revolutionären Atlantiks”, Zeitschrift für Weltgeschichte, 10 (2009), pp. 159–169. 69   Peter Linebaugh and Marcus Rediker, The Many-Headed Hydra: Sailors, Slaves, Commoners, and the Hidden History of the Revolutionary Atlantic (London, 2000); Rediker, The Slave Ship; Christopher, Pybus, and Rediker, Many Middle Passages; Marcel van der Linden, Transnational Labour History: Explorations (Aldershot, 2003); Karl-Heinz Roth and Marcel van der Linden, “Karl Marx und das Problem der Sklavenarbeit”, in Marcel van der Linden and Karl-Heinz Roth (eds), with the collaboration of Max Henninger, Über Marx hinaus. Arbeitsgeschichte und Arbeitsbegriff in der Konfrontation in den globalen Arbeitsverhältnisse des 21. Jahrhunderts (Berlin [etc.], 2009), pp. 581–586. 70  Paul Gilroy, The Black Atlantic: Modernity and Double Consciousness (London, 1993); Michael Mann, “Die Mär von der freien Lohnarbeit. Menschenhandel und erzwungene Arbeit in der Neuzeit. Ein einleitender Essay”, in idem (ed.), Menschenhandel und unfreie Arbeit, (Leipzig, 2003), also Comparativ, 13:4 (2003), pp. 7–22; Christine Chivallon, “Beyond Gilroy’s Black Atlantic: The Experience of African Diaspora”, Diaspora: A Journal of Transnational Studies 3:11 (2002), pp. 359–382; Dale W. Tomich, “Atlantic History and World Economy: Concepts and Constructions”, Proto Sociology: An International Journal of Interdisciplinary Research, 20 (2004), pp. 102–121. 71  Brent D. Shaw, “‘A WOLF BY THE EARS’: M. I. Finley’s Ancient Slavery and Modern Ideology in Historical Context”, in Moses I. Finley, Ancient Slavery and Modern Ideology, expanded edn, Brent D. Shaw (ed.) (Princeton, NJ, 1998), pp. 3–74 (latest edn, with an excellent introduction).

Historiography and Research Problems of Slavery

265

sociologist Orlando Patterson meant by “ultimative slave” and “social death”.72 Patterson’s cultural-sociological concept of “social death” is particularly interesting in this connection, and particularly contested.73 It holds that slaves were not only torn away from their community of origin, but also as members of the same group as their master or mistress they were non-persons in the enslaving society. They had no “honour” and were not “persons”, as defined by the community in which they were forced to live. They had no “freedoms”, in the sense of the European medieval discourse, or at least only very few, and even these were in general not formally defined. They were thus excluded from normal life, and formed almost a kind of social zombie condition, the undead dead.74 Their position was bound up with shame, dishonour, insecurity, lack of “freedoms” and loss of status, fatherlessness, as well as such degrading characteristics as slave names. Anthropologists, historians and social scientists have developed other models of slavery (e.g. Herman Jeremias Nieboer, Moses I. Finley, Claude Meillassoux, Paul E. Lovejoy, Albert Wirz, S. Fenoaltea, Martin A. Klein, Joseph E. Inikori, David Eltis, and John K. Thornton),75 or else used no explicit models 72  Orlando Patterson, Slavery and Social Death: A Comparative Study (Cambridge, MA, 1982). 73  David Brion Davis, “The Ancient Foundations of Modern Slavery”, in idem, Inhuman Bondage: The Rise and Fall of Slavery in the New World (New York, 2006), pp. 27–47, 30–32. 74  Leonhard Schumacher, Sklaverei in der Antike. Alltag und Schicksal der Unfreien (Munich, 2001), p. 59. 75  Herman J. Nieboer, Slavery as an Industrial System: Ethnological Researches (The Hague, 1900; 2nd expanded edn 1910; repr. New York, 1971); Claude Meillassoux, Anthropologie de l’esclavage—le ventre de fer et le argent (Paris, 1986), p. 20. Meillassoux’s work is intended as a theoretical (and thus very systematic) essay on the institution of slavery, on the basis of the author’s knowledge of the historical Sudan and West Africa in general. Paul E. Lovejoy has spoken of slavery in Africa as “a marginal feature of society”, “slavery as an institution”, and “slavery as a mode of production”, in the sense of the graduated dependence of different African societies on slavery. The African forms of slavery would in his view have been “transformed” under the pressure of the European slave trade (“transformation thesis”). By and large, there was supposedly in Africa, where enslavement took place, slavery as an institution, in America slavery as a system. See Lovejoy, Transformations in Slavery. Albert Wirz speaks of plantation slavery, lineage, and domestic slavery; see A. Wirz, Sklaverei und kapitalistisches Weltsystem (Frankfurt, 1984), pp. 14 ff.; S. Fenoaltea, “Slavery and Supervision in Comparative Perspective: A Model”, Journal of Economic History, 44 (1984), pp. 635–668; Martin A. Klein, “Introduction: Modern European Expansion and Traditional Servitude in Africa and Asia”, in idem, Breaking the Chains: Slavery, Bondage, and Emancipation in Modern Africa and Asia (Madison, WI, 1993), pp. 3–36; Joseph E. Inikori and Stanley L. Engerman (eds), The Atlantic Slave Trade: Effects on Economies, Societies, and Peoples in Africa, the Americas, and Europe (Durham,

266

Zeuske

(Georges Scelle,76 and Charles Verlinden77). Werner Scheidel has repeatedly indicated the characteristics of “slave economies” as distinct from Finley’s concept of “slave society”, and discussed various models of relationship between societies and slaveries.78 In some cases, models of slavery start from a large number of enslaved individuals and their fundamental importance for the economy, social structures, and social processes, in part also for the mentality and social psychology of most societies in world history. In the most important new global-historical approach, Joseph C. Miller has attempted to combine all the aspects of slavery (the capture, trade and transport of slaves, as well as the different sectors of slavery) under the concept of “slaving”, also in order to dynamize the extreme structuralizing of the concepts of “slavery systems” or Finley’s “slave societies” and “societies with slaves”.79 Miller presents tables on epochs of slaving, stretching from prehistory into the future, and appends to these three global-historical tables on “novelties in the Atlantic”, designed to emphasize what was specific about Atlantic slavery.80 Of particular weight for a world and global history of slavery is Moses I. Finley’s attempt, which as I said, I greatly respect, to elaborate the absoluteness of antique slavery vis-à-vis other forms of forced and involuntary labour.81 Many of Finley’s arguments still deserve consideration today, as Brent D. Shaw also stresses in his introduction to Finley’s works. One thing, however, is missing: it is always historians of classical antiquity who emphasize the idea of the exceptionality of their object of investigation. Two of Finley’s most important arguments are that of the “individuality” of enslavement and that of the inadequate reproduction of slave groups. The reality of the individuality of enslavement as a common basis of all conceivable types of slavery and unfree labour NC [etc.], 1992); Thornton, Africa and the Africans, passim. Thornton’s speciality is the historical Congo and the Atlantic dimension of the African diaspora; David Eltis, The Rise of African Slavery in the Americas (Cambridge, 1999). 76  Georges Scelle, La traite négrière aux Indes de Castille: Contrats et traités d’assiento, 2 vols (Paris, 1905–1906). 77  Charles Verlinden, L’Esclavage dans l’Europe medieval, 2 vols (Bruges, 1955 and Gand, 1977). 78  Walter Scheidel, “The Comparative Economics of Slavery in the Greco-Roman World”, in Enrico Dal Lago and Constantina Katsari (eds), Slave Systems: Ancient and Modern (Cambridge [etc.], 2008), pp. 105–126. 79  Joseph C. Miller, “Slaving as Historical Process: Examples from the Ancient Mediterranean and the Modern Atlantic”, in Dal Lago and Katsari, Slave Systems, pp. 70–102. 80  Ibid., pp. 100–102. 81  M. I. Finley, “The Emergence of a Slave Society”, in idem, Ancient Slavery, pp. 135–160, 137–145.

Historiography and Research Problems of Slavery

267

still requires further discussion and analysis—particularly in terms of world history. My own position is rather that of an actual similarity of basic characteristics of coerced and unfree labour as defined on the basis of slaveries (and of their historic individualities in particular contexts), above all because only this concept makes it possible for us to conceptualize in an integrative way the types and forms of slaveries in world and global history—from the perspective of the twenty-first century. The “individuality” of antique slavery, in view of the mass enslavement of women, children, and prisoners of war in kin-slaveries and worldwide with the rise of other slaveries “without the name of slavery”, and in cultures with different norms, is rather a product on the one hand of the process of the sharper characterization of antique slave status in Roman law, and on the other hand of the reconstruction of Roman law in the thirteenth century against the background of the Italian slave trade (Genoa and Venice), above all along the line “from the Mongols to the Mamelukes”, but also of the global spread of Roman law with the European expansion and the “new” Atlantic slavery from c. 1300– 1650. One position seems to be clearer and clearer with ongoing research: the “new” Atlantic slavery raids in Africa, the peripheral Iberians, and later other Europeans, used the forces of transculturation/creolization and a technological complex around their ocean-going ships to overtake and control the Middle Passage.82 The African and Atlantic slave trade, however, as well as Atlantic slaveries in modern times, are at the same time far more. In the rise of the “creole space” (the Atlantic, ocean shores and coastal zones, but also used for places around the Indian Ocean), as well as of trans-local and transimperial spaces of slavery, “before the nation”, as it were, and also in the fate of those affected, we see the founding violence, zones of influence, the spirit of the time, underlying structures, languages, and the beginnings of globalizations; slaveries, and especially the “new” Atlantic slavery and globalization, also show the “meaning” of early colonial expansions. It is with the slave trade and slaveries that the diasporas of African people in the Atlantic world (and beyond) arose. The “creole space”83 of Atlantization and early capital formation was marked by the most varied creole languages, displaying a common feature of “African” basic structures and European vocabulary. The most important Atlantic bearers of the culture of the trade in humans and the Atlantization of Africa, Europe, and the Americas were Atlantic creoles 82  Tobias Green, The Rise of the Trans-Atlantic Slave Trade in Western Africa, 1300–1589 (Cambridge [etc.], 2011). 83  Angela Bartens, Der kreolische Raum: Geschichte und Gegenwart (Helsinki, 1996).

268

Zeuske

even “beyond the Atlantic” (by and large, in the first generation, descendants of European fathers and indigenous mothers, whose families were generally in charge). In the early period of Atlantization (1300–1580), Atlantic creoles (tangomaõs and lançados) played a leading role in trade in humans, particularly including the early transport of slaves to America after 1493. They had to deal very soon with the counter-reactions of monopolization and company formation, as well as marginalization and attempts at exclusion on the part of all other profiteers in the slave and human trade, including Europe-, Africaand America-based sovereigns, monopolists, and wholesale banker-merchants (shipowners and armadores). For this reason, they often found themselves alongside former slaves, pirates, corsairs, and other anti-monopolists, or in general enemies of state regulations (cimarronaje as Caribbean-Atlantic culture).84 In the wake of the formation of the Anglo-Atlantic and a deliberate policy of Christianization and Europeanization (generally also with national exclusiveness), Atlantic creoles were often, though not always, pressed into service functions in relation to the European-Christian and “white”-dominated slave trade. Despite this, they remained, even en masse, bearers of a still scarcely researched Atlantic culture, and bearers of networks between America and Africa. As opposed to the enslaved, who undoubtedly also included a number of Atlantic creoles (Ira Berlin), slave traders or their employees habitually moved between continents on the sea (or on rivers into the continental mass), i.e. back and forward, not just as slaves from Africa to America.85 The spread of maize, manioc (yucca and cassava), European cattle, horses, herd and guard animals (in particular, cattle, mules, horses, and dogs), groundnuts and tobacco, as well as Atlantic epidemics, formed an irregular biological background to these diasporas. The spread of yucca/manioc and of preservable tapioca flour characterized the early slave empire of Brazil-Congo-Angola. Alcohol and tobacco were a particular consumption vice of the slave trade and the enslaved, but also goods for exchange in the slave and human trade. The extent to which opium and cocaine also played a role alongside alcohol and tobacco remains to be studied.

84  Jane G. Landers, “Cimarron Ethnicity and Cultural Adaptation in the Spanish Domains of the Circum-Caribbean, 1503–1763”, in Lovejoy, Identity in the Shadow of Slavery, pp. 30–54. 85  Michael Zeuske, “Mongos und Negreros: Atlantische Sklavenhändler im 19. Jahrhundert und der iberische Sklavenhandel 1808/1820–1873”, Periplus. Jahrbuch für außereuropäische Geschichte, 20 (2010), and Christine Hatzky and Ulrike Schmieder (eds), Sklaverei und Postemanzipationsgesellschaften in Afrika und in der Karibik, (Münster, 2010), pp. 57–116.

Historiography and Research Problems of Slavery

269

The capture of slaves, the slave trade, and slaveries were part of an earlyhuman capitalism, which from the start of the modern age in the narrower sense (c. 1870) drew Africa into active capitalist development, but a development that in the Christian Atlantic states was deliberately marginalized (racist arguments about civilization and Christianity), and only by the fixing of capital value in money, along with European manufactured goods, became the instrument of early European trading capitalism in Europe and in the European colonies, summarized in the term rescate and broadly traceable in the history of the word razzia (trade/theft/enslavement).86 The trading routes of early mercantile capitalism were often mainly the routes of slave ships and slave caravans; the real rise in value took place on the marginalized Atlantic lines of the Middle Passage between Africa and the Americas. And both Atlantic slavery and the slaveries on the shores of the Indian Ocean and in the maritime lands of South-East Asia offer the most important and largest historical foundation for all forms of unfree labour in modern times.

The Dominance of “Great Slaveries”

A basic problem of the structure of slavery and slave-trade research with a global-history orientation is its fixation on “great” slaveries. The reason for this is evidently that these “great” slaveries, meaning from a European perspective above all Roman slavery, produced a great corpus in terms of texts, legal discourses, essays, and historiographies, and still continue to produce these— studies of antiquity are among the strongest legitimation of European and neo-European superiority. With the other “great” hegemonic slavery: i.e. the Islamic-Persian-Egyptian-Indian, which produced a similar wealth of texts and historiographies, the European or North American perspective already has a harder time as well as with the “great slaveries” in Africa.87

86  Germán Peralta Rivera, Los mecanismos del comercio negrero (Lima, 1990). 87  Gabriel Baer, “Slavery in Nineteenth Century Egypt”, Journal of African History ( JAfrH), 8 (1967), pp. 417–441; Paul E. Lovejoy, “Plantations in the Economy of the Sokoto Caliphate”, JAfrH, 19 (1978), pp. 341–368; idem, “The Characteristics of Plantations in the NineteenthCentury Sokoto Caliphate (Islamic West Africa)”, American Historical Review, 84 (1979), pp. 1267–1292; Frederick Cooper, From Slaves to Squatters: Plantation Labor and Agriculture in Zanzibar and Coastal Kenya 1890–1925 (New Haven, CT, 1980); idem, Plantation Slavery on the East Coast of Africa (Portsmouth, NH, 1997); Lovejoy, Transformations in Slavery; Stephanie Beswick and Jay Spaulding (eds), African Systems of Slavery (Trenton, NJ, 2010).

270

Zeuske

Studies on the long-run history of the most varied local “small” slaveries “without the name of slavery” (but with a large number of specific names and legal conceptualizations), or of slaveries based on kinship, raiding or sacrifice, can be found in the realm of ethnology and non-European anthropology; also in the local historiographies of many colonial areas and expansion processes (including Europe, for example, Carolingian and Viking expansion), precolonial regions and zones of contact between Europeans, their successors, and non-Europeans.88 If there are beginnings of research into “prehistoric” forms of slavery or slaveries outside antique Greece and Rome (Gronenborn, Peschel, Heinen, Sommer, and Taylor),89 most of these are overlaid by hegemonic slaveries, and have been conceptualized as “special forms” (like the slavery of the helots, a kind of slavery Sonderweg), quite markedly in the case of Finley’s “individuality” of antique Mediterranean slavery; “overwritten”, as media theory would put it. In this way, the history of slavery is repeatedly canonized in a kind of formation theory after the model “ancient-East-antiquity—(more recently) 88  As representative of these: Walter Rodney, “African Slavery and Other Forms of Social Oppression on the Upper Guinea Coast in the Context of the Atlantic Slave Trade”, Journal of African History, 7 (1966), pp. 431–443; António Carreira, O tráfico de escravos nos rios de Guiné e Ilhas de Cabo Verde (1810–1850): Subsidios para o seu estudio (Lisbon, 1981); Robertson and Klein, Women and Slavery in Africa; Suzanne Miers and Igor Kopytoff (eds), Slavery in Africa: Historical and Anthropological Perspectives (Madison, WI, 1977); Suzanne Miers and Richard Roberts (eds), The End of Slavery in Africa (Madison, WI, 1988); Suzanne Miers and Martin A. Klein, Slavery and Colonial Rule in Africa (London, 1998), also published as a special issue of Slavery & Abolition, 19:2 (1998); Lovejoy, Transformations in Slavery; John K. Thornton, “Slavery and African Social Structure”, in idem, Africa and the Africans, pp. 72–97; idem, “Africa: The Source”, in Mariners’ Museum, Captive Passage: The Transatlantic Slave Trade and the Making of the Americas (Washington DC [etc.], 2002), pp. 35–51; James F. Brooks, Captives & Cousins: Slavery, Kinship, and Community in the Southwest Borderlands (Chapel Hill, NC, 2002); Pauline T. Strong, “Transforming Outsiders: Captivity, Adoption, and Slavery Reconsidered”, in Philipp J. Deloria and Neil Salisbury (eds), A Companion to American Indian History (Maiden, MA, 2004), pp. 339–356. 89   Detlef Gronenborn, “Zum (möglichen) Nachweis von Sklaven/Unfreien in prähistorischen Gesellschaften Mitteleuropas”, pp. 1–42, and Heinz Heinen, “Sklaverei im nördlichen Schwarzmeerraum: zum Stand der Forschung”, pp. 487–503, both in Heinz Bellen and Heinz Heinen (eds), Fünfzig Jahre Forschungen zur antiken Sklaverei an der Mainzer Akademie 1950–2000. Miscellanea zum Jubiläum (Stuttgart, 2001); Karl Peschel, “Archäologisches zur Frage der Unfreiheit bei den Kelten während der vorrömischen Eisenzeit”, Ethnographisch-Archäologische Zeitschrift, 31 (1990), pp. 370–417; Timothy Taylor, “Believing the Ancients: Quantitative and Qualitative Dimensions of Slavery and the Slave Trade in Later Prehistoric Eurasia”, World Archaeology, 33 (2001), pp. 27–43.

Historiography and Research Problems of Slavery

271

Islam-American-plantation-slavery-abolition-end”. The problem of a global perspective on slaveries and human trafficking in the twenty-first century is that today visible “great” slaveries in the tradition of “hegemonic” slaveries or legal ownership over human bodies no longer exist. The most interesting cases of different forms of slaveries (colonate, Leibeigenschaft, Viking and other Scandinavian slaveries, slaveries of the Mongols, Russians, or Chinese, or the slavery of gypsies in Romania) and slave trades are debated today in medieval Eurasian history, far from this formation theory and in a comparative perspective with African forms of slavery.90 All the “prehistoric”, “small” slaveries belong to the major genus of kin-and age-group slaveries. Without the types of kin-slavery and transitional forms to larger slaveries (as with the Phoenicians and Etruscans),91 it is impossible to understand slavery outside the realm in which the concept of slavery in the “Roman” tradition applies; neither pre-colonial slaveries and the slave trades in Africa92 and in the Americas “without Europeans” nor elsewhere on the globe, including the dynamics of early slave exchange between Africans and Europeans, or Europeans and indigenous peoples, in the Americas. The sources of the Atlantic slave trade and the “great” slaveries formed on this basis lie in the “small” slaveries and the dynamic raiding slaveries of Africa, and, despite royal prohibition, also in the “small” slaveries on the peripheries of the European colonial empires in the Americas.93 Kin slaveries also existed within the “great” plantation slaveries, as Gilberto Freyre has impressively shown from the example of Brazil (this despite his conceptual errors). In other plantation societies, too, slaveowners and their 90  Joseph E. Inikori, “Slaves or Serfs? A Comparative Study of Slavery and Serfdom in Europe and Africa”, in Okpewho, Davies, and Mazrui, African Diaspora; Michael McCormick, Origins of the European Economy: Communication and Commerce, AD 300–900 (Cambridge, 2001); idem, “New Light on the ‘Dark Ages’: How the Slave Trade Fuelled the Carolingian Economy”, Past and Present, 177 (2002), pp. 17–54; Joachim Henning, “Strong Rulers—Weak Economy? Rome, the Carolingians and the Archaeology of Slavery in the First Millennium AD”, in Jennifer Davis and Michael McCormick (eds), The Long Morning of Medieval Europe: New Directions in Early Medieval Studies (Aldershot, 2008), pp. 33–53; Viorel Achim, The Roma in Romanian History (Budapest, 2004). 91  Michael Sommer, Die Phönizier Handelsherren zwischen Orient und Okzident (Stuttgart, 2005), pp. 249–264. 92  Patrick Manning, Slavery and African Life: Occidental, Oriental and African Slave Trades (Cambridge, 1990); idem (ed.), Slave Trades, 1500–1800: Globalisation of Forced Labour (Aldershot, 1996). 93  Michael Zeuske, “Sklaven und Kin-Sklavereien”, Amerindian Research. Zeitschrift für indianische Kulturen von Alaska bis Feuerland, 5:2 (no. 16) (2010), pp. 92–104.

272

Zeuske

employees had children with slave women. And in terms of perspective, it is particularly important to point out that present-day forms of slavery can no longer be grasped in terms of the concepts of “great” slaveries, but only those concepts of “small” slavery and concealed debt slaveries. Today, in 2012, far more slaves (and human trafficking in different kinds) exist than at any time in the past (estimates reaching from 27 million to as much as 270 million). The problem of the literature dealing with today’s forms of slavery and trade in humans is, that, with few exceptions (for example Kevin Bales),94 it is too much focused on prostitution, transnational crime, or illegal migration, and separated from history in the very long run (longue durée), particularly that of these “small” forms of slavery and pawn-slaveries.

German Historiography on Slavery

Research on Atlantic slavery and other types of slavery in Germany can be described as an alternation between brilliance and silence. The brilliance was in the past, the silence is today.95 There is scarcely anything theoretical any more. Some of the earliest works of synthesis on the slave trade came from the pens of German historians (especially Römer, Sell, Sprengel, Hüne, and Häbler).96 The most important liberal manifesto against slavery in the 94  Kevin Bales, Disposable People: New Slavery in the Global Economy (Berkeley, CA, 2004); idem, The Slave Next Door: Human Trafficking and Slavery in America Today (Berkeley, CA, 2009). 95  This is of course meant relatively. See, for example: Heinz Heinen (ed.), Menschenraub, Menschenhandel und Sklaverei in antiker und moderner Perspektive. Ergebnisse des Mitarbeitertreffens des Akademievorhabens Forschungen zur antiken Sklaverei (Mainz, 10 October 2006) (Stuttgart, 2008); Elisabeth Hermann-Otto, “Die antike Sklaverei und ihre Rezeption”, in idem, Sklaverei und Freilassung in der griechisch-römischen Welt (Hildesheim, 2009), pp. 43–50; Heinz Heinen (ed.), Antike Sklaverei: Rückblick und Ausblick. Neue Beiträge zur Forschungsgeschichte und zur Erschließung der archäologischen Zeugnisse (Stuttgart, 2010); see also the chapter “Die hässlichen Seiten des Lebens—Diskriminierung, Gewalt und Verbrechen”, in Eberhard Schmitt (ed.), Dokumente zur Geschichte der europäischen Expansion, 5 vols (Munich, 1986–1988); I–IV, and V: Eberhard Schmitt and Thomas Beck (eds) Das Leben in den Kolonien (Wiesbaden, 2003), pp. 396–477. 96  Ludwig Ferdinand Römer, Die Handlung verschiedener Völker auf der Küsten Guinea und in Westindien (Copenhagen, 1758); idem, Nachrichten von der Küste Guinea. Mit einer Vorrede v. D. Erich Pontoppidan, aus dem Dänischen übersetzt (Copenhagen [etc.], 1769); Matthias Christian Sprengel, Vom Ursprung des Negerhandels, ein Antrittsprogramm (Halle, 1779); Johann Jakob Sell, Versuch einer Geschichte des Negersclavenhandels (Halle, 1791); Albert Hüne, Vollständige historisch-philosophische Darstellung aller Veränderungen

Historiography and Research Problems of Slavery

273

nineteenth century was Alexander von Humboldt”s Essay über die Insel Cuba.97 Karl Marx had at least certain theoretical ideas about slavery (his misunderstanding of slavery as “an anomaly in capitalism”),98 and on the connection between “primitive” accumulation, English industrial capitalism, the slave trade, and slavery. It is clear today that accumulation is (up to now) eternal and that slavery is indeed capitalism.99 It is also increasingly clear that the Islamic territories and Africa also had a centrality because of their dynamic slaving, and it was peripheral spaces in the economic system (seen from northwestern Europe), as in modern times the Americas (especially the USA, but also Brazil, Mexico, Cuba, Venezuela, Colombia and many other places) whose world-historical importance profited from the slave trade, slavery, forced migration, implanted resources, natural conditions and racism.100 des Negersclavenhandels von dessen Ursprunge an bis zu seiner gänzlichen Aufhebung, 2 vols (Göttingen, 1820); Johann Andreas Riemer, Missionsreise nach Surinam und Berbice zu einer am Surinamflusse im dritten Grad der Linie wohnenden Freynegernation: Nebst einigen Nachrichten über die Missionsanstalten der Bruderunität zu Paramaribo (Mit Kuppfern) (Zittau [etc.], 1801); Konrad Häbler, “Die Anfänge der Sklaverei in Amerika”, Zeitschrift für Social- und Wirtschaftsgeschichte, 4 (1896), pp. 176–223. 97  Alexander von Humboldt, Essai politique sur l’Ile de Cuba, avec une carte et un supplément qui renferme des considérations sur la population, la richesse territoriale et le commerce de l’Archipel des Antilles et de Colombia, 2 vols. (Paris, 1826); idem, Cuba-Werk. Hrsg. u. komm. von Hanno Beck in Verbindung mit W.-D. Grün et al. (Darmstadt, 1992), Alexander von Humboldt Studienausgabe. Sieben Bände. Bd. III) (= Humboldt, Cuba-Werk); see also my essay on Humboldt: M. Zeuske, “Humboldt, Historismus, Humboldteanisierung”, part 1, Humboldt im Netz (HiN), International Review for Humboldtian Studies, 2:3 (2001), www.unipotsdam.de/u/romanistik/humboldt/hin/hin3.htm; part 2, HiN, 3:4 (2002), www. unipotsdam.de/u/romanistik/humboldt/hin/hin_4.htm. 98  Karl Marx, Grundrisse (London, 1973), p. 463; Wilhelm Backhaus, Marx, Engels und die Sklaverei: zur ökonomischen Problematik der Unfreiheit (Düsseldorf, 1974); Claude Meillassoux, Antropología de la esclavitud (Mexico City, 1990), pp. 20–23; M. Zeuske, “Sklaven in der deutschen Geschichtsschreibung”, in idem (ed.), Nach der Sklaverei. Grundprobleme amerikanischer Postemanzipationsgesellschaften (Leipzig, 1997), = COMPARATIV. Leipziger Beiträge zur Universalgeschichte und zur vergleichenden Gesellschaftsforschung, 7:1 (1997), pp. 7–17. 99  Marcel van der Linden, “Eine einfache und dennoch schwer zu beantwortende Frage: Warum gab (und gibt) es Sklaverei im Kapitalismus?”, in M. Kabadayi, M. Erdem, and Tobias Reichardt, Unfreie Arbeit. Ökonomische und kulturgeschichtliche Perspektiven (Hildesheim [etc.], 2007), pp. 260–279; Roth and Van der Linden, “Karl Marx und das Problem der Sklavenarbeit”, pp. 581–586. 100  Margarete Grandner and Andrea Komlosy (eds), Vom Weltgeist beseelt. Globalgeschichte 1700–1815 (Vienna, 2004); David Eltis, Frank Lewis, and Kenneth Sokoloff (eds), Slavery in the Development of the Americas (Cambridge, 2004).

274

Zeuske

In the German language there is today no modern scientific synthesis on Atlantic slavery, and little genuine research. Important works stem from the Africa historians Heinrich Loth, Albert Wirz, Helmut Bley, and Jan-Georg Deutsch,101 as well as from world-history oriented historians such as Wolfgang Reinhard102 und Jürgen Osterhammel;103 historians of Latin America have dealt with the slave trade and slavery more in essays.104 Loth published an analysis of the slave trade. Albert Wirz’s book is a survey in the style of Wallerstein, focused on the British and Anglo-American realm. A preferable source work is that of Peter Martin on Africans in the history and consciousness of Germans. Wolfgang Binder and Rüdiger Zoller have published the proceedings of congresses on slavery in the Americas.105 The synthesis History of African Americans,106 published in 1999 by Norbert Finzsch and fellow authors, places US slavery above all in the context of the “race and racism debate”.

101  Jan-Georg Deutsch, “Sklaverei als historischer Prozeß”, in idem and Albert Wirz (eds), Geschichte in Afrika. Einführung in Probleme und Debatten (Berlin, 1997), pp. 53–74. 102  Wolfgang Reinhard, “Frühneuzeitliche Negersklaverei und ihre Bedeutung für Wirtschaft und Gesellschaft”, Geschichte in Wissenschaft und Unterricht, 37 (1986), pp. 660–672. Reinhard’s major works on the history of colonialism include further details on slavery. 103  Jürgen Osterhammel, Sklaverei und die Zivilisation des Westens (Munich, 2000). 104  Heinrich Loth, Sklaverei: die Geschichte des Sklavenhandels zwischen Afrika und Amerika (Wuppertal, 1981); idem, Das Sklavenschiff: die Geschichte des Sklavenhandels Afrika, Westindien, Amerika (Berlin, 1984); Wirz, Sklaverei und kapitalistisches Weltsystem, also idem, “Transatlantischer Sklavenhandel, Industrielle Revolution und die Unterentwicklung Afrikas”, Geschichte und Gesellschaft, 8 (1982), pp. 518–557; Helmut Bley et al. (eds), Sklaverei in Afrika: afrikanische Gesellschaften im Zusammenhang von europäischer und interner Sklaverei und Sklavenhandel (Pfaffenweiler, 1991); Horst Pietschmann, “Der atlantische Sklavenhandel bis zum Ausgang des 18. Jahrhunderts— Eine Problemskizze”, Historisches Jahrbuch, 107 (1987), pp. 122–133; Hans-Jürgen Puhle (ed.), Sklaverei in der modernen Geschichte, Geschichte und Gesellschaft. Zeitschrift für Historische Sozialwissenschaft, 16:2 (1990); Wolfgang Reinhard, Parasit oder Partner?: europäische Wirtschaft und neue Welt 1500–1800 (Münster [etc.], 1998). 105  Binder, Slavery in the Americas; Rüdiger Zoller (ed.), Amerikaner wider Willen: Beiträge zur Sklaverei in Lateinamerika und ihre Folgen (Frankfurt, 1994); Peter Martin, Schwarze Teufel, edle Mohren. Afrikaner in Geschichte und Bewußtsein der Deutschen (Hamburg, 2001). 106  Norbert Finzsch, James O. Horton, and Lois Horton, Von Benin nach Baltimore. Die Geschichte der African Americans (Hamburg, 1999).

Historiography and Research Problems of Slavery

275

Conclusion The view represented by the author, however (on the basis of field and archival work), that the history of “slavery” today should be not so much a history of the institution of slavery, but rather and especially a trans-cultural history of actors, in the first place of slaves of both sexes (because there is least research on these), as well as of individual slave-owners, slave traders, and their ancillaries between the micro and macro history of different spaces (above all seas, oceans, islands and coasts), has not prevailed. In global history, analysis of slavery should be replaced by the history of slaveries, or of actors in these slaveries, in the tradition of “small” and kin slaveries, which extend up to the present.107 The syntheses published around 2008 for the 200th anniversary of the abolition of the slave trade by Britain and America were rather hasty anniversary-oriented essays, without any basis in research (with the exception of the Atlantic sections in Schwarzes Amerika),108 a fact that may be due not least to a mistaken publication policy (and ignorance) on the part of publishers, particularly clearly expressed in a so-called Weltgeschichte der Sklaverei (Munich, 2009), which is more like a Christian-globalist pamphlet directed against what the author terms “Islamic slavery”. The history of slaves is, still more than the history of the economic macrostructure of the “Atlantic slave trade”, or of the violent institution of “slavery” or the great ideology “racism”, a transcultural subject, but one partialized 107  M. Zeuske (with Rebecca J. Scott), “Le ‘droit d’avoir des droits’. Les revendications des ex-esclaves à Cuba (1872–1909)”, Annales HSS, 3 (2004), pp. 521–545; idem, “Comparing or Interlinking? Economic Comparisons of Early Nineteenth-Century Slave Systems in the Americas in Historical Perspective”, in Dal Lago and Katsari, Slave Systems, pp. 148–183; idem, “Unfreiheit abhängiger Landbevölkerung im atlantischen Raum und in den Amerikas, 15. bis 18. Jahrhundert—Prolegomena, Typologien der Anfänge, Bedingungen und lange Linien”, in Elisabeth Hermann-Otto (ed.), Unfreie und abhängige Landbevölkerung (Hildesheim [etc.], 2008), pp. 71–157; idem and Laviña, “Failures of Atlantization”, in Tomich and Zeuske, The Second Slavery; Michael Zeuske and Orlando García Martínez, “La Amistad de Cuba. Ramón Ferrer, contrabando de esclavos, captividad y modernidad atlántica”, Caribbean Studies, 37 (2009), pp. 97–170; M. Zeuske, Globalgeschichte der Sklaverei. Menschen als Kapital gestern und heute (Munich, 2012, in preparation). 108  Jochen Meissner, Ulrich Mücke, and Klaus Weber, Schwarzes Amerika. Eine Geschichte der Sklaverei (Munich, 2008); see also Klaus Weber, “Deutschland, der Atlantische Sklavenhandel und die Plantagenwirtschaft der Neuen Welt”, Journal of Modern European History, 7 (2009), pp. 37–67.

276

Zeuske

nationally or regionally by scientific traditions, perspectives, and the professionalization of historiographies; paradigmatically, this partialization is expressed in the context of a “nation” in Colombia or in still quasi-colonially structured spaces, such as the Caribbean, or in the slavery historiography of Venezuela, which is marked very strongly by Hubert Aimes (1907), Cuban historiography (Fernando Ortiz), as well as the Venezuelan “nationalization of Marxism” (Miguel Acosta Saignes).109 At the same time, the rise of a new historical problem consciousness for the history of slaves and slavery can be studied from the rise of these historiographies. Translation: David Fernbach 109  Alfonso Múnera, “Balance historiográfico de la esclavitud en Colombia”, in idem, Fronteras imaginadas. La constructión de las razas y de la geografía en el siglo XIX colombiano (Bogota, 2005), pp. 193–225; Maya Restrepo and Luz Adriana, Brujería y reconstructión de identidades entre los africanos y sus descendientes en la Nueva Granada, siglo XVII (Bogotá, 2005); Michael Zeuske, Schwarze Karibik. Sklaven, Sklavereikulturen und Emanzipation (Zurich, n.d. [2004]); Alberto Abello Vives and Bassi Arévalo, “Un Caribe por fuera de la ruta de la Plantación”, in A. Abello Vives (ed.), Un Caribe sin plantatión. Memorias de la cátedra del Caribe colombiano (San Andrés, 2006), pp. 11–43; “Historiografía y Esclavitud en Venezuela, 1937–2003”, in Dora Dávila Mendoza, La sociedad esclava en la provincia de Venezuela, 1790–1800 (Caracas, 2009), pp. 17–60.

Antiquity to the Early Modern Period

⸪

Between Slavery and Freedom* Moses I. Finley

I

I have taken my title from the Onomastikon or Word-Book of an Alexandrian Greek of the second century of our era named Julius Pollux. At the end of a longish section (3.73–83) listing, and sometimes exemplifying, the Greek words which meant “slave” or “enslave”, in certain contexts at least, Pollux noted that there were also men like the helots in Sparta or the penestae in Thessaly who stood “between the free men and the slaves”. It is no use pretending that this work is very penetrating or systematic, at least in the abridged form in which it has come down to us, but the foundation was laid in a much earlier work by a very learned scholar, Aristophanes of Byzantium, who flourished in the first half of the third century BC. The interest in the brief passage I have cited is that it suggests in so pointed a way that social status could be viewed as a continuum or spectrum; that there were statuses which could only be defined, even if very crudely, as “between slavery and freedom”. Customarily Greek and Roman writers were not concerned with such nuances. To be sure, the Romans had a special word for a freedman, libertus, as distinguished from liber, a free man. When it came to political status, furthermore, distinctions of all kinds were made, necessarily so. But for social status (which I trust I may be permitted, at this stage, to distinguish from political status), and often for purposes of private law, they were satisfied with the simple antinomy, slave or free, even though they could hardly have been unaware of certain gradations. There is a Greek myth which neatly exemplifies the lexical point, a myth certainly much older than its first surviving literary reference in the Agamemnon Source: Finley, Moses I., “Between Slavery and Freedom,” Comparative Studies in Society and History 6(3) (1964): 233–49. © Cambridge University Press, reproduced with permission. *  This is the text, slightly modified, of a lecture I gave to the Royal Anthropological Institute in London on May 30, 1963. I have kept annotation to a bare minimum. For Graeco-Roman slavery, see the bibliographical essay in the volume of articles I edited, Slavery in Classical Antiquity (Cambridge, Eng., 1960). My colleague, Mr. E. R. Leach of King’s College, kindly read and criticized the manuscript. I have also benefited from the opportunity to discuss the servi Caesaris with Mr. P. R. C. Weaver of the University of Western Australia, and problems of American slavery with Professor A. A. Sio of Colgate University.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_011

280

Finley

of Aeschylus produced at Athens in 458 BC. Hercules was afflicted with a disease which persisted until he went to Delphi to consult Apollo about it. There the oracle informed him that his ailment was a punishment for his having killed Iphitus by treachery, and that he could be cured only by having himself sold into slavery for a limited number of years and handing over the purchase price to his victim’s kinsmen. Accordingly he was sold to Omphale, queen of the Lydians (but originally a purely Greek figure), and he worked off his guilt in her service. The texts—which are fairly numerous and scattered over a period of many centuries—disagree on several points: for example, whether Hercules was sold to Omphale by the god Hermes or by friends who accompanied him to Asia for the purpose; whether his term of servitude was one year or three, and so on.1 One has no right to expect neatness in a myth, of course, nor, for that matter, in the legal institutions of the archaic society in which this particular myth arose. The ancient texts all speak of Hercules being “sold”, and to describe his status while in Omphale’s service they employ either doulos, the most common Greek word for “chattel slave”, or latris, a curious word that meant “hired man” and “servant” as well as “slave”. The word latris upsets modern lexicographers and legal historians, but the historical situation behind the lexical “confusion” is surely that in earlier Greece, as in other societies, “service” and “servitude” did in fact merge into each other. The Biblical code was explicit (Deuteronomy 15.12–17): “If thy brother … be sold unto thee and serve thee six years; then in the seventh year thou shalt let him go free from thee … And it shall be, if he say unto thee, I will not go out from thee; because he loveth thee and thy house, because he is well with thee; then thou shalt take an awl, and thrust it into his ear unto the door, and he shall be thy servant for ever.” Cynical remarks are tempting. Quite apart from the very real possibility that the six-year limitation was, as one distinguished authority has phrased it, “a social programme rather than actually functioning law”,2 there is an odd ring about “if he say unto thee, I will not go out from thee; because he loveth thee and thy house”. One suspects that the transition from a more limited bondage to outright slavery was neither so gentle nor so voluntary; that, unlike Hercules, the victims in real life, once caught up in bondage, had little hope of release; that, as in peonage, their masters could find devices enough by which to hold 1  The most important sources are Sophocles, Trach. 68–72, 248–54, 274–76 (with scholia); Apollodorus, Bibl. 2.6.2–3; Diodorus 4.31.5–8. 2  D. Daube, Studies in Biblical Law (Cambridge, Eng., 1947), p. 45; cf. the important recent article of E. E. Urbach, “The Laws regarding Slavery … of the Period of the Second Temple, the Mishnah and Talmud”, Annual of Jewish Studies, 1 (London, 1963), pp. 1–54.

Between Slavery And Freedom

281

them in perpetuity. The sixth-century Athenian statesman Solon, referring to debt-bondsmen, used these words: “I set free those here [in Athens] who were in unworthy enslavement, trembling at the whim of their masters.”3 And the Greek words he employed were precisely those which became the classical terminology of chattel slavery: douleia-slavery; despotes-master; eleutheros-a free man. Modern scholars, too, regularly speak of enslavement for debt. Why not? Why play with words? Why draw elaborate, abstract distinctions? The men Solon liberated belonged to a restricted though numerous class: they were Athenians who had fallen into bondage to other Athenians in Athens. His programme did not extend to non-Athenians, outsiders, who were slaves in Athens, just as the Biblical six-year limitation was restricted to “thy brother”, a fellow-Hebrew, and did not extend to the Gentile. Nor was this merely a sentimental distinction, empty rhetoric holding up vain hopes to the in-group, pretending that they were different from the outsiders when in fact they shared the latter’s fate. The whole story of Solon (like the closely analogous struggles in early Roman history) proves that the distinction was meaningful, though it may have been in abeyance in any individual case or in any given span of time. For Solon was able to abolish debtbondage—indeed, he had been brought to power for that express purpose—following a political struggle that bordered on civil war. Athenian bondsmen had remained Athenians; now they reasserted their rights as Athenians, and they forced an end to the institution— servitude for debt—which had deprived them de facto of all or most of those rights. They were not opposed to slavery as such, only to the subjection of Athenians by other Athenians. Hence, whatever the superficial similarity, this was not a slave revolt; nor did ancient commentators ever make such a connection, despite their resort to slave terminology. I am not now concerned with the history of debt-bondage and its abolition or of clientage in Athens or Rome, nor for the moment with giving precise content to the notion of “rights”. I am merely trying, as a preliminary, to establish the need to distinguish among kinds of servitude, even though contemporaries were themselves not concerned to do so, at least not in their vocabulary. The matter of revolts is worth pursuing a little further in this connection. The debtrevolt syndrome was one of the most significant factors in the early history of both Greece and Rome, and it even survived into classical history. Helot revolts were equally important and very persistent in the history of Sparta. Chattel slaves, on the other hand, showed no such tendency at any time in Greek history and only for a brief period, between about 135 and 70 BC, were there massive 3  Quoted in Aristotle, Const, of Athens 12.4.

282

Finley

slave revolts in Roman history.4 Towards the end of antiquity, finally, there was more or less continual revolt in Gaul and Spain by depressed and semi-servile peasants and slaves acting in concert.5 To explain the differences in the revolt pattern, and particularly in the propensity to revolt, by the differences in treatment, by the relative harshness or mildness of the masters, will not do. The one distinction which stands out most clearly is this, that the chattels, who were both the most rightless of all the servile types and the most complete outsiders in every sense, were precisely those who showed the weakest tendency to cohesive action, the weakest drive to secure freedom. Under certain conditions individual slaves were permitted considerable latitude and eventual emancipation was often held out as an incentive to them. That is another matter, however. Slaves as slaves showed no interest in slavery as an institution. Even when they did revolt, their objective was either to return to their native lands or to reverse the situation where they were, to become masters themselves and to reduce to slavery their previous masters or anyone else who came to hand. Insofar as they thought about freedom, in other words, they accepted the prevailing notion completely: freedom for them, as individuals, included the right to possess other individuals as slaves. Debt-bondsmen and helots, in contrast, fought—when they fought—not only to transfer themselves, as individuals, from one status to the other, but to abolish that particular type of servitude altogether (though not, significantly, to abolish all forms, and particularly not chattel slavery).

II

To a Greek in the age of Pericles or a Roman in Cicero’s day, “freedom” had become a definable concept, and the antinomy, slave-free, a sharp, meaningful distinction. We are their heirs, and also their victims. Sometimes the results are amusing, as in the first efforts in the Far East in the nineteenth century to cope with the word “freedom” for which they had no synonym and which

4  The basic, though brief, study is now Joseph Vogt, Struktur der antiken Sklavenkriege (Akad. d. Wiss. u. d. Lit., Mainz, Abhandlungen, 1957, no. 1). See also Peter Green, “The First Sicilian Slave War”, Past & Present, 20 (1961), pp. 10–29, with discussion in no. 22 (1962), pp. 87–92; Claude Mossé, “Le rôle des esclaves dans les troubles politiques du monde grec …”, Cahiers d’histoire, 6 (Lyon, 1961), pp. 353–60. 5  See E. A. Thompson, “Peasant Revolts in Late Roman Gaul and Spain”, Past & Present, 2 (1952), pp. 11–23.

Between Slavery And Freedom

283

till then was “scarcely possible” in, say, Chinese.6 And sometimes the results are very unfunny, as when western colonial administrators and well-meaning international organizations decree the immediate abolition of such practices as the payment of bride-wealth or the “adoption” of debtors on the ground that they are devices for enslavement.7 My subject, however, is not current social or political policy but history; the simple slave-free antinomy, I propose to argue, has been equally harmful as a tool of analysis when applied to some of the most interesting and seminal periods of our history. “Freedom” is no less complex a concept than “servitude” or “bondage”; it is a concept which had no meaning and no existence for most of human history; it had to be invented finally, and that invention was possible only under very special conditions. Even after it had been invented, furthermore, there remained large numbers of men who could not be socially located as either slave or free, who were “between slavery and freedom”, in the loose language of Aristophanes of Byzantium and Julius Pollux. Let us look at one particular case which came before the royal court of Babylonia in the middle of the sixth century BC, in the so-called neo-Babylonian or Chaldaean period.8 A man borrowed a sum of money from a woman who was head of a religious order, and gave her his son as a debtbondsman. After four years the woman died and both the debt and the debt-bondsman were transferred to her successor. The debtor also died and his son, now his heir, found himself in the position of being simultaneously the debtor and the debt-bondsman (an oddity in the ancient Near East, I may add parenthetically, where the transfer of wives and children for debt was common, but the transfer of the debtor himself was rare, unlike the Graeco-Roman practice). After ten years the bondsman paid over a quantity of barley from his own resources and went to court. The judges made a calculation, according to the conventional ratios, translating each day’s service into barley and then translating the barley (both the real barley and the fictitious barley) into money; this arithmetic produced a sum which was equal to the original loan plus 20% interest per annum for ten years; the court ruled, accordingly, that the debt was now paid up and the bondsman was liberated. 6  E. G. Pulleyblank, “The Origins and Nature of Chattel Slavery in China”, J. Econ. Soc. Hist. Orient, 1 (1958), pp. 185–220, at pp. 204–205. 7  See, e.g., H. N. C. Stevenson, The Economics of the Central Chin Tribes (Bombay, 1943), pp. 175–80. 8  V. Scheil, “La libération juridique d’un fils donné en gage … en 558 av. J.-C.”, Rev. d’Assyriologie, 12 (1915), pp. 1–13; cf. H. Petschow, Neubabylonisches Pfandrecht (Abh. Akad. Leipzig, Phil.-hist Kl., 48, no. 1, 1956), pp. 63–65.

284

Finley

During his ten years of service was the bondsman who was working off his father’s debt (which became his debt) a free man or a slave? Were the Israelites in Egypt slaves because they were called upon, as were most native Egyptians, to perform compulsory labor for the Pharaoh? The answer seems clearly to be “Neither”; or better still, “Yes and no”. In analogous situations the Greeks and Romans defined such service obligations as “slavelike”, and that catches the correct nuance. There were in Babylonia and Egypt chattel slaves in the strict property sense, whose services were not calculated at so much barley or so much anything per day, who could not inherit, own property or take a matter to court. But there was no word in the languages of these regions to encompass all the others, those who were not chattel slaves. To call them all “free” makes no sense because it wipes out the significant variations in status, including the presence of elements of unfreedom, among the bulk of the population. If one examines the various law codes of the ancient Near East, stretching back into the third millennium BC, whether Babylonian or Assyrian or Hittite, the central fact is the existence of a hierarchy of statuses from the king at the top to the chattel slaves at the bottom, with rules—in the penal law, for example—differentiated among them. Translators often enough employ the term “a free man”, but I believe this to be invariably a mistranslation in the strict sense, the imposition of an anachronistic concept on texts in which that concept is not present. It is enough to read the commentaries appended to the translations to appreciate the error: each such rendition requires the most complex contortions in the commentary if the various clauses of the codes are not to founder in crass inner contradictions once “free man” has been inserted. What the codes actually employ are technical status-terms, which we are unable to render precisely because in our tradition the hierarchy and differentiation of statuses had been different. Hence, for example, careful Hittitologists resort to such conventional renditions as “man of the tool”, which may not be very lucid but has the great advantage of not being downright misleading. The English word “slave” is a reasonable translation of one such status-term, but it is then necessary to emphasize the fact that slaves were never very significant and never indispensable in the ancient Near East, unlike Greece or Rome. The neo-Babylonian case I have discussed took place 60 or 70 years before the Persian Wars, by which time the Greek city-state had achieved its classical form, in Asia Minor and the Aegean islands as well as on the mainland of Greece, in southern Italy and Sicily. Proper analysis of classical Greece would require far more space than I have at my disposal, for the society was not nearly so homogeneous throughout the many scattered and independent Greek

Between Slavery And Freedom

285

communities as we often pretend.9 I shall confine myself to two cities, Athens and Sparta, in the fifth and fourth centuries BC, the two cities which the Greeks themselves considered the best exemplars of two sharply contrasting social systems and ideologies. Athens is, of course, the Greek city that comes first to mind in association with the word “freedom”. And Athens was the Greek city which possessed the largest number of chattel slaves. The actual number is a matter of dispute—as are nearly all ancient statistics or, better, statistical guesses—but much of the debate is largely irrelevant since no one can seriously deny that they constituted a critical sector of the labour force (in a way which slaves never did in the ancient Near East). My own guess is of the order of 60–80,000, which would give a ratio to the free population about the same as in the southern states of the United States in the first half of the nineteenth century, but with a different distribution pattern. Proportionally more Athenians than Southerners owned slaves, but there were few if any great concentrations in single hands because there were no plantations, no Roman latifundia. For our present concern there are a number of points to be made about slavery in Athens, which I shall run through briefly. 1)

2)

There were no activities in which slaves were not engaged other than political and military, and even those two categories must be understood very narrowly, for slaves predominated in the police and in what we should call the lower civil service. Contrariwise, there were no activities in which free men were not engaged, which slaves monopolized: they came nearest to achieving that in mining and domestic service. In other words, it was not the nature of the work which distinguished the slave from the free man but the status of the man performing the work. Slaves were outsiders in a double sense. After Solon’s abolition of debt bondage, no Athenian could be a slave in Athens. Hence all slaves to be found there had either been imported from outside the state or had been born within to a slave mother. “Outside the state” could mean a neighboring Greek state as well as Syria or southern Russia—the law never forbade Greeks to enslave other Greeks, as distinct from Athenians and Athenians—but the evidence seems to show that the great majority were in fact non-Greeks, “barbarians” as they called them, and that is why I say “outsiders in a double sense”.

9  On this point see Finley, “The Servile Statuses of Ancient Greece”, Rev. int. des droits de l’antiq., 3rd ser., 7 (1960), pp. 165–89; D. Lotze, Metaxy eleutheron kai doulon. Studien zur Rechtsstellung unfreier Landbevölkerungen in Griechenland … (Berlin, 1959).

286 3)

4)

Finley

Slaveowners had the right, essentially without restriction, to free their slaves, a right which seems to have been exercised with some frequency, especially among domestic servants and skilled craftsmen, though, as usual, we are unable to express the pattern numerically. The contemporary attitude was summed up by Aristotle when he wrote (Rhetoric 1367a32): “The condition of the free man is that he does not live under the restraint of another.” In that sense, manumitted slaves were free men, if we ignore, as we legitimately may in this outline analysis, conditional manumissions and minor obligations towards the exmaster. But in another sense “free man” is an excessively loose category. The distinction between citizens and free non-citizens was not merely political—the right to vote or hold office—but went much deeper: a noncitizen could not own real property, for example, except by special grant of that privilege by the popular assembly, a grant which was rarely made. Nor, for much of the period under consideration, could a non-citizen marry a citizen; their children were by definition bastards, subject to various legal disabilities and excluded from the citizen-body. Manumitted slaves were not citizens, though free in the loose sense, and hence they suffered all the limitations on freedom I have just mentioned. In addition, it should be noted that insofar as slaves were often freed relatively late in life, and insofar as any children born to them were not freed along with them-practices which existed though we do not know in what proportion of the cases—to that extent freed women were effectively denied the right to procreate free children.

Now let us look at Sparta in the same period, the fifth and fourth centuries BC, and in the same schematic way. 1) 2)

3)

The Spartiates proper were a relatively small group, perhaps never more than 10,000 adult males and declining from that figure more or less steadily during our period. Such chattel slaves as existed were wholly insignificant. In their place there existed a relatively numerous servile population known as helots (a word with a disputed etymology) who were scattered over extensive territories in the southern and western Peloponnese, in the districts of Laconia and Messenia. Again we lack figures, but it is certain that the helots outnumbered the Spartiates, perhaps several times over (in contrast to Athens where the proportion of slaves to free was probably of the order of 1:4, of slaves to citizens less than 1:1). Who the helots were in origin is disputed. They may even have been Greeks to begin with, but whether so or not, they were the people of

Between Slavery And Freedom

4)

5)

287

Laconia and Messenia, respectively, whom the Spartans subjugated and then kept in subjection in their own home territories. That immediately distinguished them—and distinguished them sharply—from the chattel slave “outsiders”, not only genetically but also in later history, for they were bound together by something far more than just the weak negative factor of sharing a common fate, by ties of kinship, nationhood (if I may use the term) and tradition, all perpetually reinforced through their survival on their native soil. Insofar as it makes any sense to use the terminology of property, the helots belonged to the state and not to the individual Spartiates to whom they were assigned. (Parenthetically I should say that the word “belonged”, which explains the willingness of the Greeks to call the helots “slaves”, is justified by the existence of a further Peloponnesian population who were politically subject to Sparta but were at the same time free and citizens of their own communities, the perioeci, whom I am ignoring in this discussion.) It follows from the previous point that only the state could manumit helots. They did so only in one type of situation: when military service by helots was unavoidable, those selected were freed, either beforehand or as a subsequent reward. Once freed they did not become Spartiates but acquired a curious and distinct status, as did Spartiates who lost their standing for one reason or another, so that, as in Athens, the category of “free men” was a conglomeration, not a homogeneous single group.

These points do not exhaust the picture, nor do they by any means exhaust the range of differences between Athens and Sparta, but I trust I have said enough to make it clear not only that the differences were very sharp but also that the number of status possibilities was very considerable. It remains to add that whereas for our subject Athens was typical of the more highly urbanized Greek communities on the mainland of Greece and in the Aegean islands, Sparta was, taken whole, unique. However, if we narrow our focus solely to the helots, then parallels were far from uncommon, less so in Greece proper than in the areas of Greek dispersion east and west, such as Sicily or the regions bordering on the Black Sea, where native populations were reduced to a status sufficiently like that of the helots to warrant their being bracketed with them, as Pollux did, under the rubric “between the free men and the slaves”.10

10  In addition to Lotze, op. cit., see D. M. Pippidi, “Die Agrarverhältnisse in den griechischen Städten der Dobrudscha in vorrömischer Zeit”, in Griechische Städte und einheimische Völker des Schwarzmeergebietes (Berlin, 1961), pp. 89–105.

288

Finley

Now, merely to illustrate the variety which actually existed, I want to look briefly at the institution we know from the so-called law code of Gortyn in Crete.11 The text we have was inscribed on stone in the fifth century BC but the provisions may be much older. The code is far from complete, and there are some devilishly difficult problems in interpretation. It is clear, however, that there was a servile population which in some sense “belonged” to individual Gortynians who could buy and sell them (apparently with restrictions hinted at, but not clarified, in the code), unlike the situation in Sparta with which too easy comparisons are often made. Yet this same servile population had rights which slaves in Athens lacked. For example, the rules regarding adultery and divorce and the provisions regulating relations between bondsmen and free women leave no doubt that it is proper to speak of marriage, of a relationship which was more than the Roman contubernium between slaves, because it created enforceable rights, but which was at the same time far less than a marriage between free persons. For one thing, an unfree husband was not his wife’s tutor; that role was fulfilled by her master. For another, such a marriage did not lead to the creation of a kinship group, although it created the elementary family for certain purposes. Hence a composition payment for adultery could be arranged with the kinsmen of a free woman, but only with the master of a servile woman. (Parenthetically, I should also note that debt-bondsmen are clearly differentiated in the code from the bondsmen I have been discussing.) After the conquests of Alexander the Great, finally, when Greeks and Macedonians became the ruling class in Egypt, Syria and other lands of the ancient Near East, they found no difficulty in adapting themselves to the social structure which had been in existence there for millennia, modifying the top of the pyramid more than the bottom. A city in the Greek style like Alexandria had its chattel slaves just as in Athens; in the Egyptian countryside, however, the peasantry remained in its traditional status, neither free nor unfree. Royal grants of land to favorite ministers included whole villages along with their inhabitants. Compulsory labor services of various kinds were imposed on them, precisely as on the Israelites a thousand years earlier. Our greatest historian of this era, Rostovtzeff, has written of this peasantry that “They possessed a good deal of social and economic freedom in general and of freedom of movement in particular,… And yet they were not entirely free. They were bound to the government and could not escape from this bondage, because on it depended

11  See the works cited in n. 9 and D. Lotze, “Zu den woikees von Gortyn”, Klio, 40 (1962), pp. 32–43.

Between Slavery And Freedom

289

their means of subsistence. This bondage was real, not nominal.”12 Which both makes my point and illustrates, in the vagueness and inadequacy of its formulations, how far we still are from a proper analysis of the social pattern. The Romans, who eventually replaced the Greeks as rulers of this whole area, had a history of servitude more like that of Athens than of Sparta or the Near East, but with features of their own worth our notice. They, too, had an internal crisis in the archaic period brought about by massive debt-bondage. They, too, then turned to chattel slaves on a large scale, the form of dependent labor which was characteristic of Rome in what I shall arbitrarily define as its classical period, roughly speaking, the three centuries between 150 BC and 150 AD “Rome” is here ambiguous: we normally use it to refer both to the city on the Tiber and to the whole of the Roman Empire, which by the end of the classical age extended from the Euphrates to the Atlantic. I want to focus on neither, however, but on Italy, the Latin heartland of the Empire, which had become sufficiently uniform socially and culturally to warrant our treating it as a unit. And I want to single out a few characteristics of slavery in Italy which contribute new dimensions to the picture I have drawn so far. 1)

2)

The great landed estates of Italy, the latifundia, which specialized in ranching, olive- and wine-production, remained, at least until the American South replaced them, the western model of slave agriculture par excellence. Slave numbers there, and in the rich urban households, reached proportions far exceeding anything in Greece. In the final struggle between Pompey and Caesar, for example, Pompey’s son enlisted 800 slaves from his shepherds and personal attendants to add to his father’s army.13 In a law of 2 BC, Augustus restricted to 100 the number of slaves a man could manumit in his will, and only an owner of 500 or more was permitted to free that many.14 A certain Pedanius Secundus, who was prefect of the city in AD 61, maintained 400 slaves.15 These are examples at the upper end of the scale, to be sure, but they help fix the whole level. Upon manumission a freedman acquired the status of his ex-master, so that the freed slave of a Roman citizen became a citizen himself, distinguished by certain minor disabilities (chiefly with respect to his former master) but none the less a citizen with the right to vote and to marry in

12   The Social & Economic History of the Hellenistic World (3 vols., Oxford, corr. ed., 1953), I  320. 13  Caesar, Bell civ. 3.4.4. 14  Gaius, Inst. 1.43. 15  Tacitus, Ann. 14.43.4.

290

3)

4)

Finley

the citizen class. This last had interesting and amusing implications. Within the Roman imperial territory there was a complicated variety of free statuses in the sense that there were numerous non-Romans, free and citizens of their communities, who lacked both the political rights of Roman citizenship and the ius conubii, the right to contract a marriage with a Roman citizen. But an ex-slave, by the mere private act of manumission, which required no approval from the government, automatically jumped the queue, in law at any rate, provided his master was a Roman citizen. A significant proportion of the industrial and business activity in Rome and other cities was carried on by slaves acting independently, controlling and managing property known as a peculium. This was a legal device invented in the first instance to permit adults to function independently while still technically in patria potestas, the tenacity of which in Rome is one of the most remarkable features of the social history of that civilization. The extension of peculium to slaves created legal problems of great complexity—in the event of a lawsuit, to give the most obvious example—but they do not concern me now, apart from one notable anomaly. It was possible, and by no means rare, that a peculium included one or more slaves, leaving the slave in charge of the peculium in the position of owning other slaves de facto, though not de jure. The reason I have singled out peculium can perhaps best be clarified by some rhetorical questions. In what sense were a slave loaded with chains in one of the notorious agricultural ergastula and a slave managing a sizeable tannery which was his peculium both members of the same class we (and the Romans) call “slaves”? Who was more free, or more unfree, a slave with a peculium or a “free” debtbondsman? Can the concept of freedom be usefully employed at all in such comparisons? In order to insure their administrative control, the early emperors, beginning with Augustus and reaching a crescendo under Claudius and Nero, made extensive use of their own familiae in running the Empire. The servi and liberti Caesaris, the emperor’s own slaves and freedmen, took charge of the bureaus and even headed them for a time. Careful investigation has shown that even among these imperial slaves their children were not as a rule freed along with them if they were also slaves—there are complications here, arising from the status of the mothers, which I need not go into—but stayed on as servi Caesaris, advancing in the service if they were capable and earning their own freedom in time. Hence the interesting situation was created in which important civil servants not only came out of the slave class but left their children behind in that class. And more

Between Slavery And Freedom

291

interesting still, the generalization may be made that in Rome of the first century of our era, much the greatest opportunity for social mobility lay among the imperial slaves. No one among the free poor could have risen to a status like that of head of the bureau of accounts, or, for that matter, to anything like the many lower posts in the administration. I doubt if I need make further comment.

III

All the societies I have been discussing, from those of the Near East in the third millennium BC to the end of the Roman Empire, shared without exception, and throughout their history, a need for dependent, involuntary labor. Structurally and ideologically, dependent labor was integral, indispensable. In the first book of the Pseudo-Aristotelian Oeconomica we read: “Of property, the first and most necessary kind, the best and most manageable, is man. Therefore the first step is to procure good slaves. Of slaves there are two kinds, the overseer and the worker.” Just like that, without justification or embellishment. There is no need to pile on the quotations; it is simpler to note that not even the ancient believers in the brotherhood of man were opponents of slavery: the best that Seneca the Stoic and St Paul the Christian could offer was some variation on the theme, “status doesn’t matter”. Diogenes the Cynic, it is said, was once seized by pirates and taken to Corinth to be sold. Standing on the auction block he pointed to a certain Corinthian among the buyers and said: “Sell me to him; he needs a master.”16 Most revealing of all is the firm implication in many ancient texts, and often the explicit statement, that one element of freedom was the freedom to enslave others. Aristotle wrote the following in the Politics (1333b38ff., translated by Barker): “Training for war should not be pursued with a view to enslaving men who do not deserve such a fate. Its objects should be these—first, to prevent men from ever becoming enslaved themselves; secondly, to put men in a position to exercise leadership …; and thirdly, to enable men to make themselves masters of those who naturally deserve to be slaves.” It may be objected that I am unfair to select a text from Aristotle, the most forthright exponent of the doctrine of natural slavery, a doctrine which was combatted in his own day and generally rejected by philosophers in later generations. Let us then try another text. About the year 400 BC an Athenian cripple who had been taken off the dole on the ground that the amount of property he owned made him 16  Diogenes Laertius 6.74.

292

Finley

ineligible, appealed formally to the Council for reconsideration of his case. One of his arguments was that he could not yet afford to buy a slave who would support him, though he hoped eventually to do so.17 Here was no theorist but a humble Athenian addressing a body of his fellow-citizens in the hope of gaining a pittance from them. The implications—and the whole psychology— could scarcely be brought out more sharply. I do not propose to revive the old question of the origin of the inequality of classes, to ask why dependent labor was indispensable. My starting-point is the fact that everywhere in the civilizations under consideration, as far back as our documentation goes (including the new documentation provided by the Linear B tablets), there was well established reliance on dependent labor. All these societies, as far back as we can trace them, were already complex, articulated, hierarchical, with considerable differentiation of functions and division of labor, with extensive foreign trade and with well-defined political and religious institutions. It is rather what happened thereafter which interests me now: the essentially different development as between the Near East and the Graeco-Roman world, and, in the latter, the sharp differences in different periods as well as the unevenness of development in different sectors. I have already indicated the most fundamental difference, namely, the shift among Greeks and Romans from reliance on the half-free within to reliance on chattel slaves from outside, and as a corollary, the emergence of the idea of freedom. A wholly new social situation emerged, in which not only some of the components were different from anything known before but also the relationships and spread among them, and the thinking. We may not be able to trace the process but we can mark its first literary statement beyond any doubt, in the long poem, the Works and Days, in which Hesiod, an independent Boeotian landowner of the seventh century BC, presumed freely to criticize his betters, the “bribe-devouring princes” with their “crooked judgments”. In another poem, the Theogony, also attributed to Hesiod—and it does not matter whether the attribution is right or not, for the Theogony and the Works and Days were approximately contemporary, which is enough for this discussion—the same new social situation found expression in another area of human behavior, in man’s relations with his gods. As Frankfort phrased it, the author of the Theogony “is without oriental precedent in one respect: the gods and the universe were described by him as a matter of private interest. Such freedom was unheard of in the Near East …”18 It was a firm doctrine in 17  Lysias 24.6. 18  H. Frankfort, ed., Before Philosophy (Penguin ed., 1949), p. 250.

Between Slavery And Freedom

293

the ancient Near East that man was created for the sole and specific purpose of serving the gods: that was the obvious extension by one further step of the hierarchical structure of society. Neither Greek nor Roman religion shared that idea. Man was created by the gods, of course, and he was expected to serve them in a number of ways, as well as to fear them, but his purpose, his function, was not that, and surely not that alone. Institutionally the distinction may be expressed this way: whereas in the Near East government and politics were a function of the religious organization, Greek and Roman religion was a function of the political organization. Hesiod is often called a peasant-poet, which is inexact, for Hesiod was not only himself an owner of slaves but he assumes slavery as an essential condition of life for his class. From the first, therefore, the slave-outsider was as necessary a condition of freedom as the emancipation of clients and debt-bondsmen within. The methods by which outsiders were introduced into the society need not concern us. But it is worth a moment to consider one aspect of the outsider situation, the “racial” one, which is being much discussed today, both by historians and sociologists, chiefly with reference to the American South. It is important to fix in mind that “outsiders” were often neighbors of similar stock and culture; that though the Greeks tried to denigrate the majority of their slaves with the “barbarian” label and though Roman writers (and their modern followers) are full of contemptuous references to “Orientals” among their slaves and freedmen, the weaknesses of this simple classification and its implications were apparent enough even to them. The decisive fact is that widespread manumission and the absence of strict endogamy together destroy all grounds for useful comparison with the American South on this score. I need not go into the variations in antiquity with respect to rights of marriage—I have already indicated the significance of the Roman practice which granted freedmen full rights of conubium, for example, and that should be enough to show that the “racial” element in the concept of the outsider, though not zero, was essentially irrelevant, both in fact and in the ideology. When the Roman lawyers agreed on the formulation, “Slavery is an institution of the ius gentium whereby someone is subject to the dominium of another, contrary to nature,”19 they were saying in effect that slavery was indispensable, that it was defensible only on that ground, and that one was liable to be enslaved just because one was an outsider. An outsider, in short, was an outsider. That tautological definition is the best we can offer. Hence the expansion of the Roman Empire, for example, automatically converted blocks of outsiders to free insiders. 19   Digest 1.5.4.1.

294

Finley

Why, we must then ask, was the historical trend in some Greek communities, such as Athens, and in Rome towards the polarity of the free insider and the slave outsider, while elsewhere no comparable development occurred (or where incipient signs appeared, they soon proved abortive)? Max Weber suggested that the answer lay in the loosening of the royal grip on trade and the consequent emergence of a free trading class who acted as social catalysts.20 I have no great confidence in this hypothesis, which can neither be verified nor falsified from Greek or Roman evidence. The decisive changes occurred precisely in the centuries for which we lack documentation, and for which there is no realistic prospect of new documentation being discovered. I must confess immediately that I have no alternative explanation to offer. Reexamination of the body of Greek and Roman myth may help, but the hope lies, in my opinion, in the very extensive documentation of the ancient Near East. I say “hope”, and no more, because it is no use pretending that study of Near Eastern servitude has taken us very far. One reason is the primitive classification into slave and free which has been my theme, and I now want to return to this and suggest an approach. Merely to say, as I have thus far, that there were statuses between slavery and freedom is obviously not enough. How does one proceed to formulate the differences between a Biblical bondsman who hoped for release and the man who chose to remain a slave in perpetuity and had his ear bored to mark his new status? Or between a helot in Sparta and a chattel in Athens? The Sicilian Greek historian Diodorus, writing as a contemporary of Julius Caesar, gives us the following variation on the Hercules-Omphale myth. Hercules, he says, produced two children during the period of his stay with the Lydian queen, the first by a slave-woman while he was in servitude, the second by Omphale herself after he had been restored to freedom. Unwittingly Diodorus has pointed the way. All men, unless they are Robinson Crusoes, are bundles of claims, privileges, immunities, liabilities and obligations with respect to others. A man’s status is defined by the total of these elements which he possesses or which he has (or has not) the potential of acquiring. Actual and potential must both be considered: the potential of the servi Caesaris, for example, was always a factor in the psychology of status in the early Roman Empire, and sometimes it became an actuality, when one of them climbed high enough on the civil service ladder and was freed. Obviously none of this can be expressed in numerical, quantitative terms: it is not a matter of one 20   “Agrarverhältnisse im Altertum”, in his Gesammelte Aufsätze zur Sozial- und Wirtschaftsgeschichte (Tübingen, 1924), pp. 1–288, at pp. 99–107.

Between Slavery And Freedom

295

man having one more privilege or one more liability than another. Rather it is a matter of location on a spectrum or continuum of status; the servi Caesaris as a class, in this language, stood nearer the freedom end than did the servi of any private owner in Rome. It is possible, furthermore, to work out a typology of rights and duties. By way of illustration, I suggest the following rough scheme:21 1)

2) 3) 4) 5)

6) 7)

Claims to property, or power over things—a category which is itself complex and requires further analysis: for example, the difference between the power of a slave over his peculium and the power of an owner in the strict sense; or differences according to the different categories of things, land, cattle, money, personal possessions, and so forth. Power over human labor and movements, whether one’s own or another’s—including, of course, the privilege of enslaving others. Power to punish, and, conversely, immunity from punishment. Privileges and liabilities in judical process, such as immunity from arbitrary seizure or the capacity to sue and be sued. Privileges in the area of the family: marriage, succession and so on—involving not only property rights and rights of conubium, but, at one step removed, the possibility of protection or redemption in case of debt, ransom or blood-feud. Privileges of social mobility, such as manumission or enfranchisement, and their converse: immunity from, or liability to, bondage, penal servitude and the like. Privileges and duties in the sacral, political and military spheres.

I have said enough, I trust, to forestall any suggestion that I am proposing a mechanical procedure. In Athens chattel slaves and wealthy free non-citizens (Aristotle, for example) were equally barred from marriage with a citizen; in terms of my typology, they both lacked the privilege of conubium. It would be absurd, however, to equate them in a serious sense just on that score. Or to take a more meaningful instance of quite another kind: Athenian slaves and Spartan helots both belonged to someone, but the fact that the someone was a private individual in the one case, the Spartan state in another, introduced a very important distinction. These various combinations must be weighed and judged in terms of the whole structure of the individual society under examination. 21  This is substantially the scheme I first formulated in the article cited in n. 9.

296

Finley

If I am then asked, What has become of the traditional property definition of a slave? Where on your continuum do you draw the line between free and slave, free and unfree?—my answer has to be rather complicated. To begin with, the idea of a continuum or spectrum is metaphorical: it is too smooth. Nevertheless, it is not a bad metaphor when applied to the ancient Near East or to the earliest periods of Greek and Roman history. There one status did shade into another. There, although some men were the property of others and though the gap between the slave and the king was as great as social distance can be, neither the property-definition nor any other single test is really meaningful. There, in short, freedom is not a useful category and therefore it is pointless to ask where one draws the line between the free and the unfree. In classical Athens and Rome, on the other hand, the traditional dividing line, the traditional distinction according to whether a man is or is not the property of another, remains a convenient rule of thumb for most purposes. For them the metaphor of a continuum breaks down. But the problem has not been to understand those two, relatively atypical, societies, but the others, societies which we have not understood very well just because, in my view, we have not emancipated ourselves from the slave-free antinomy. And if my approach proves useful, I suggest it will lead to a better understanding of Athens and Rome, too, where the category of “free man” needs precise subdivision. I might close with a highly schematic model of the history of ancient society. It moved from a society in which status ran along a continuum towards one in which statuses were bunched at the two ends, the slave and the free—a movement which was most nearly completed in the societies which most attract our attention for obvious reasons. And then, under the Roman Empire, the movement was reversed; ancient society gradually returned to a continuum of statuses and was transformed into what we call the medieval world.

A Scientific Approach to Ancient Slavery? Niall McKeown The last twenty years have seen the increasing prominence of a very different style of history writing from that examined in our last few chapters. It is an approach which seems to promise an escape from literary evidence which can be used (apparently) to ornament both sides of an argument. It focuses instead on statistics and on probability. While our knowledge of the date of our own death is (usually and thankfully) vague, the probable mortality and fertility patterns of large groups of people are easier to predict. Those who investigate the demography of the ancient world do not claim to deal in exact answers, only such probabilities. It is these, they argue, that allow one to talk sensibly about the plausible and the probable rather than running around in circles trying to interpret a character in a Roman comedy. Their tables of statistics certainly look scientific. Let us examine their approach by focussing on just one important question they may help to answer: where did the Romans get their slaves, particularly at the height of empire after ad 14? The three main ways of entering slavery were probably capture in war or by pirates, being born a slave, and being left to die by one’s parents but being found and then raised as a slave. Adding up the number of references Romans made to each possibility would probably tell us more about what interested contemporary writers than proving which was the most important. Keith Bradley (1987) argued that all three potential sources were important, but that the evidence might be so impressionistic we can never hope to reach any more exact answer (cf. Parkin 1992: 121–2 and Frier 2000: 808). Walter Scheidel argued that as many as 75–80% of Rome’s slaves were the children of slaves (Scheidel 1997: 167). William Harris believed that this figure was too high and that a significant proportion of Rome’s slaves came instead from outside the empire or from children left to die inside the empire (see Harris 1980 and especially 1999: 62–3). I want to look at the debate in the late 1990s between Harris and Scheidel, but I need to begin by introducing some of the tools demographers use to examine populations.

Source: McKeown, Niall, “A Scientific Approach to Ancient Slavery?” in Niall McKeown, The Invention of Ancient Slavery?, London: Duckworth, 2007, 124–140. © Niall McKeown, 2007, The Invention of Ancient Slavery?, and Bristol Classical Press, an imprint of Bloomsbury Publishing Plc.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_012

298

McKeown

Demographers use something called ‘Model Life Tables’ to predict the age profile of a given population (see Coale, Demeny and Vaughan 1983). These Tables are based on different geographical areas which are able to supply us with reliable statistical information. ‘Model South’ is popular with ancient historians since it is based on Mediterranean statistics and assumes high infant mortality, apparently the case in ancient Rome. ‘Model West’ is even more popular. It is an average of a whole series of ‘Life Tables’ and is probably the safest to generalise from in the absence of solid information. Each ‘Model’ is then subdivided into different population tables determined by the average life expectancy at birth. Life expectancy in modern Britain is now over 70. So far as we can tell it was much lower in pre-industrial societies (largely because of the effect of high infant mortality decreasing the overall average). Most of our (inadequate) ancient evidence tends to produce a range of possibilities of between 20 and 30 years, with the most probable average at the lower end. This corresponds to ‘Model West’ [or ‘South’] Level 2. For a society simply to reproduce itself each woman must, on average, give birth to a female child who survives long enough to be a mother in her turn. The problem for an ancient Roman family would have been the high chance of a daughter dying young. To provide a replacement, the average woman in Italy would have had to give birth to an average of three girls (if Model West Level 2 is anything like a good match for the ancient Roman empire). That means about six births (both male and female) in total. That might seem a high number (certainly if you are a woman) but without it the population would have collapsed. It is time to make a few more assumptions. Walter Scheidel estimated that the population of the Roman empire ultimately reached about 60 million (1997: 158). He assumed that about 6 million (or 10%) of the population were slaves (we will examine some of his evidence later). He argued that the population outside the empire from which the Romans could import slaves was perhaps 20 million, or one third of that of the empire (1997: 159–60). This figure is obviously open to challenge. The Romans, however, generally expanded in most directions until they hit ecological boundaries that made further advance uneconomic (e.g. deserts), or else met political entities which could successfully resist them (e.g. Parthia). It seems logical, therefore, to suppose that the area from which the Romans could draw an external supply of slaves was considerably less populous than the empire itself. Crucially it is not precise numbers that are important here, but the assumptions Scheidel makes about the ratios between free and slave within the empire (9:1) and between the (exploitable) population inside and outside the empire (3:1). Could the slave population have been maintained without needing to draw substantially on the free population for replacements? To answer the

A Scientific Approach To Ancient Slavery ?

299

question we need to know how many babies slave women had each year, and the proportion of the 6 million slaves who either died or were freed each year and who would therefore need replacing. Literary evidence doesn’t give us this information. This is, however, where demography proves its worth. First, how many slaves died each year? Scheidel assumed that slaves didn’t live as long as free people (though cf. Scheidel 2005: 74, 76). Comparative statistics from modern slave societies in the Caribbean and the US suggest 2–5% more deaths among slaves than free people every year. Scheidel chose a lower estimate for Rome, 1% extra (1997: 163–4, with n31). It is difficult to guess how many slaves were freed. Scheidel decided to use a thought experiment. He imagined three different possible levels of manumission, one ‘low’, one ‘intermediate’, one ‘high’. Let’s look at his ‘intermediate’ model. This assumed that 10% of all slaves are freed at age 25 and 10% of the remainder every five years thereafter on average (1997: 160). Freeing slaves would have increased the numbers of replacement slaves needed each year and also have decreased the number of slave children born (1997: 163). One might ask why. The key is the legal status of children. The child of a slave was a slave. But if a woman was freed at the age of 30 (for example) her future children were generally free and so lost to the slave population. The exact proportion lost depended on how often and how early female slaves were manumitted. Hence Scheidel’s different models. If Romans did have anything like Scheidel’s ‘intermediate’ level of manumission, they would have required the equivalent of just under 320,000 babies a year to make into slaves (1997: 164). If one assumes that the proportion of female to male slaves was roughly equal, Scheidel expected just over 200,000 slave births a year. This would still have left a deficit of 110,000–120,000 future slaves. They had to be found among the free population either inside or outside the empire. Scheidel assumed that no more than half of these replacements came from the free population outside the empire (1997: 164). He argued that any higher number would have put ludicrous pressure on the population there because (as we saw) it was possibly only a third of the size of that within the empire. 40,000 captives a year would have filled half of the 110,000–120,000 deficit. That may not seem like half, but it is. Infancy was the most dangerous period in life, but imported slaves would have been older. Since captives had generally already survived infant mortality, fewer were required than babies. That still left almost 60,000 ‘baby equivalents’ needed from within the empire. Adult self-enslavement was known, but it is difficult to judge how frequently it happened. The exposure of children (who might then be raised as slaves) was probably more common. Scheidel argued that is hardly

300

McKeown

reasonable to assume that all Roman families left children out to die because they couldn’t afford to raise them: the richest quarter of the population would surely not have faced this problem. The potential ‘reservoir’ for exposed free children was therefore not the 54 million of the free population but only the poorest threequarters of it, perhaps 40 million (1997: 159). Those 40 million probably produced (according to Model West Level 2) about 1.8 million children each year (1997: 160). In order to fill the gap in slave supply approximately 60,000, or about 1 in 30, must have become slaves (1997: 164). We noted earlier that, simply to keep the population stable, Roman mothers must have had an average of 6 births each. On average therefore 1 in 5 mothers would (at some point of their life) have had to abandon one of their children to slavery if Scheidel’s ‘intermediate model’ approximates to reality. Of course, not all the children exposed would have survived, increasing the number of abandoned children required. Scheidel suggested that perhaps half died, so 1 in 15 surviving children must become slaves. This implies that, on average, 1 in 2 or 1 in 3 of all mothers in the Roman empire would have had to expose a child to produce the required number of slaves, something he saw as straining credulity. Scheidel’s essential argument was that this would have been a phenomenon of unheralded enormity and pain. We ought, therefore, to see copious reference to that pain in our sources. We don’t. This suggested to Scheidel that it is unlikely infant exposure was such an important source of slaves. Scheidel therefore argued, by a process of elimination, that breeding must have been the primary source of slaves, and so even his ‘Intermediate’ model, assuming 67% replacement by breeding, was too conservative. The real figure may have been even higher. There are a number of implications. If 75–80% of slave replacements came from breeding, female slaves could not have been freed as often as the ‘intermediate’ model suggests. Hence Scheidel developed what he saw as a rather more realistic ‘low’ model assuming 10% manumission at age 30 (not 25) and 10% manumission for surviving slaves every 5 years thereafter (1997: 166). He also allowed for a smaller extra slave mortality figure than his ‘intermediate’ model (0.5% rather than 1.0%). Taken together, these changes in his model would allow for 82% of slaves to be bred from existing slaves, with the remaining 55,000 coming from the free population within and without the empire. The key point is that the Romans would have had good reason to prevent female slaves of childbearing age from gaining freedom. Scheidel felt that there was evidence from some Egyptian census statistics to back this up, though he recognised that the numbers involved (44 females) were rather small for a detailed statistical examination (161–2; cf. Bradley 1978).

A Scientific Approach To Ancient Slavery ?

301

If Scheidel is correct, freedwomen (and, by extension, freedmen, who seem to have preferred to marry within their own status group) would have had little chance to contribute to the genetic makeup of the population (see 1997: 167– 8). This contradicts the arguments we saw in Chapter 1 that the population of Rome became a bastardised slavish rabble. If correct, Scheidel’s model allows us to move beyond the biases of our sources (and, to some extent, the biases of modern historians) towards a more objective picture of ancient slavery. There are, however, some potential criticisms of Scheidel’s approach. William Harris (1999) criticised it for: (1) not giving enough emphasis to evidence for slave-imports in our literary sources; (2) exaggerating the number of children born of slaves; (3) underestimating how many free children might have been exposed. (1) Had Scheidel underplayed the evidence of imports? Probably not. Scheidel accepted that there were perhaps 10,000–15,000 slaves imported each year (1997: 167). This a small fraction of the replacements needed, but still very noticeable to contemporaries and it would explain the number of comments about imports we find in our sources. On the other hand, Scheidel’s arguments against high levels of importation could themselves be attacked. Here is how Scheidel put his case against imports being a high as 40,000 per annum (1997: 164n34): The occidental slave trade to the Americas reached an annual average of 70,000 in the late eighteenth century … More than ten million Africans reached the New World as slaves from 1500 to 1900. The rate of intake for the Roman empire assumed here would have been considerably larger in the long run (at four million per century)—once again, a rather unlikely supposition. Should this comparison rule out higher Roman slave imports? (2) Was Scheidel assuming unreasonably high levels of fertility among slaves? Harris argued that for Scheidel to be correct, slave girls would have had to breed at quite striking rates. He gave several sets of figures for the average number of births required because, like Scheidel, there was an important variable he couldn’t pin down: the ratio of female to male slaves (Harris 1999: 67ff.). The higher the proportion of male slaves, the fewer the number of potential slave mothers, and the higher the number of children required of each mother. The evidence seems to suggest more male slaves than female slaves (though, pace Harris, this might reflect a lack of interest in mentioning women rather than reality). Harris argued that, for Scheidel’s model to work, slave women would have had to produce as many as seven or eight children each. That was,

302

McKeown

perhaps, several higher than the average for free women. Harris argued that this was unlikely. Conditions for slaves may not have favoured access to partners or privacy nor have encouraged the creation of large families (1999: 65ff.). Scheidel recognised these potential problems, but argued that the USA in the nineteenth century showed that the natural reproduction of a slave population was possible (Scheidel 1997: 168–9; cf. 2005: 71–3 and 75). The important question, Scheidel knew, was whether Rome, at least in this respect, was similar to the USA. He believed it was. Harris believed it wasn’t. Harris argued that Scheidel had not taken into account the harshness of Roman slavery. In an echo of some of the debates we have seen in earlier chapters Harris stressed the inhumanity of Roman slavery (1999: esp. 68). Scheidel, on the other hand, argued as follows (1997: 169): In the absence of statistical data, structure and fertility of Roman slave families remain obscure. The evidence from the Roman census returns suggests that slave fertility was similar to that of all free women, though well below that of all married women: Bagnall and Frier op. cit. (n. 16), 158. However, this information comes from households with only a few slaves where family formation may have been difficult; moreover, we must allow for the possibility that some children of slave mothers did not appear in the returns because they had already been sold (ibid., 158n. 85). As for larger slaveholdings in Roman Italy, there is only sparse impressionist evidence: App., BC 1.1.7 takes it for granted that slaves on large agricultural estates would have numerous offspring, a notion that is consistent with Varro, RR 2.10.6 and Colum., RR 1.8.19. The same may have been true for large urban households. Both sides quote ancient evidence to show that breeding either was or wasn’t widespread. We are dragged back into the kind of discussion we were hoping to escape by using demography. Harris’s difficulty, however, is that Scheidel’s US example does at least show that breeding was a possible way of reproducing a slave population. (3) Was Scheidel underestimating the number of exposed freeborn children? He was aware that nineteenth-century Europe had seen some extremely high rates of exposure. He believed, however, that these were unusual, the product of the foundation of foundling hospitals in urban areas. This might have increased the likelihood of exposure, since parents might be more willing to abandon a child if they believed it would survive (165n37). The Romans didn’t have such hospitals:

A Scientific Approach To Ancient Slavery ?

303

… there is nothing to suggest that the modern data could be of much relevance for antiquity. We are here, of course, sucked back into trying to decide which set of potential comparative modern statistics is the ‘exception’—Milan’s 30% exposure rate in the nineteenth century, or Lombardy’s 4.8%. While Romans didn’t have foundling hospitals, they might, nonetheless, have used places to expose children that they expected to be visited. Here the difficulty is Scheidel’s. Harris has shown that high levels of child exposure could have formed an important source of slaves. Harris may have successfully raised a number of doubts about some of Scheidel’s arguments. This did not mean, however, that he was necessarily successful in proving that slave breeding was less important than Scheidel imagined. The problem was that Harris shared some of his opponent’s starting premises. His estimate of the proportion of slaves in the population of the empire, 15% +, was even higher than Scheidel’s (see Harris 1999: 65). The higher the proportion of slaves, the higher the number of replacements required each year, and the less plausible the filling of the gap with exposed freeborn children and with imports (as noted by Scheidel 1997: 165). How could the periphery of the empire supply such numbers? Why doesn’t child exposure figure even more prominently in our sources if it was at very high levels? Scheidel was winning on points. But perhaps we should look at the rules of the game. Scheidel’s position actually rested on a number of key assumptions which were left unchallenged by Harris. He was perhaps right to leave them unchallenged. Changing one or all of them, however, would have important effects on all the impressive statistics quoted. First I will suggest ways in which one could change the rules of the game, but probably shouldn’t (I will give reasons why not). Then I will suggest one very major ‘rule of the game’ that really might be challenged. First, the more ‘minor’ rules. (1) Both authors believed Roman life expectancy was in the low 20s. If it was actually closer to 30 there would be a significantly higher potential reservoir of babies among the free and at the same time fewer slave replacements would be needed. That, however, would seem to run counter to most of the indicators we do have on mortality. (2) If slaves lived as long as free people (or even longer), the number of necessary replacements would also fall. Longer life expectancy for slaves is possible (they were, after all, an investment to be looked after), but it does, nonetheless, seem unlikely.

304

McKeown

(3) Could slave women have produced significantly more children than free women? Again, this is possible, but, given what we know of slave living conditions, perhaps not plausible. (4) Scheidel didn’t argue that all the free population might have exposed babies, only the poorest threequarters. For his intermediate model to work, 1 in 30 children from poor families would have to be exposed to be raised as possible slaves. This is literally ‘1 in 40’ of all children. However, while this makes some difference to the statistics, it is not a massive difference. One might also note also that some women might have regularly exposed their children (reducing the average number who would have done so). (5) Scheidel assumed that more than half all exposed children died. If a significantly higher proportion than 40–50% of exposed children survived to be raised as slaves, however, one would require a much lower number of exposures to satisfy the need for new slaves. That said, such a high survival rate does seem unlikely. (6) Could more slaves have come from outside the empire than Scheidel believed? 70,000 imported slaves a year would produce perhaps a third of all requirements. One might then argue for another third from breeding and a third from exposure. It would, however, as Scheidel noted, have come at a dramatic demographic cost to the populations outside the empire. We also have little way of proving it. We have already seen that there is, however, another, vital, assumption in Scheidel’s arguments. He assumed that 10% of the population of the empire were slaves. Imagine, however, that only 5% were slaves. One would then require only half the number of replacements Scheidel suggested (now approximately 150,000 each year). It would also, however, make exposure and importation more plausible, and breeding less vital, as major sources of slaves (see Harris 1999: 64). 40,000 imports might then produce about a third of the replacements. Exposure rates less than half of Scheidel’s intermediate rate could produce a sixth, with half coming from breeding, bringing the situation closer perhaps to that implied by Harris. The estimate that 10% (or more) of the population were slaves is therefore very important. On what is it based? Very little. There are only two regions of the empire where historians have been prepared to make serious guesses. The first is Italy. In his book Italian Manpower Peter Brunt argued (1971: 124): In my view we could put the number of slaves at about 3,000,000, out of a total population of no more than 7,500,000 (Cisalpina included). This ratio of slaves to free men is extraordinarily high.

A Scientific Approach To Ancient Slavery ?

305

It is indeed high: 40% of the Italian population would have been slaves. That is higher than either Scheidel or Harris initially suggested. It is, however, only an estimate for Italy. Italy was the wealthiest part of the empire and probably the likeliest to use slaves. If Brunt were correct we would, however, already be half-way to the 6 million slaves suggested by Scheidel for the whole empire. Brunt, however, was principally concerned to undermine an earlier attempt (by Beloch) to argue that 2 million was the maximum number of slaves Italy could support. The little evidence he gives for his preferred estimate is surprisingly impressionistic. For example (122): Beloch remarked that Spartacus’ army never exceeded 120,000. In fact the number of slaves who took part in his rising appears to have been rather larger, say 150,000 … The loss of so many slaves must have been a grave blow to their employers, but there is no suggestion in our evidence that the latifundia were denuded of labour, and I incline to think that we should have heard of this, if it had been the case. This leads me to suppose that the rebels constituted only a small part of the total servile labour force. There is also reference to a literary source claiming that perhaps a quarter of the population of Pergamum (in Turkey) were slaves, but this throws us back again to the question of the reliability of literary impressions. How would ancient authors have known these figures? And how widely can we generalise from what they say to the whole of the empire? We need to avoid misunderstandings. There were a lot of slaves in Italy (and particularly in Rome). I am not arguing against the idea that they produced the bulk of the wealth of the elite. I am merely pointing out how soft some of the ‘statistics’ may be, even for Italy. It is worth noting that Wim Jongman (using his own variety of cliometrics) has recently argued that Italian slavery was primarily an urban phenomenon and that the number of rural slaves has been exaggerated (Jongman 2003). Indeed Scheidel himself, following Jongman, now doubts the Brunt figure for Italy, and has argued for a figure perhaps half as large or less (Scheidel 2005, though he may be preparing to lower his estimate of the size of the free population too). The second area where historians have been prepared to make estimates for the slave population has been Roman Egypt. Surviving census statistics indicate that about 11% of the population was slave. Or do they? One might leave aside traditional arguments that Egypt was not necessarily representative of the rest of the empire (a view difficult to prove or disprove). One might also ignore the sparse number of surviving census returns (1,100 people, including

306

McKeown

just over 100 slaves, scattered over several centuries). This probably represents less than 0.01% of the total, but modern polling organisations would not necessarily be unhappy with this size of sample. But how representative is the census material? Most of it comes from just two of the ‘nomes’ or administrative districts of Egypt. Urban areas seem to be over-represented and seem to have had more slaves than rural areas (13% as opposed to 7%). The ratio of male to female (roughly one to two) also seems very odd. It has been argued recently that this is due to the sample being skewed by a small number of households that had a strong bias (for some reason) towards female slaves (Bagnall and Frier 1994: 94). This irregularity is thought-provoking in itself. To summarise: if the slaves made up 10% of the population of the empire, it is highly probable that breeding was the source of the clear majority of slaves and that female slaves were unlikely to be freed while they could produce children. The burden of proof would certainly rest on those who would claim that breeding was not the source of most slaves. But the ‘if in the first sentence of this paragraph is critical. If only 5% of the empire’s inhabitants were slaves, the proportion of home-bred slaves may have been much lower than Scheidel suggested. I have tried to show that the course of the demographic debate remains very sensitive to its starting premises. Those starting premises are in turn easily affected by the kind of messy debates about the interpretation of sources we saw in earlier chapters. The argument between Scheidel and Harris is massively affected by the estimate of the proportion of slaves within the overall population. Ironically, Scheidel himself may be in the process of re-evaluating this number. Whatever its advantages and disadvantages, however, the demographic approach at least operates by debating the limits of uncertainty, and as such it acts as a good metaphor for how the history of slavery ought to be written, even if it isn’t. And it does tell us something important. Even if we reduced the proportion of slaves to 5% of the population, breeding would probably remain the single most important source for slaves. That means the chance of female slaves gaining their freedom during their child-bearing age was probably greatly restricted. We don’t therefore live in a ‘postmodern’ trap of utter uncertainty. On the other hand, we need always to be aware of just how provisional our answers are, and just how much they are determined by conscious and unconscious starting assumptions.

A Scientific Approach To Ancient Slavery ?

307

Bibliography of Works Cited

Ancient Texts

Translations and texts of the ancient texts cited can be found listed under the ancient author’s name in the Loeb Classical Library published by Harvard University Press, with the exception of: Daley, L. W (trans.) (1961) Aesop Without Morals: The Famous Fables, and a Life of Aesop (New York: T. Yoseloff). Perry, B. (1952) Aesopica (Urbana: University of Illinois Press). Watson, A. (trans.) (1985) The Digest of Justinian (Philadelphia: University of Pennsylvania Press). White, R. J. (trans.) (1992) Artemidorus: The Interpretation of Dreams. Translation and Commentary (Park Ridge, New Jersey: Noyes Classical Studies).



Modern Works

Alföldy G. (1972) ‘Die Freilassung von Sklaven und die Struktur der Sklaverei in der römischen Kaiserzeit’, Rivista Storica dell’Antichità 2: 97–129. Bagnall, R. and Frier, B. (1994) The Demography of Roman Egypt (Cambridge: Cambridge University Press). Baisdon, J. P. V. D. (1979) Romans and Aliens (London: Duckworth). Barrow, R. H. (1928) Slavery in the Roman Empire (London: Methuen). Bettini, M. (1991) Verso un antropologia dell’ intreccio e altri studi su Plauto (Urbino: Quattro Venti). Boulvert, G. (1970) Esclaves et affranchis impériaux sous le haut-empire romain: rôle politique et administratif (Napoli: Jovene). Bradley, K. R. (1978) ‘The age and time of sale of female slaves’, Arethusa 11: 243–52. Bradley, K. R. (1984) Slaves and Masters in the Roman Empire (Bruxelles: Latomus). Bradley, K. R. (1987) ‘On the Roman slave supply and slave-breeding’, in M. I. Finley (ed.) Classical Slavery (London: Frank Cass) 42–64; new edn (2003)53–81. Bradley, K. R. (1989) Slavery and Rebellion in the Roman World, 140 BC‒70 BC (London: Batsford). Bradley, K. R. (1994) Slavery and Society at Rome (Cambridge: Cambridge University Press). Bradley, K. R. (2000) ‘Animalising the slave: the truth of fiction’, Journal of Roman Studies 90: 110–25. Bradley, K. R. (2001) ‘Imagining slavery: the limits of the plausible’, Journal of Roman Archaeology 14: 473–7.

308

McKeown

Braund, S. M. (trans.) (2004) Juvenal and Persius (Cambridge, Mass.: Harvard University Press). Brunt, P. (1971) Italian Manpower (Oxford: Clarendon Press). Coale, A., Demeny, D. and Vaughan, B. (1983) Regional Model Life Tables and Stable Populations (New York: Academic). Chantraine, H. (1967) Freigelassene und Sklaven im Dienst der römischen Kaiser (Wiesbaden: Steiner). Dill, S. (1904) Roman Society from Nero to Marcus Aurelius (London; New York: Macmillan). duBois, P. (2003) Slaves and Other Objects (Chicago: University of Chicago Press). Duff, A. M. (1928) Freedmen in the Early Roman Empire (Oxford: Clarendon Press). Dumont, J. Chr. (1974) ‘Guerre, paix et servitude dans les Captifs’, Latomus 33:505–22. Dumont, J. Chr. (1987) Servus: Rome et l’esclavage sous la République (Rome: École française de Rome). Dunkin, P. S. (1946) Post-Aristophanic Comedy: Studies in the Social Outlook of Middle and New Comedy at both Athens and Rome (Urbana: University of Illinois Press). Fay, B. and Pom, P. (eds) (1998) History and Theory (Oxford: Blackwell). Finley, M. I. (1998) Ancient Slavery, Modern Ideology (Princeton, New Jersey: Markus Wiener). Fitzgerald, W. (1989) ‘Horace, pleasure, and the text’, Arethusa 22.1: 81–104. Fitzgerald, W. (2000) Slavery and the Roman Literary Imagination (Cambridge: Cambridge University Press). Fox-Genovese, E. et al. (eds) (1999) Reconstructing History (London: Routledge). Frank, T. (1916) ‘Race mixture in the Roman empire’, American Historical Review 21.4: 689–708. Frier, B. (2000) ‘Demography’, in A. K. Bowman, P. Garnsey and D. Rathbone (eds) The Cambridge Ancient History, vol. XI (Cambridge: Cambridge University Press) 787–816. Gamaulf, R. (2001) ‘Zur Frage “Sklaverei und Humanität” anhand von Quellen des römischen Rechts’ in H. Bellen and H. Heinen (eds) Fünfzig Jahre Forschungen zur antiken Sklaverei an der mainzer Akademie 1950–2000 (Stuttgart: Steiner) 51–72. Garnsey, P. (1975) ‘Descendants of freedmen in local politics: some criteria’, in B. Levick (ed.) The Ancient Historian and His Materials: Essays in Honour of C. E. Stevens (Farnborough: Gregg) 167–80. Garnsey, P. (1.981) ‘Independent freedmen and the economy of Italy under the Principate’, Klio 63: 359–71. Garnsey, P. (1996) Ideas of Slavery from Aristotle to Augustine (Cambridge: Cambridge University Press). Garrido-Hory, M. (1981) Martial et l’esclavage (Paris: Belles Lettres). Garrido-Hory, M. (1998) Juvénal: esclaves et affranchis à Rome (Besançon: Presses Universitaires Franc-Comtoises).

A Scientific Approach To Ancient Slavery ?

309

Genovese, E. (1974) Roll, Jordan, Roll: The World the Slaves Made (New York: Pantheon Books). Giardina, A. and Schiavone, A. (1981a) Società romana e produzione schiavistica [1]: L’Italia: insediamenti e forme economiche (Roma-Bari: Laterza). Giardina, A. and Schiavone, A. (1981b) Società e produzione schiavistica [2]: merci, mercati e scambi nel mediterraneo (Roma-Bari: Laterza). Giardina, A. and Schiavone, A. (1981c) Società romana e produzione schiavistica: [3]: modelli etici, diritto e trasformazioni sociali (Roma-Bari: Laterza). Gordon, M. L. (1931) ‘The freedman’s son in municipal life’, Journal of Roman Studies 21: 65–77. Griffin, M. (1976) Seneca: A Philosopher in Politics (Oxford: Clarendon Press). Harris, W. V. (1980) ‘Towards a study of the Roman slave trade’, Memoirs of the American Academy at Rome 36: 117–40. Harris, W. V. (1999) ‘Demography, geography, and the sources of Roman slaves’, Journal of Roman Studies 89: 62–75. Highet, G. (1973) ‘Libertino patre natus’, American Journal of Philology 94: 268–81. Hopkins, M. K. (1966) ‘On the probable age structure of the Roman population’, Population Studies 20.2: 245–64. Hopkins, M. K. (1978) Conquerors and Slaves (Cambridge: Cambridge University Press). Hopkins, M. K. (1993) ‘Novel evidence for Roman slavery’, Past and Present 138: 3–27. Hunt, P. (1998) Slaves, Warfare, Ideology in the Greek Historians (Cambridge: Cambridge University Press). Hunter, R. L. (1985) The New Comedy of Greece and Rome (Cambridge: Cambridge University Press). Iggers, G. I. (2005) Historiography in the Twentieth Century: From Scientific Objectivity to the Postmodern Challenge (Middletown, Conn.: Wesleyan University Press). Isaac, B. (2004) The Invention of Racism in Classical Antiquity (Princeton: Princeton University Press). Jenkins, K. (1991) Re-thinking History (London: Routledge). Jenkins, K. (1995) On “What is History?’: From Carr and Elton to Rorty and White (London: Routledge). Jenkins, K. (1997) The Postmodern History Reader (London: Routledge). Johnstone, S. (1998) ‘Cracking the code of silence: Athenian legal oratory and the histories of slaves and women’ in S. R. Joshel and S. Murnaghan (eds) Women and Slaves in Greco-Roman Culture: Differential Equations (London: Routledge) 221–36. Jongman. W (2003) ‘Slaves and the growth of Rome’ in G. Woolf and C. Edwards (eds) Rome the Cosmopolis (Cambridge: Cambridge University Press) 100–22. Joshel, S. R. (1992) Work, Identity and Legal Status at Rome (Norman, Oklahoma: University of Oklahoma Press). Klees, H. (1975) Herren und Sklaven: Die Sklaverei im oikonomischen und politischen Schrifttum der Griechen in klassischer Zeit (Wiesbaden: Steiner).

310

McKeown

Klees, H. (1998) Sklavenleben im klassischen Griechenland (Stuttgart: Steiner). Konstan, D. (1983) Roman Comedy (Ithaca; London: Cornell University Press). Kudlien, F. (1991) Sklaven-Mentalität im Spiegel antiker Wahrsagerei (Stuttgart: Steiner). Kyrtatas, D. (1987) The Social Structure of the Early Christian Communities (London: Verso). Kyrtatas, D. (1995) ‘Slavery as progress: pagan and Christian views of slavery as moral training’, International Sociology 10: 219–34. McCarthy, K. (1998) ‘Servitium Amoris: Amor Servitii’ in S. Murnaghan and S. Joshel (eds) Women and Slaves in Greco-Roman Culture: Differential Equations (London: Routledge) 174–92. McCarthy, K. (2000) Slaves, Masters, and the Art of Authority in Plautine Comedy (Princeton, N.J.; Oxford : Princeton University Press). McKeown, N. (1999) ‘Some thoughts on Marxism and ancient Greek history’, Helios 26.2: 103–28. McKeown, N. (forthcoming (a)) ‘The sound of John Henderson laughing: Pliny 3.14 and Roman slaveowners’ fear of their slaves’, in A. Serghidou (ed.) Fear of Slavery, Fear of Enslavement in the Ancient Mediterranean (Bescançon: Presses universitaires de Franche-Comté). McKeown, N. (forthcoming (b)) ‘Resistance among chattel slaves in the Classical Greek world’, in P. A. Cartledge and K. R. Bradley (eds) The Cambridge World History of Slavery, vol. 1 (Cambridge: Cambridge University Press). Manning, C. (1989) ‘Stoicism and slavery in the Roman empire’, Aufstieg und Niedergang der römischen Welt 2.36.3: 1518–43. Martin, R. (1971) Recherches sur les agronomes latins et leurs conceptions économiques et sociales (Paris: Les Belles Lettres). Mazza, M. (1975) ‘Prefazione’ in E. M. Shtaerman and M. K. Trofimova, La schiavitù nell’Italia imperiale: I‒III secolo (Roma: Editori Riuniti) vii‒xxxv. Meyer, E. A. (1990) ‘Explaining the epigraphic habit in the Roman empire: the evidence of epitaphs’, Journal of Roman Studies 80: 74–96. Morley, N. (1999) Writing Ancient History (Ithaca: Cornell University Press). Mouritsen, H. (2004) ‘Freedmen and freeborn in the necropolis of imperial Ostia’, Zeitschrift für Papyrologie und Epigraphik 150: 281–305. Mouritsen, H. (2005) ‘Freedmen and decurions: epitaphs and social history in imperial Italy’, Journal of Roman Studies 95: 38–63. Muecke, F. (1993) Horace Satires II (Warminster: Aris & Phillips). Oliensis, E. (1998) Horace and the Rhetoric of Authority (Cambridge: Cambridge University Press). Parker, H. (1989) ‘Crucially Funny or Tranio on the Couch: the servus callidus and jokes about torture’, Transactions of the American Philological Association 119: 233–46.

A Scientific Approach To Ancient Slavery ?

311

Parkin, T. G. (1992) Demography and Roman Society (Baltimore: Johns Hopkins University Press). Perry, B. (1952) Aesopica (Urbana: University of Illinois Press). Phillips, U. B. (1929) Life and Labor in the Old South (Boston: Little, Brown). Prachner, G. (1980) Die Sklaven und Freigelassene im arretinischen Sigillatagewerbe (Wiesbaden: Steiner). Raditsa, L. (1980) ‘Augustus’ legislation concerning marriage, procreation, love affairs, and adultery’, Aufstieg und Niedergang der römischen Welt 2.13: 278–339. Raskolnikoff, M. (1975) La recherche en Union Soviétique et l’histoire économique et sociale du monde hellénistique et romain (Strasbourg: A. E. C. R.). Roberts, G. (ed.) (2001) The History and Narrative Reader (London; Routledge). Rostovtzeff, M. (1926) Social and Economic History of the Roman Empire (Oxford: Clarendon Press). Ste. Croix, G. E. M. de (1975) ‘Early Christian attitudes towards property and slavery’, Studies in Church History 12: 1–38. Ste. Croix, G. E. M. (1983) The Class Struggle in the Ancient Greek World (London: Duckworth). Sailer, R. (2001) ‘The family and society’, in J. Bodel (ed.) Epigraphic Evidence: Ancient History from Inscriptions (London: Routledge) 95–117. Scheidel, W. (1997) ‘Quantifying the sources of slaves in the early Roman empire’, Journal of Roman Studies 87: 156–69. Scheidel, W. (2005) ‘Human mobility in Roman Italy II: the slave population’, Journal of Roman Studies 95: 64–79. Shtaerman, E. M. and Trofimova, M. K. (1975) La schiavitù nell’ltalia imperiale: I‒III secolo (Roma: Editori Riuniti). Scott, J. (1985) Weapons of the Weak: Everyday Forms of Peasant Resistance (New Haven; London: Yale University Press). Scott, J. (1990) Domination and the Arts of Resistance: Hidden Transcripts (New Haven; London: Yale University Press). Segal, E. (1987) Roman Laughter: The Comedy of Plautus (Oxford: Oxford University Press). Sherwin-White, A. N. (1967) Racial Prejudice in Imperial Rome (Cambridge: Cambridge University Press). Spranger, P. P. (1984) Historische Untersuchung zu den Sklavenfigur des Plautus und Terenz (Wiesbaden: Steiner). Stampp, K. (1956) The Peculiar Institution: Slavery in the Ante-Bellum South (New York: Alfred A. Knopf). Taylor, L. R. (1961) ‘Freedmen and freeborn in the epitaphs of imperial Rome’, American Journal of Philology 82: 113–32.

312

McKeown

Thalmann, W. G. (1996) ‘Versions of slavery in the Captivi of Plautus’, Ramus 25.2: 112–45. Thompson, H. (1993) ‘Iron age and Roman shackles’, Archaeological Journal 150: 57–168. Tosh, J. (ed.) (2000) Historians on History (Harlow: Longman). Treggiari, S. (1969) Roman Freedmen During the Late Republic (Oxford: Clarendon Press). Vernant, J. P. (1980) Myth and Society in Ancient Greece (Brighton: Harvester Press). Vogt, J. (1974) Ancient Slavery and the Ideal of Man (Oxford: Blackwell). Waldstein, W. (1986) Operae libertorum: Untersuchungen zur Dienstpflicht freigelassener Sklaven (Stuttgart: Steiner). Waldstein, W. (2001) ‘Zum Menschsein von Sklaven’ in H. Bellen and H. Heinen (eds) Fünfzig fahre Forschungen zur antiken Sklaverei an der mainzer Akademie 1950–2000 (Stuttgart: Steiner) 31–49. Wallace-Hadrill, A. (1994) Houses and Society in Pompeii and Herculaneum (Princeton: Princeton University Press). Ward-Perkins, B. (2005) The Fall of Rome: and the End of Civilization (Oxford: Oxford University Press). Watson, A. (trans.) (1985) The Digest of Justinian (Philadelphia: University of Pennsylvania Press). Weaver, P. R. C. (1972) Familia Caesaris (Cambridge: Cambridge University Press). Weiß, A. (2004) Sklave der Stadt: Untersuchungen zur öffentlichen Sklaverei in den Städten des römischen Reiches (Stuttgart: Steiner). White, H. (1978) Tropics of Discourse: Essays in Cultural Criticism (Baltimore; London: Johns Hopkins University Press). White, H. (1987) The Content of the Form: Narrative Discourse and Historical Representation (Baltimore; London: Johns Hopkins University Press). Wiedemann, T. (1985) ‘The regularity of manumission at Rome’, Classical Quarterly 35: 162–75. Williams, G. (1995) ‘Libertino patre natus: true or false?’ in S. J. Harrison (ed.) Homage to Horace (Oxford: Clarendon Press) 296–313. Woolf, G. (1996) ‘Monumental writing and the expansion of Roman society in the early empire’, Journal of Roman Studies 86: 22–39. Zanker, P. (1975) ‘Grabreliefs römischer Freigelassener’, Jarhrbuch der deutschen archäologischen Instituts und archäologischen Anzeiger 90: 267–315.

A Life-Course Approach to Household Slaves in the Late Third Millennium BC Laura Culbertson* Despite the frozen social and legal taxonomies of the cuneiform record, slavery in Mesopotamian societies was not an absolute category. Rather, it was “a more diffuse and malleable, less crisply decisive act” (Adams 2010: 2.6; see also Diakonoff 1987). Because of the fixity of terminology, however, learning about slavery beyond the level of linguistics, paleography, and legal language is difficult if not prohibitive in many contexts. The impressive temporal and regional consistency of Sumerian terms such as arad and geme, words we translate as male and female slave respectively, suggests that slavery was a well-defined practice, standardized, and a permanent social category. But society changes regardless of the persistence of writing conventions, requiring historians to question static or inflexible characterizations and definitions of slavery while seeking methods of recovering the dynamics of enslaved people in context. In this chapter I examine the complexities of enslavement trajectories during a short, half-century window of documentation by adopting a life-course approach to household slavery. Sumerian legal, administrative, and economic documents from the twentyfirst century BC in Mesopotamia, more specifically from the period known as the Third Dynasty of Ur (henceforth Ur III), constitute a prodigious body of textual sources that reference facets of daily economic life, making reference to slaves. The historical setting from which our sources originate is a century of rapid political change. In southern Mesopotamia, the twenty-first century BC witnessed a unification of city-states by the kings of Ur, Ur-Namma and his son Šulgi. The degree to which this centralization affected local politics and Source: Culbertson, Laura, “A Life-Course Approach to Household Slaves in the Late Third Millennium B.C.,” in Laura Culbertson, ed., Slaves and Households in the Near East, Chicago: Oriental Institute of the University of Chicago, 2011, 33–48. * This paper is modified from the original version presented at the seminar entitled, “Un-free Children of Elite Households in Early Mesopotamia.” I wish to thank Andrea Seri, Kathryn Babayan, Martha Roth, and JoAnn Scurlock for their comments about this paper at the seminar and in various phases of its progress. I also thank Gil Stein for encouraging the “lifecourse” idea and giving the paper a new angle.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_013

314

Culbertson

society is debatable, complicated by the uneven coverage of written evidence. The nature of the ensuing, short-lived polity consequently also remains the subject of debate. This polity was organized into provinces with capital cities. At the core of the state, the provincial cities housed networks of wealthy households of elites who held important institutional titles and affiliations. Specialists suspect that the kings of Ur did implement some unifying administrative practices and economic patterns, at least affecting local record-keeping practices in different provinces. Midway through the forty-eight-year reign of Shulgi, about halfway through this political experiment, an explosion of cuneiform documentation occurred in the provincial cities of the Ur III heartland. Most of the documentation selected for this study originated during the reigns of Šulgi’s sons and successors, the kings Amar-Sin and Šu-Sin, dating to a roughly twenty-year window. Textual production declined in the early years of their successor, the final king Ibbi-Sin, shortly before the poorly understood and unevenly documented disintegration of the state at the end of the century. These historical, political, and administrative circumstances result in a uniquely high volume of documents covering a short period of several decades, barely one generation, presenting an opportunity for microscopic examination of slavery during a restricted moment in history. The cuneiform documents used for this study primarily concern economic and administrative activities of the affairs of elite, private citizens with institutional affiliations and titles. In the provincial capital cities of Umma and Lagash, for example, documents show that the urban elites, including members of the households of governors, participated in local systems of dispute resolution, negotiating professional and household investments and entitlements. Because the results of resolution proceedings affected networks of relationships, these resolution systems can be considered local political orders, reflecting the ever-shifting distribution of power among the urban community (Culbertson 2009). Even though most elites engaged in a range of activities and businesses during this period, the administrative and legal language of documents condenses information, collapsing many aspects of social life and interactions into succinct reports. The documents are abbreviated, formulaic records of brief flashes of events, largely reflecting the interests and objectives of urban elites and their household assets. For this reason, social and political dynamics during this period, such as those between slaves and households, are difficult to reconstruct. One traditional method of studying slavery using legal sources from the cuneiform world, which we may call the synthetic approach (VerSteeg 2000), involves identifying attestations of the terminology associated with slavery in the documents and compiling the attestations into surveys. Syntheses of these data may be organized according to a legal question (e.g., sale, manumission,

A Life-Course Approach to Household Slaves

315

marriage), a type of source (e.g., Law Code entry, receipt, letter, court record), or chronology.1 Scholars, linguists, and historians can use the synthesis of data and attestations to make inferences about the terminology or legal question under investigation, or locate analytical problems by noting gaps, inconsistencies, and idiosyncrasies. Importantly, information compiled in a synthetic approach can be distilled into definitions or descriptions about the legal and economic context of enslaved people, allowing scholars to overview the possible range of attested activities associated with slaves in particular contexts. This approach is additionally invaluable for tracing the longevity and expanse of written traditions across time and place. Despite the profitable findings of synthetic studies of slavery, a danger of flattening persists. Even if administrative and legal terminology retain a common standardization across time or across a politically or culturally unified region, society and the personal relationships operating within it cannot be standardized to such an extent. Words such as arad and geme last for centuries but meanings and contexts change, and the ascription of a legal designation (e.g., “slave”) to a person in an economic or legal record may thus have little reflection of daily interactions. As Adams has recently put it, “the inclusion of an individual in the formal slave category might or might not establish that individual’s breadth of prescribed activities, treatment, or even social position” (2010: 2.3). The synthetic approach can focus on individuals only in the moment reported by a single document or legal problem and cannot explore variations and idiosyncrasies in matters such as how people entered or exited enslavement. As a result, slavery is condensed to one social category, generalized from terminology and from the empirical observation that slaves are listed in the records in connection with an owner or overseer, and therefore slavery must be a generally downward experience. To further understand what household slavery looked like in the Ur III period, how this status, predicament, or transition was established or transmuted, and how the wider context of private household politics and dynamics figured into the process, I propose to experiment with a life-course approach to household slavery. A life-course approach (taking inspiration from Manning 1990: 113) focuses on the transitional moments in a person’s life and involves compiling series of transitions into larger chronologies. Using the approach, we can highlight the varieties of slavery, leaving a bit more room to notice the mutable aspects as well as the permanent. The life-course approach looks at events in people’s lives using multiple kinds of documents to map out the 1  Invaluable examples containing this approach can be found in VerSteeg 2000; Westbrook 1995 and 1998; and Wilcke 2007: 53–58.

316

Culbertson

range of possible courses intersecting with the key moments captured by the written records. These include the creation of enslavement through sale, transfer, manumission, and any other inferred or mentioned events occurring before, between, or after these phases. To gain a meaningful perspective on these events in the lives of the enslaved, we can align life-courses with the events and transitional moments of households, finding pivotal correspondences between the two. In short, instead of defining slavery by the use of legal terminology and reconstructing information about slavery from single terms or formulae, the life-course approach allows consideration of slavery as “an interactional process,” to use one aspect of Orlando Patterson’s definition (1982: 13). As is already known, people enslaved in elite Ur III households were often not born into slavery, nor did persons who were born into household enslavement necessarily remain slaves for the duration of their lives, nor did all slaves of private households die in slavery. Collecting a range of life-courses will help us better understand these variables. The life-course approach first involves a reassessment of how to look for slavery in the documentation given its limitations and problems. Ideally, we would have numerous forms of documentation detailing the vagaries of every life trajectory, but such comprehensive documentation is lacking for any person who lived during the twenty-first century BC, let alone any enslaved person. Access to life details about household slaves, outside of the parts of their lives spent in slavery, are generally non-existent and must be inferred or reconstructed from sources if pertinent information is provided. Not all documents mentioning slaves can fruitfully be included in this approach if they do not provide contextual information about the enslaved person or people in question. Prosopographical research allows the tracing of individual people across multiple texts in a few instances, but enslaved people are rarely identified by name in more than one document. We are thus woefully reliant on textual synthesis, survey, and description, but the comparison of multiple categories of documentation, telling about multiple stages of enslavement, opens up different views and allows us to approach slavery as a process. For the sake of manageability in this short contribution, the discussion is restricted to instances of children or unmarried young adults who were sold into slavery by their parents or family members. This is the form of enslavement creation for which we have the most evidence, with multiple types of documentation referencing the events surrounding the sale and enslavement, providing a longer picture. These principal sources are sale documents and court records, both of which bring limitations and benefits. Importantly, both types of documents deal with transitional moments for both slaves and households. Sale documents record the moment of a person’s enslavement and his

A Life-Course Approach to Household Slaves

317

or her transfer between households, reported in the context of an economic transaction between two parties. A sale itself may represent the fragmentation of the household from which the enslaved child came. The persons sold, exchanged, or enslaved via these transactions were often children or unmarried young adults. Extraction of generalizations about the nature of slavery from these documents alone has been justifiably avoided for two reasons. First, the moment of sale or exchange constitutes only one moment out of the term of enslavement, and second, we can safely assume that the newly enslaved children and young adults grew up and experienced other things beyond the single moment of sale. Slaves are also mentioned in court records, the second type of documentation consulted for this study. Court records are reports about the dispute resolution proceedings of urban elites. Household turmoil and transition were often the catalyst for such disputes. At these occasions, enslaved household members could exploit status ambiguities and household fractures to argue for release, sometimes changing the course of their enslavement. Because the court records make reference to past events and time, we can deduce the length of enslavement for certain people or infer the origins of their enslavement. It is often said that Ur III society was rigidly hierarchical, offering few opportunities for mobility, even among free persons. A life-course examination of sale documents and court records, however, shows that slaves did travel through different social contexts over the course of their lives, for better or worse, and by tracing these paths we find that the very sale and enslavement of children and adults could send them in a variety of directions. While some slaves spent the duration of their enslavement, and possibly lives, protesting their status, there is evidence that others remained in their owner’s household. In either case, household slaves enjoyed social privileges and affiliations not accessible to most members Ur III society, including access to court. As Neumann has overviewed (this volume), the households of Ur III elites were multifaceted and complex.2 In addition to the household family, which could consist of a household head, his wife, their children from one or multiple marriages, and sometimes their siblings, each household could also host a combination of house-born slaves and their offspring, purchased male and female slave children, and other kinds of tenants, employees, laborers, and subordinates. The mixed make-up of these environments meant potential for interactions between persons of different statuses and backgrounds on a daily basis, and suggests that the experience of slavery in a household was 2  I refrain from offering a composite or archetype household structure model to avoid generalization.

318

Culbertson

conditioned by a variety of variables, contingencies, and relationships, even if we know little about the degree of harmony or discord among household members. This preliminary application of the life-course approach will show that the lives of slaves in households followed no single trajectory and a variety of conditions, circumstances, and contingencies must have factored into these situations.

Sale Documents: Children into Slaves

We start with the issue of how people entered into elite households as slaves, either by birth to enslaved parents or by sale from their parents. It should be noted before embarking on this topic that “child slavery” is not a viable concept for studies of early Mesopotamian societies. The study of “child slaves” in specific societies is a productive branch of slave studies, warranting specific theoretical examination of children in slavery after other groups (e.g., women) have been isolated for study, reducing the emphasis of slave studies away from the long-standing focus on adult men (Campbell, Miers, and Miller 2009: 1–18). While many societies and historical contexts offer suitable justification for positing a phenomenon of slavery specific to children,3 a study of “Ur III child slavery” does not qualify because children in this period were neither a legal category nor a specific segment of the labor force that could be isolated into a single category. These points are only worth mentioning to preface the argument that the courses traveled by enslaved children in Ur III households were predicated upon the events, decisions, and circumstances of their families and the households to which they belonged, and were not predetermined by rigidly demarcated economic or legal classifications based on concepts of child or minority status. Enslaved and low-ranking non-slave children in Ur III sources were a substantial part of the manual workforce. It is well known, for example, that women and children comprised the backbone of the textile industry (Waetzoldt 1972). Many thousands of children are mentioned in labor rosters and ration records of institutions, listed as recipients of subsistence rations disbursed to laborers. These children, often identifiable as children because of the reduced quantities of subsistence rations they received relative to adults, were listed in ration records alongside their parents or as groups of siblings. Children could also be 3  See the arguments about the existence or non-existence of “child slavery” in Roman society offered in Laes 2006.

A Life-Course Approach to Household Slaves

319

listed without mention of having families, or as groups of nameless, orphaned boys who were seized and forced to work for a temple institution.4 However, non-slave children appear in these administrative records in the same fashion, receiving remunerations for menial labor performed alongside their parents or other adults. Determining whether administrative rosters are dealing with slaves or dependent but essentially “free” laborers is difficult.5 In either case, most of the administrative sources from the third millennium BC do not unambiguously demarcate a labor force specific to children, indicating rather that free and un-free families performed the same unskilled jobs for comparable rates, side by side (Studevent-Hickman 2006: 137, 161ff.). Legal sources from Ur III society also did not postulate a formal category for children, nor do they suggest a category akin to the concept of minority. In practice, children were the property of their parents, whose socioeconomic status they typically inherited. The rights parents held over their children are evidenced by the reality that both poor and wealthy parents were entitled to sell their offspring in a financial crisis. Within the households, children of household slaves were considered the property of the household owner, his wife, or his heirs, inheriting the status of their parents. Given all these issues, I use the term “child” to indicate an unmarried dependent who lived with his or her parent(s). Children entered household slavery by birth or sale. Ur III sources make a terminological distinction between house-born and purchased slaves, but usually not in the context of court records or sale documents. Children born to household slaves were legally considered the property of the household head, who theoretically retained rights to sell or free them. This right was often challenged at certain phases of the child’s life or the household’s history, as is discussed in the following section. In the context of insurmountable financial crisis, debt, or obligation, enslavement could be established to address the situation. In some instances people sold themselves into slavery to mitigate their debts (e.g., NG 38), but most enslavement-sales involve children. Cases of child sale highlight the dynamics between household and families because they concern the physical and legal transfer of a child from his or her blood family to a new household. Other than marriage, this was another means in Ur III society by which a person was separated from blood relatives and attached to a new family. As indicated above,

4  For example, on the dumu-dab5-ba “seized children,” see Gelb 1979 or Heimpel 2009. 5  See Tenney’s discussion of this problem in this volume.

320

Culbertson

enslaved families in institutional slavery were more often kept together,6 while children without families, having been abandoned or orphaned, worked in labor gangs under the authority of institutional overseers and officials. The records of sale-enslavements, including contracts and receipts, provide only the most basic of information about these transactions and leave aside the matters of circumstances, motivations, and the respective assumptions of buyers and sellers. The records are written with the interests of the buyers in mind, and, despite any illusion of their formulaic, legal character, they cannot be approached as neutral compositions equally representing the interests of all parties. Steinkeller (1989: 8ff.) has already outlined the formulary of Ur III sale documents and discussed the problems of identifying patterns in their construction, but a brief summary is useful here. Sale documents regarding household slaves always provide (1) an identification of the slave by name and blood parentage, the name of the buyer, and the name of the seller; (2) the price of the slave; (3) an indication of the status of the payment; (4) the names of witnesses to the sale and authorizing officials, if present; and (5) the date. The documents may additionally provide some combination of the following information: (6) a no-contest clause, in which either the seller promises not to deny or contest the sale at a later date or both buyer and seller reciprocally promise not to overturn the sale; (7) a record of any oaths taken to bolster such clauses or cement the sale; (8) a legal clause referring to symbolic or ritual acts performed at the time of the sale; (9) highly abbreviated summaries of any terms and conditions about the sale or the duration of enslavement; and (10) the name of a guarantor who was present for the sale. Thus the bulk of information on these tablets concerns names of participants and clauses protecting the buyer’s investment, while little or no information concerning the characteristics or condition of the slave or specific circumstances leading to their sale are provided. Consistent with the general nature of Ur III sale records, records of slave sales presuppose that the sale was a completed, irreversible transaction. Despite the limitations of these documents, we can reconstruct a preliminary picture of the events surrounding the sale of children, the context in which this occurred, the creation of enslavement, and transference to a new household. According to records of non-institutional slave transactions, parents pledged or sold members of the family when faced with financial emergency, and they most often selected their unmarried, especially female, 6  On the evidence of laborers as families, see Studevent-Hickman 2006: 137, 161ff. See also, for examples of sources, Iraq 41 125 2 (record of various marsh workers employed by a temple, listing groups of up to three sisters, organized by the siblings’ patronymic) or the document mvn 6 538.

A Life-Course Approach to Household Slaves

321

children. Loan contracts show that parents pledged specific children as collateral for advances; court records indicate that creditors could and did lay claim to pledged children, taking them into their own households, when the conditions of a contract were breached.7 In other situations, indebted men and women made contractual agreements with their creditors to temporarily loan children to work off the debt over a fixed amount of time (e.g., Molina 2008: no. 2). Some documents explicitly state that the child was purchased to perform domestic tasks for his or her owners (e.g., UET 3 51). In a presumably more permanent transaction, a parent could sell members of the family outright and collect payment in exchange. In general, the sale of children into slavery is largely associated with financial upheaval adversely affecting a household and requiring parents to implicate their children in resolving the obligation or debt. Thus two asymmetrically linked households were involved in these transactions in some manner. Both mothers and fathers sold their children in the Ur III period. In the cases of mothers, there is too little evidence to determine if these sellers were widowed or perhaps forced into hardship after abandonment, but the possibility exists.8 According to one court record (NG 44), a woman disputed the enslavement of her whole family, husband included, suggesting that women played a decisive role in matters of slavery, even if married (Siegel 1947: 13). Siblings and grandparents are also attested as having sold female children; one document (NG 55) reports that a grandmother participated in the sale of her female grandchild. We may surmise that the father is deceased or absent in these cases, but the circumstances are regrettably unclear (see Siegel 1947: 23). Even though sale records report the name of parents and sellers, few of these people are identifiable through prosopographical investigation, further obstructing an understanding of the circumstances and contexts of child sales. Several documents contain seal impressions of the mother or father (e.g., faos 17 23, 46, 49, and 78) suggesting that the sellers in these instances had socially recognized institutional affiliations and titles, even if their names are untraceable to us now. This indicates that not all families who sold children were always impoverished and lowly. By contrast, some buyers can be identified owing to a prodigious presence in the Ur III legal and economic documentation, including the well-known entrepreneur si.a-a to name but one example.

7  Examples of these lawsuits are found in the documents ng 35, in which the disputant is a girl sold by her mother, ng 37, in which a litigant is a girl sold by her father, and ng 204, concerning a father who sold his daughter. 8  Receipts of sale by mother: faos 17 42, 46, 49. Disputes involving the sale of children by the mother: ng 35, 53, 55.

322

Culbertson

The sale itself was a formal event that usually took place at the household of the buyer (Steinkeller 1989: 41). The transaction involved the physical participation of the sellers or parents, multiple witnesses, and a person who would act as a guarantor.9 The purpose of this temporary function of guarantor is never made explicit, but from the sources we can infer that he or she was somebody who vouched for availability of the child, especially promising that the child had not been pledged to another party. The guarantor also acted as a hypothetical co-seller, inasmuch as he or she could be implicated along with the seller in the event of any conflict following the sale. Often the guarantor belonged to the nuclear family of the child, perhaps a parent or sibling (e.g., faos 17 46) who promised to personally replace the child as a slave in the event that another household issued a claim. Guarantors are less frequently named in sale records involving adults or persons who were already enslaved before the time of sale, as far as can be determined from the available evidence. The function appears to be associated with sales involving the creation of slavery and is thus restricted to the sale of children or unmarried young women, who, legally speaking, had the same status as children.10 The need for such a function in the sale of children and unmarried women also perhaps suggests that child sale was not an option to most debt-ridden families and occurred only under very specific circumstances. Presumably, buyers and creditors were wary of buying a child from a family in dire financial circumstances given the likelihood that other creditors might attempt to claim the child. At the event of sale, payment for the new slave was made in full, with occasional exceptions.11 Ownership of the child, and the child him- or herself, was then transferred or delivered to the buyer and his household. Ritual acts accompanied this phase of the sale and symbolized the physical transfer of the slave to the new household. One sale receipt from the city of Nippur uses an expression of legal symbolism, stating that the buyer made the enslaved girl 9  Sumerian lu2-gi-na-ab-tum2, on which see Falkenstein 1956–57: vol. 1, p. 126; Steinkeller 1989: 80–90 and his commentary on text 127. 10  See faos 17 25 and 33 (involving the sale of an unmarried woman). foas 17 33 includes gab2-gi-in sag-kam “guarantor of the slave,” the function of which is unclear. Steinkeller (1989) renders the term, “guarantor of the head,” understanding sag as “head” rather than “slave” and perhaps alluding to the guarantor’s job of attesting to the quality and health of the slave. 11  faos 17 25 deals with the sale of a young, unmarried woman, stating that, “out of the price, 2 shekels are paid” (line 7). The court records ng 176 cases 1 and 2 deal with settled disputes over the non-payment of money for slaves. See also ng 63, 65, 66, 131, and 207 case 1, for other examples from Girsu, and ng 48, 49, 51, and Molina 2008: no. 9 for examples from Umma.

A Life-Course Approach to Household Slaves

323

“cross a wooden pestle,” representing the full transfer.12 Indeed, many of these sales did involve the crossing of socioeconomic and political boundaries for these children from an insolvent house to one with the means and standing to collect. The sale records imply that, at some moment during the transaction, the parties swore oaths promising not to contest the status and transfer of the child in the future. When the sale activity was complete, a document was drawn up to record the pertinent information and the transaction was theoretically complete. Unless specific provisions were drafted to release the child at a certain time, any children born of the slave would traditionally be considered the property of the buyer and his household. Many of the documented slaves did eventually bear children. In the case of young women, it is likely that the children were fathered by the household head himself, evidenced by cases in which slave women and household heirs disputed the status of her children after the death of the household head (see below). A final observation about sale records is that the contracts pertaining to child sales imply a buyer’s indifference about the child personally, which by extension indicates a dislocation between the child and his or her ability to perform service for the household from the buyer’s point of view. For example, a sale document from Nippur reports that a girl named Allani was sold into enslavement for 5 shekels by her brothers. The document stipulates, “they (i.e., her brothers) will be enslaved if she ceases to work” (faos 17 45). Despite being named in the document, the girl was regarded as interchangeable by her new owner, at least at the time of sale or from the point of view represented in the document. In summation, records of child sales suggest that enslavement could emerge from two situations aside from birth. In the first case, parents sold or pledged children when the household encountered a financial emergency, seeking a remedy by turning to another household to make the sale or collected an advance. In the second, children were implicated in enslavement when members of two different households engaged in financial or business relationships and one party could not uphold his or her side of the contract. The records of child sales indicate that the sale of children and young adults was a one-way and permanent transaction, severing the blood ties of the children, stripping them of their birth status and family affiliations, and transforming them and any potential offspring into what we might call household “chattel.” If we continue to 12  faos 17 41 line 8: giš gin7 i3-na-ra-bala. The document reports the sale of an unmarried woman, with a guarantor present. See Steinkeller 1989: 34ff. on this clause, which appears in a number of slave sales.

324

Culbertson

trace the life-courses of people pledged or sold into enslavement and supplement this picture with the information from court records, we discover a more complicated picture of slavery.

Court Records and Contestations of Life-Courses

Court records from the Ur III period contain summaries of public proceedings during which disputing urban elites presented cases before third parties in pursuit of resolution or settlement. Like the sale records, these documents strictly adhere to an administrative template, presenting the following sets of information in part or full: (1) the names of disputants; (2) the object, person, or entitlement under dispute and stock declarations or statements made by the disputants; (3) the prehistory of the case, with mention of earlier court sessions, events, or transactions; (4) the evidentiary measures presented by the winning party; (5) a decision or settlement, if one was reached; (6) provisions or stipulations about the settlement; (7) oaths, usually cementing promises among the parties not to return to court again; (8) lists of officiating parties, including judges and other high-ranking figures from urban society; and (9) the date.13 Of course, court records pose the same limitations as sale records, reporting only brief periods of activity out of what were often prolonged, labored conflicts and cumbersome household or family transitions. Disputes could persist for decades, with disputants resurfacing in courts multiple times and before several groups of arbitrators in pursuit of resolution or a confirmation of conditions. Although conflicting perspectives are evident in the dispute and audit records, the presentation of information in court records is one-sided, with emphasis on the party who won the case. The scribes who composed these records reduced complex networks of people and interests to basic casts of characters, conflating elaborate conflicts into stock paradigms of buyer versus seller, slave versus master, or brother versus brother. Consequently, the ambiguities and events that sparked disputation are not readily apparent in the documents and must be inferred by a close reading of the case and the references to past proceedings and events. Judging from past events mentioned in the texts, many disputes emerged within the context of public audits of massive household estates.14 Both dispute 13  Overviews of Ur III records and their construction are available in Falkenstein 1956–57; Molina 2000; and more recently Culbertson 2009 for a full list of publications. 14  Examples include ng 99, 211, and 213.

A Life-Course Approach to Household Slaves

325

proceedings and household audits corresponded to household transition or turmoil, death of the household head, financial emergency, or other kinds of crises that demanded evaluation of the household assets and a public clarification about the precise status and entitlements of household members, including the slaves. Resolution and audit proceedings were therefore consequential social acts by which household members or assets were reconfigured, dissolved, or dispersed into new households. Among the issues of contention during household transitions and disputes were slaves. Despite the unambiguous language and finality expressed in sale records, ensuing confusion and controversy about the exact terms and conditions of slave sales seems to have been common because the majority of documented court cases from this period—possibly upwards of 65 percent (Culbertson 2009)—deals with disputes about the status, sale, and ownership of domestic slaves.15 The court documents provide a partial view of what happened to children who were born or sold into enslavement with private households, showing that a variety of conditions and social courses could result from household enslavement. Broadly speaking, slaves appear in two capacities in the documents of relevance to our discussion. In the first situation, slaves were implicated in court proceedings as the subject of disputes between two or more elites, such as a buyer and seller from different households or between siblings from the same household who stood to inherit after the death of their mother or father. Despite the fact that people retained written records of sales and contractual agreements about the slaves, the court records indicate frequent disagreement and misunderstanding over the terms of sales. Amid the confusion, parties present for sale transactions and oaths, including guarantors, reconvened before judges to argue about the legitimacy and conditions of enslavement.16 In some disputes, the point of conflicts was whether the sale or payment occurred (e.g., NG 196 case 1; see also Molina 2008: no. 10), or whether the slave was delivered to the new household after the payment was made. Court proceedings were used to establish that both payment and delivery occurred in attempt to prevent further claims and misunderstandings (e.g., NG 176 case 2 or NG 207 case 1). 15  See Falkenstein 1956–57: vol. 1. This high percentage is partially explained by the fact that there are no lawsuits or court records dealing with immovable property, which predominate in records from the adjacent Sargonic and Old Babylonian periods. 16  In ng 45, in which the sale record is produced in court to demonstrate the existence of sale witnesses followed by the summoning of those witnesses, and Molina 2008: no. 8; see also ibid., no. 3 for a registration of the death of the guarantor, possibly drafted in anticipation of future conflict.

326

Culbertson

As sellers, blood parents of enslaved children attempted to retract the sale of their children or claim that they never received payment. For example, NG 45 reports that a woman appeared before judges and denied selling her daughter and receiving 5 shekels in payment, but witnesses and written documentation overturned her claim; the judges made the woman swear an oath promising not to claim her daughter again. NG 47 similarly reports that a man named Kamu denied receiving 6 shekels as payment for his daughter,17 but the verdict of the case is broken. In the court record NG 48, from Umma, a man denied selling his son for the relatively small amount of 2/3 shekels of silver eight years after the fact, but the court denied the claim for unspecified reasons. At stake in some disputes was the matter of whether enslavement explicitly involved labor, the requirement to work off a fixed amount of debt, permanent residency in the household as a dependent, unfree servant, or ongoing performance of some other unspecified function (NG 7; Sigrist 1995: 612 and no. 2). For example, one dispute record reports that, even though a young enslaved woman somehow paid off her debt and was therefore relieved of enslavement, she was still to remain attached to the household in a transitional status while performing domestic service until her owner’s death.18 The case reported in Molina 2008: no. 2 indicates that a woman left her enslavement without completing her term and headed to another household. The sellers therefore had to refund the 5 shekels paid for her. The death of the household head who purchased the slave was of central concern to this question as many disputes center upon the question of whether the enslaved child was designated to remain in the household until the owner died (e.g., NG 58). Disputes among family members about the inheritance of household slaves are also among the cases concerning slaves. In attested cases, widows fought their sons, daughters, or stepchildren over who would inherit the slaves of the deceased husband/father (NG 28), and one case involves investigation of which person bought the slave (NG 99). In the second situation under which slaves appear in court records, slaves are actors in the dispute, appearing in court on their own behalf. Disputes over the inheritance patterns of slaves and their children were common and surfaced in court upon household transitions brought on by the death of the household head. NG 7, 33, and 34, for example, involve disputes between slaves and their masters’ heirs. In these examples, male and female slaves argued before judges 17  Probably five to six years before, given the partial names of judges in lines 10–11 who were prevalent during the middle of Šhu-Sin’s reign (Culbertson 2009: 127). 18  uet 3 51. On this matter, see Falkenstein 1956–57: vol. 1, p. 95. See Kleber’s discussion of paramone, this volume.

A Life-Course Approach to Household Slaves

327

that they were to have been freed upon their master’s death, but were unable to corroborate their claims (via witnesses or written documents) in order to prove that their owner had indeed provided freedom upon his death. In similar cases resulting from the death of the household father, widows were prompted to dispute their husband’s heirs over the fate of household slaves (e.g., NG 7, 28, 83, 99), usually implicating witnesses to the purchase of the slaves or agreements struck with the household head over inheritance plans. In NG 37, an allegedly enslaved girl named Ninzagesi pled before judges that her father did not sell her. The girl was confirmed to enslavement after witnesses contradicted her testimony and her owner swore an oath declaring that the sale took place. Similarly, NG 35 reports that a girl named Etamuzu challenged her enslavement to the household of a high-ranking cook, reportedly stating, “I am not your slave-girl (geme)” before the judges (line 9), only to be contradicted by her mother, Atu, who also attended the proceedings and corroborated the sale of her daughter (see NG 36 for a similar case). Several court records contain stock first-person denunciations of personal enslavement, purporting to give the voice of a disputing slave with the construction: “The slave so-and-so appeared before judges and said, ‘I am not a slave’” or “I am not so-and-so’s slave.”19 These declarations are intriguing given that the entitlement to participate in the urban court system as a disputant was not guaranteed to all members of Ur III provincial communities, let alone to slaves. Whether or not these slaves were indeed afforded the opportunity to represent him- or herself in real-life proceedings will remain unknown to us; both the Ur III dispute resolution system and the textual documentation it yielded were intended to protect the interests of elites. Yet there are several references in the court records to slaves who claim to possess sale records or other forms of written proof attesting to the conditions of their enslavement, and there is at least one case in which a slave girl actually produces a document in court (NG 205; see also Sigrist 1995: no. 1). Many documented court sessions occurred long after the sale and, in many instances, the slave must have become a grown adult by the time they appeared before judges. The court records make reference to the dates of these alleged sales, and, in a few instances where dates are cited we can reconstruct the amount of time elapsed since the child was sold or enslavement was established:20

19   n g 32, 33, 34, 35; BM 106451; and Sigrist 1995: no. 1. 20   n g 41 deals with enslavement as punishment for murder, not sale.

328

Culbertson

Court record

Documented period of contested enslavement

ng 31 ng 34 ng 41 ng 48 ng 65 ng 67 ng 70 ng 71 ng 88 ng 192 ng 205 case 1

20 years 15 years 5 years 8 years 10 years 6 years 4 years? 13–14 years 10 years 20 years 36 years

It should be noted that these figures represent only the time elapsed between the cited date of the sale and the cited date of the specific court document that concerns a dispute over the sale. The actual duration of enslavement, conflict, or ambiguity may have been much longer than these figures admit. In most of these cases, the lag between sale and disputes represents the time between initial acquisition of the slave and the death of his or her buyer. In the context of public disputes and audits and corresponding upheaval of the household and competition among members, enslaved members of the household seized opportunity to contest their enslavement, to varying degrees of success. The death of the household head prompted slave and nonslave family members to determine inheritance patterns and articulate the precise ownership of the household assets, including the slaves. This must be the case with Irišaga, an enslaved man who was purchased and then given to the buyer’s son. Thirty-six years later, the buyer having died, his other children sued their brother for rights over Irišaga, taking the matter before governors at least three times before they were finally compelled to cease (as far as we know). In other situations, the conflict seems to have been ongoing over the years, as in the case of the male slave Ahuma, who is cited as having protested his enslavement in court at least three times (see NG 33 and 34) over a course of more than fifteen years. It seems facile and anachronistic to assume that enslaved disputants spent years contesting their enslavement in pursuit of an abstract notion of

A Life-Course Approach to Household Slaves

329

“freedom.” A variety of other reasons for contestation can be hypothesized. For example, some enslaved people seem to have retained the identity of their former household even after sale or transfer, possibly harboring connections to their blood families for decades after the separation. The court records indicate an ambiguity of slave status, identifying household slaves according to their blood parents (i.e., sellers), often21 providing patronymics of male slaves even if also identifying the owner. Ahuma, for example, is called “Ahuma son of Lumarza, slave (arad) of Kudamartu” (NG 33). Female slaves are given a partronymic if their father is alive; if a mother sold her daughter, no patronymic appears in the court records and the relationship is only clear by context. In addition to maintaining previous social and legal identities, other bonds may account for the motivation to leave enslavement. One court record involves a young man’s unsuccessful attempt to free his enslaved sister, implying that familial ties were not severed during the sale (Molina 2008: no. 7). In addition to disputation in court, another manifestation of confusion over the nature and terms of enslavement involves escape. Court records make reference to slaves who fled house, city, and region, and one court record reports that a man escaped as far away as Anshan in Iran, where he was found and captured (Molina 2008: no. 4). Other cases of flight are less extreme and involve situations in which the slave simply left the household under the impression that his or her service had been completed or was supposed to be temporary (Molina 2008: no. 2). Escaped slaves were returned if found, and court proceedings solidified the (re)enslavement before the community. Clearly, many purchased domestic slaves at least fostered an expectation of return and there were different methods for pursuing freedom. Returning to the issue of house-born slaves, children of enslaved parents were also considered slaves. Despite having been born into the status, children of household slaves nonetheless contested their enslavement in court and this was also undertaken when the household head died. The results of these cases varied. In one case, for example, a male slave lost the dispute because, according to the document, he was born and raised in the household (NG 32). In other similar situations, the children of purchased slaves were freed as a result of a plea in court (e.g., NG 99, NG 205 case 3, and Molina 2008: no. 5). Among these instances, however, the parents of the newly freed children were rarely freed as well, and thus the group remains with the elite household as a subsidiary family of mixed statuses. Not all enslavements involved futile attempts to detach from a household. Children sold to private households could reach eventual release, and several 21  Not always: ng 45.

330

Culbertson

records attest to the termination of slave status. One record indicates that a slave was ultimately returned to his or her parent’s household while still a child (e.g., Molina 2008: no. 5). In other situations, however, the released slave was awarded the status of a freeborn adult. In NG 75, for example, an adult man named Ursagub and his children are declared free before the court and community, and the text explicitly states that Ursagub is “returned” to the status of a free citizen (see also NG 76).22 In other examples of this kind of “manumission,” the freed slave remains in his master’s household, where he is regarded as eligible for inheritance of household property upon the death of its head. Conclusions Reading two corpuses of sources against one another, on the basis of these piecemeal excerpts of enslavement, we find that the initial enslavementthrough-sale of young, unmarried individuals could result in years of uncertainty, resistance, and dispute. Misunderstandings about the fate of household slaves is evident in the exhaustively preemptive language of the sale documents, while records of court proceedings confirm that disagreement was nonetheless frequent, not only among slaves and owners but also among nonslave household members. Some enslavements ended with the transformation of the slave into a semifree member of the household or a free person who was eligible to start a new household. It is not possible, given all these examples and possible results of enslavement, to characterize the sale of individuals into slavery as a full “social death,” as Orlando Patterson (1982) claimed. That is, household enslavement did not necessarily entail the loss of family ties, identity, social affiliations and opportunity, and honor, as demonstrated by the court documents discussed here. While the condition of household enslavement did not guarantee upward mobility, it did afford slaves the opportunity to participate in a restricted court system and occasionally led to a non-slave status with elite affiliations. Above all, the sale of children into slavery could propel the slave through different social contexts over the course of his or her enslavement, and in some situations, the slave had a degree of control over their fate if he or she seized opportunity to argue in court at key moments in the household’s history. In short, the different paths taken by un-free children in elite households were varied and defy any universal characterizations.

22   n g 75 lines 22–23, literally, “as if he is ‘returned to the previous status of freeborn.’” On this terminology, see Lafont and Westbrook 2003: 197.

A Life-Course Approach to Household Slaves

331

A number of variables and conditions seemed to have directed the course of the household slave. On the one hand, the sale of a child to an elite household meant the slave had a protected status, which could be extended to the offspring of the child in later years, as a confirmed member of a household. For some household slaves, this affiliation granted access to judges and the entitlement to plead cases in court. Occasionally this participation resulted in the release of the slave and/or his or her children, a break from the normal inheritance pattern. On the other hand, the sale of a child entailed the potentially traumatic separation of a child from their blood family, accompanied by a loss of inheritance rights, and, one must assume, possible exposure to abuse or violence. Given that many children were sold in the context of financial crisis, their fate may have been improved through their sale. The idea of an unsettling double potential of child domestic slavery—that is, possible abuse, degradation, and social death, versus possible mobility, safety, or opportunity—can be detected in ancient and modern records about domestic slavery in a variety of contexts, from debates about whether the mui tsai (female domestic slaves) custom in Hong Kong represents the enslavement of girls or a philanthropic tradition that saves them from poverty and social oblivion,23 to debates about brutal conditions and social possibilities of household slavery in ancient Rome.24 In these cases, we find nothing short of a variety of experiences and perspectives about household slavery. Abbreviations bm faos Iraq 41 mvn 6 ng uet 3

Signature of the British Museum, London 17 Steinkeller 1989 George 1979 Pettinato 1977 Falkenstein 1956–57 Legrain 1937

23  Pedersen 2001; Pomfret 2008; Poon 2009. While some British imperial officials and interest groups viewed the selling of young girls into domestic service as slavery, various local officials defended the custom, arguing that it was a charitable, philanthropic tradition that offered poor girls a better life and protected status as legal members of a household. The alternative—no home and no legal status—was far worse, they argued. A web of competing interests of course informed all these perspectives, and the voices of the mui tsai are less represented. It is known that not all mui tsai were given education and a comfortable life, many of them facing abuses, violence, and restricted freedoms. 24  Laes 2008, especially pp. 257 and 262f.

332

Culbertson

Bibliography Adams, Robert McC. (2010) “Slavery and Freedom in the Third Dynasty of Ur: Implications of the Garshana Archives.” Cuneiform Digital Library Journal 2010/2. Campbell, Gwyn; Suzanne Miers; and Joseph C. Miller, editors (2009) Children in Slavery through the Ages. Athens: Ohio University Press. Culbertson, Laura (2009) Dispute Resolution in the Courts of the Third Dynasty of Ur. Ph. D. dissertation, University of Michigan. Diakonoff, Igor M. (1987) “Slave-Labour vs. Non-Slave Labour: The Problem of Definition.” In Labor in the Ancient Near East, edited by M. A. Powell, pp. 1–4. American Oriental Series 68. New Haven: American Oriental Society. Falkenstein, Adam (1956–57) Die neusumerischen Gerichtsurkunden. 3 volumes. Bayerische Akademie der Wissenschaften, Philosophisch-Historische Klasse. Abhandlungen, n.F., 39, 40, and 44. Munich: Verlag der Bayerischen Akademie der Wissenschaften. Gelb, Ignace J. (1979) “Definition and Discussion of Slavery and Serfdom.” UgaritForschungen 11: 283–97. George, A. R. (1979) “Cuneiform Texts in the Birmingham City Museum.” Iraq 41: 121–40. Heimpel, Wolfgang (2009) “The Location of Madga,” Journal of Cuneiform Studies 61: 25–62. Laes, Christian (2008) “Child Slaves at Work in Roman Antiquity,” Ancient Society 38: 235–83. Lafont, Bertrand, and Raymond Westbrook (2003) “Neo-Sumerian Period (Ur III).” In A History of Ancient Near Eastern Law, Volume 1, edited by R. Westbrook, pp. 183– 226. Handbook of Oriental Studies, Section One, The Near and Middle East 72. Leiden: Brill. Legrain, Leon (1937) Business Documents of the Third Dynasty of Ur. Ur Excavations, Text 3. London: Harrison & Sons. Manning, Patrick (1990) Slavery and African Life: Occidental, Oriental, and African Slaves Trades. Cambridge: Cambridge University Press. Molina, Manuel (2000) La ley más antigua: Textos legales sumerios. Pliegos de oriente, Textos 5. Madrid: Editorial Trotta. Molina, Manuel (2008) “New Ur III Court Records Concerning Slavery.” In On the Third Dynasty of Ur: Studies in Honor of Marcel Sigrist, edited by P. Michalowski, pp. 125– 43. Journal of Cuneiform Studies Supplement Series 1. Boston: American Schools of Oriental Research. Patterson, Orlando (1982) Slavery and Social Death: A Comparative Study. Cambridge: Harvard University Press. Pedersen, Susan (2001) “The Maternalist Moment in British Colonial Policy: The Controversy Over ‘Child Slavery’ in Hong Kong 1917–1941.” Past and Present 171: 161–202.

A Life-Course Approach to Household Slaves

333

Pettinato, Giovanni (1977) Testi economici di Lagas del Museo di Istanbul, Part 1: La. 7001–7600. Materiali per il vocabolario neosumerico 6. Rome: Multigrafica Editrice. Pomfret, David M. (2008) “ ‘Child Slavery’ in British and French Far-Eastern Colonies 1880–1945.” Past and Present 201: 175–213. Poon, Pauline Pui-ting (2009) “The Weil-Being of Purchased Female Domestic Servants (mui tsai) in Hong Kong in the Early Twentieth Century.” In Children in Slavery through the Ages, edited by G. Campbell, S. Miers, and J. C. Miller. Athens: Ohio University Press. Siegel, Bernard J. (1947) Slavery during the Third Dynasty of Ur. Memoir Series of the American Anthropological Association 66. Menasha: American Anthropological Association. Sigrist, Marcel (1995) “Some di-til-la Tablets in the British Museum.” In Solving Riddles and Untying Knots: Biblical, Epigraphic, and Semitic Studies in Honor of Jonas C. Greenfield, edited by Z. Zevit, S. Gitin, and M. Sokoloff, pp. 609–18. Winona Lake: Eisenbrauns. Steinkeller, Piotr (1989) Sale Documents of the Ur-III Period. Freiburger Altorientalische Studien 17. Stuttgart: Franz Steiner. Studevent-Hickman, Benjamin (2006) The Organization of Manual Labor in Ur III Babylonia. Ph. D. dissertation, Harvard University. VerSteeg, Russ (2000) Early Mesopotamian Law. Durham: Carolina Academic Press. Waetzoldt, Hartmut (1972) Untersuchungen zur neusumerischen Textilindustrie. Studi economici e tecnologici 1. Rome: Istituto per l’Oriente. Westbrook, Raymond (1995) “Slave and Master in Ancient Near Eastern Law.” ChicagoKent Law Review 70: 1631. Westbrook, Raymond (1998) “The Female Slave.” In Gender and Law in the Hebrew Bible and the Ancient Near East, edited by V. H. Matthews, B. M Levinson, and T. S. FrymerKensky, pp. 214–38. Journal for the Study of the Old Testament, Supplement Series 262. Sheffield: Sheffield Academic Press. Wilcke, Claus (2007) Early Ancient Near Eastern Law: A History of Its Beginnings; The Early Dynastic and Sargonic Periods. Winona Lake: Eisenbrauns.

Slavery, Freedom and Citizenship in Classical Athens: Beyond a Legalistic Approach Kostas Vlassopoulos* In a seminal article written over four decades ago M. I. Finley made the memorable statement that ‘one aspect of Greek history is the advance, hand in hand, of freedom and slavery’.1 Finley explained that Greek communities, and in particular places like Athens, passed during the archaic period through a process whose endpoint was a social division into two exclusive categories, the slave and the free. It was not that the distinction between slave and free did not exist in earlier periods or in other, earlier or contemporary, societies. But earlier Greek communities, and the older civilisations of the Near East, usually incorporated this distinction within a wider spectrum of statuses, with varying degrees of freedom and dependence from the monarch at the top to the lowest slave at the bottom. It was only in certain Greek communities that the intermediate statuses were abolished and the social body divided into two polar opposites. Finley offered his own interpretation for this unique development. It was the result of a struggle between the aristocracy and the peasantry during the archaic period, in which the lower classes were victorious and destroyed the bonds of dependency on the aristocracy by political means: the lower classes were incorporated into the body politic as citizens with equal rights. Athens was, for Finley, the case par excellence in this development. The legislation of Solon in the early sixth century bc cancelled previous debts and prohibited the enslavement of Athenian citizens for debts incurred.2 This created a significant manpower problem for the elite, which could no longer count on exploiting the peasants to produce the necessary surplus. The solution was quickly discovered. Chattel slavery existed in archaic societies, but it played a rather marginal role, mainly that of personal attendance and service. But in the aftermath of these political struggles there emerged a massive importation of chattel slaves, who now took the role of providing the elite with labour Source: Vassopoulos, Kostas, “Slavery, Freedom and Citizenship in Classical Athens: Beyond a Legalistic Approach,” European Review of History/Revue européenne d’histoire, 16 (2009): 347–363. © Taylor & Francis Ltd, reprinted by permission of Taylor & Francis Ltd, www.tandfonline.com. * Email: [email protected]. 1  Finley, Economy and Society, 115. 2  But not debt bondage: see Harris, “Did Solon abolish debt bondage?” © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_014

Slavery, Freedom and Citizenship in Classical Athens

335

and surplus. Thus, the political gains of the lower classes and the creation of an egalitarian political field could be safeguarded through the exploitation of thousands of chattel slaves. In truth freedom and slavery had progressed hand to hand through the expansion of citizenship.3 Finley’s view that the growing importance of freedom and citizenship had as its corollary the growth of chattel slavery has been widely accepted by ancient historians in the last few decades.4 There is no denying that there is a significant element of truth in it. But there are many elements of the social and political history of classical Greece that this analysis leaves without account. I shall confine myself in the present context to a re-examination of the relationship between slavery and citizenship in classical Athens. My argument, briefly, can be summarised as follows. While there was a categorical and simple division between slave and free in Athenian law, in social practice the situation was very complicated. In many cases it is simply impossible to tell whether the individual involved was free or slave; even more, there are cases in which citizens are accused of being slaves or the children of slaves. This blurring of identities was recognised by contemporaries themselves as a significant phenomenon, as we shall see. The explanation of this bifurcation between law and reality should be sought in the function of the Athenian economy and the social content of Athenian citizenship. In Athenian democracy the privileges of citizenship extended to the lower classes of peasants, artisans, shopkeepers and wage labourers. Slave and free exercised the same professions and this meant that it was often necessary to accord slaves an extensive freedom of movement and initiative. At the same time, this overlap made it impossible to differentiate status solely on the basis of profession or living conditions. Thus, a number of slaves were in a position to take advantage of this blurring of identities to escape detection, avoid maltreatment and create better conditions for themselves. From a legal perspective the distinction between slave and free was manifested in a number of ways.5 To start with, there were activities which were permitted only to freemen, while strictly prohibited to slaves. For example, only freemen had the right to own property, while the law categorically denied this right to slaves.6 Activities that were essential to the Greek sense of identity, like participation in gymnastic activities, were strictly forbidden to slaves:

3  Finley, Ancient Slavery, 67–92. 4  But see the modifications suggested by Rihll, “Origins and establishment.” 5  Hunter, “Introduction,” 5–15; Todd, Shape of Athenian Law, 167–200. 6  Harrison, Law of Athens, 236.

336

Vlassopoulos

Your fathers, when they were laying down laws to regulate the habits of men and those acts that inevitably flow from human nature, forbade slaves to do those things which they thought ought to be done by free men. ‘A slave’, says the law, ‘shall not take exercise or anoint himself in the wrestling-schools’.7 Moreover, penalties for law-breaking were clearly differentiated according to the status of the individual: Indeed, if you wanted to contrast the slave and the freeman, you would find the most important distinction in the fact that slaves pay for all offences with their body, while freemen, even in the most unfortunate circumstances, can protect their persons. For it is in the shape of money that in the majority of cases the law must obtain satisfaction from them.8 Finally, in terms of legal procedure, a slave’s testimony was admissible in court only if taken after torture; on the contrary, judicial torture was impermissible for freemen, except under very special circumstances which involved religious sacrilege or treason.9 Thus, the establishment of an individual’s status was considered necessary for the proper function of the legal procedure. The legal picture seems clear enough. Many of the legal procedures were based on the assumption that all people belonged to either the category of the free or the category of the slave; establishing the identity of a person was a necessary precondition of the function of the legal system. It goes without saying that the distinction was even more pronounced when the distinction was not simply between free and slave, but between citizen and slave. The citizen of the Athenian democracy was particularly privileged: he could vote on all essential issues of his community, he could not be maltreated without vengeance, and he had many means of protecting himself against abuse of power by the magistrates. The list can go on; against this, the impotence of the slave as regards legal protection could not be drawn more starkly. So far, Finley’s picture seems to hold good. Actual social practices, to judge from the Athenian law court speeches, may have been somewhat different from the legal theory. There is little doubt that 7  Aeschines, Against Timarchus, 138. All translations are adapted from the Loeb Classical Library, unless otherwise stated. 8  Demosthenes, Against Androtion, 55. 9  Andocides, On the Mysteries, 43; Thür, Beweisführung; Gagarin, “Torture of slaves”; Mirhady, “Torture and rhetoric”; Hunter, Policing Athens, 154–84.

Slavery, Freedom and Citizenship in Classical Athens

337

any attempt to read ‘reality’ as a simple reflection of statements found in these speeches is mistaken. These speeches narrate a number of facts which are always presented from the perspective of a conflict between the two plaintiffs.10 When we attempt to use these speeches as sources of social history and turn the various statements found in them into historical facts, we have always to take seriously into account the rhetorical aims of the speakers and the ways in which they deliberately omit or expand certain facts in order to suit their case.11 Let us start with the more elementary cases. These are cases in which it is disputed whether a certain individual is slave or free. Usually, the argument arises out of the fact that slaves can be tortured to give evidence, while freemen and freedmen cannot. It is thus in the interest of the speaker to argue in favour or against a person’s free status, depending on whether he claims to be in favour of settling an issue by exacting evidence from a slave through torture. Thus, in Demosthenes’ speech 29, the issue turns on the status of Milyas, and whether he can be legally tortured to exact his testimony.12 Demosthenes argues that Milyas was manumitted by his father on his deathbed and accordingly Milyas cannot be tortured in order to provide evidence.13 In fact, Demosthenes reads a deposition in which he claims that his opponent admitted that Milyas is free (§32). Precisely the opposite argument occurs in the dispute between Apollodorus and Timotheus in a speech of the Demosthenic corpus: Now, as to the bowls and the mina of silver, which he borrowed from my father when he sent his servant Aeschrion to my father in the night, I asked him before the arbitrator if Aeschrion was still a slave, and demanded that he be put to the test ‘in his hide’. He answered that Aeschrion was free, so I desisted from my demand; but I required him to put in a deposition made by Aeschrion as being a free man. He, however, neither provided a deposition from Aeschrion, as being free, nor would he deliver him up as a slave that proof might be had from his body; for he was afraid that, if he produced a deposition from him as being free, I should bring suit for false testimony … and if, again, he should deliver him up for the torture, he was afraid that Aeschrion would state the truth against him.14

10  See Johnstone, Disputes and Democracy, 46–69. 11  See e.g. Fisher, “Independent slaves.” 12  Thür, “Der Streit über den Status.” 13  Demosthenes, Against Aphobus, 25–6. 14  Demosthenes, Against Timotheus, 55–6.

338

Vlassopoulos

A speech by Isocrates covers a dispute between the speaker, a son of a potentate in the Kingdom of Bosporus in the Black Sea, and the banker Pasion, a former slave himself.15 According to the speaker, Pasion had defrauded him of his money which was deposited in Pasion’s bank and spirited away Cittus, the slave who was in knowledge of the transaction. But Menexenus found the slave here in the city, and having seized him demanded that he give testimony under torture about both the deposit and the charge brought by his master. Pasion, however, reached such a pitch of audacity that he secured the release of the slave on the ground that he was a freeman and, utterly devoid of shame and of fear, he claimed as a freeman and prevented the torture of a person who, as he alleged, had been stolen from him by us and had given us all that money.16 According to the speaker, when Pasion subsequently accepted that the man was a slave and surrendered him for torture, he did not allow an actual torture, but only a verbal interrogation (§17). Later on, the speaker claims that Pasion refused to sail to the Black Sea and settle the issue, as he had promised, but sent instead Cittus, the slave, who claimed to be a freeman and a Milesian by birth (§51).17 Another speech by Isaeus concerns the property of Nicostratus, an Athenian who died abroad. There were many claimants to the inheritance, but one pair made a particularly interesting claim: Ctesias of Besa and Cranaus at first asserted that Nicostratus had been condemned to pay them a talent; when they could not prove this, they pretended that he was their freedman; they were no better able to prove their statement.18 Apparently, it could be imagined that a person like Nicostratus, who had lived abroad for 11 years and had amassed a fortune of two talents, could have been a freedman. A dispute in a speech by Lysias revolves around a slave woman who, according to the speaker, was bought in common by the plaintiff and the defendant. Both of them were injured in a brawl over the woman, and the defendant claims that evidence should be extracted through torture from the woman who was a 15  See Trevett, Apollodorus, 1–9. 16  Isocrates, Trapeziticus, 13–14. 17  Thür, “Komplexe Prozessführung.” 18  Isaeus, On the Estate of Nicostratus, 9.

Slavery, Freedom and Citizenship in Classical Athens

339

witness. But the plaintiff refused to allow it, claiming that the woman was actually free, i.e. she had been manumitted.19 Another speech of Lysias also concerns a brawl, but this time over a young man (meirakion) called Theodotus.20 Theodotus is described as a young man from Plataea (§5), a Boiotian city close to Athens and with strong links to it. At the time that this speech took place Plataea had been destroyed by Thebes because of her alliance with Athens and all Plataeans had been granted Athenian citizenship as a reward for their unwavering loyalty.21 The plaintiff argued that he had a contract with Theodotus to pay him 300 drachmas for his sexual favours and accused the defendant of seducing the boy away from him (§22). The defendant denied the existence of such a contract and explained that the boy was living with him (§32). The actual status of the young man is impossible to tell. At a certain point the defendant alludes to the theoretical possibility that the young man might testify against him through torture (§33). This would imply that the young man was a slave, since no free person could be tortured for evidence.22 And yet, if Theodotus was indeed a slave, there is no reference to his master, who could presumably settle the issue by determining whether Theodotus had a contract with either of the opponents.23 It would be possible to interpret these passages by arguing that they simply reflect the cynical manipulation of slaves by their masters in their feuds and disputes; we could also claim that they are evidence of the precarious status of freedmen in classical Athens, where the absence of an obligatory registration of the manumission meant that it could often be challenged by the former owner or by other persons.24 But there is a great deal of remaining evidence which shows that the uncertainty of status which we have encountered in these cases is only an aspect of a wider phenomenon. In other words, claims that a certain slave was actually a freedman, or that a certain free person was actually a slave, were believable because of a wider background of contested claims of status. Slave owners were certainly able to manipulate the status of slaves and freedmen because there was inequality of power. But the process may have had a number of different sides, one of which was the exact opposite: 19  Lysias, On a Wound by Premeditation, 12. 20  Lysias, Against Simon. 21  Osborne, Naturalization II, 11–16; Kapparis, “Athenian decree.” 22  But see Cairns, “Civic status”; Todd, Commentary, 333–4. 23  Todd, Commentary, 279–81. 24  For the precariousness of the freedom of other free foreigners in certain circumstances, see the case in Dinarchus, Against Demosthenes, 23, where a miller is executed for keeping a free boy from a foreign city as a slave.

340

Vlassopoulos

slaves and freedmen could manipulate this state of uncertainty for their own benefit. Let us start with cases in which this lack of distinction of status is just taken for granted and is part of the background and not the main contested issue of a speech. In two passages from speeches of Demosthenes the speaker casually comments on how easily a slave boy could be taken for a citizen boy: More than this, as they were neighbours and my farm adjoined theirs, they sent into it in the daytime a young boy who was an Athenian, and put him up to plucking off the flowers from my rose-bed, in order that, if I caught him and in a fit of anger put him in bonds or struck him, assuming him to be a slave, they might bring against me an indictment for assault.25 And not only did they go off with my furniture, men of the jury, but they were even on the point of taking away my son, as though he were a slave, until Hermogenes, one of my neighbours, met them and told them that he was my son.26 But the lack of distinction did not simply concern little boys. In another speech included in the Demosthenic corpus, the speaker accuses Stephanus of certain malicious activities: For not only did Stephanus seek in this way to bring us to ruin, but he even wished to drive Apollodorus from his country. He brought a false charge against him that, having once gone to Aphidna in search of a runaway slave of his, he had there struck a woman, and that she had died of the blow; and he suborned some slaves and got them to give out that they were men of Cyrene, and by public proclamation cited Apollodorus before the court of the Palladium on a charge of murder.27 Apparently it was believable that a couple of slaves could be presented as citizens from another Greek city and have them testify in court. They must have spoken very good Greek, at the very least.28 Another interesting example is presented in a speech by Isaeus. According to the speaker, Euctemon was a rich man who owned a tenement house 25  Demosthenes, Against Nicostratus, 16. 26  Demosthenes, Against Evergus and Mnesibulus, 61. 27  Ps-Demosthenes, Against Neaira, 9. 28   For the problems in interpreting this passage, see Worthington, “The mysterious Cyreneans.”

Slavery, Freedom and Citizenship in Classical Athens

341

managed by a freedwoman who was formerly a slave of his. This freedwoman had bought a slave prostitute by the name of Alce, who exercised her profession there for many years. When she grew old, she was given charge by Euctemon of another tenement house of his. Alce had two children by a freedman called Dion. She then developed a relationship with Euctemon, who was very old by then, and convinced him to introduce her eldest son to his phratry, thus accepting him as his legitimate child and ultimately making him eligible for citizenship. Despite a quarrel with his relatives, who objected to the introduction, Euctemon carried the day and managed to have Alce’s son recognised, though, according to the speaker, under the restriction that he should only inherit a single farm.29 It is of course impossible to verify the speaker’s claims that the two children were indeed the offspring of Alce and not of another wife of Euctemon, as they claimed, or that they were the children of the freedman Dion and not actually of Euctemon. What this case shows is the possibility that interaction between citizens, freedmen and slaves could create relationships that freedmen and slaves could use for their own benefit and to enhance their position. Another example can be found in the case of Pittalacus, an individual mentioned in Aeschines’ speech against Timarchus.30 Aeschines presents Pittalacus as a public slave at the time of the events described, and he does not even allude that he might have been manumitted at that point in time; nevertheless, it is revealed that Pittalacus was affluent and had his own house, where he organised cock fights, from which presumably he reaped a considerable profit. Aeschines argues that Pittalacus had a sexual affair with Timarchus, the defendant and a scion of a wealthy Athenian family. The situation allegedly went sour when Timarchus left Pittalacus to go and live with Hegesandrus, a minor Athenian politician at the time. Pittalacus was enraged, and as a punishment for his complaining he was brutally beaten by Hegesandrus and Timarchus, who also ravaged his house and killed his beloved cocks. Aeschines claims that Pittalacus brought a legal action against Timarchus and Hegesandrus. Given that he is presented as a public slave, he had no right to bring an action; therefore, either he was already a freedman at the time of the events described, or he had to find some citizen to introduce the action on his behalf.31 In order to bar Pittalacus from bringing the action, Hegesandrus claimed that Pittalacus was his own slave. Nick Fisher has proposed an ingenious explanation of how this claim might have arisen. He has suggested that Pittalacus 29  Isaeus, On the Estate of Philoctemon, 19–24. 30  Aeschines, Against Timarchus, 54–64; Fisher, Aeschines, 189–204. 31  Zelnick-Abramovitz, Not Wholly Eree, 298; Hunter, “Pittalacus and Eucles.”

342

Vlassopoulos

might have been a slave of the genos of Salaminioi, a body with public functions to which Hegesandrus belonged. The genos was in charge of the temple of Athena Skiras, which was possibly a centre of gambling, and this could explain Pittalacus’ association with gambling. This hypothesis could thus explain both why Pittalacus is described as a public slave and how Hegesandrus could plausibly make a claim that Pittalacus belonged to him.32 Whatever the truth, Pittalacus was able to find a citizen to defend him by bringing a process known as aphairesis eis eleutherian, which aimed to ensure that a freeman was not falsely captured as slave, but was ‘taken’ by a third party into freedom. From these details it seems probable, though by no means certain, that Pittalacus had been manumitted by the time these events were taking place. Thus, we would have to assume that Aeschines presents him as still being a public slave for rhetorical reasons, in order to further emphasise the gravity of Timarchus’ having a passive sexual affair with a public slave.33 Even if we grant this, the fact that Hegesandrus tried to claim that Pittalacus was his own private slave shows the uncertainty of establishing personal status in Athens.34 But the fact that Pittalacus was able to find a citizen to defend him through a judicial process testifies again to the importance of links between slaves, freedmen and citizens. A similar case appears in a fragmentary speech of Isaeus.35 The speaker, who must be a wealthy Athenian since he had served as a trierarch, has initiated a similar procedure, asserting the liberty of the freedman Eumathes. The speaker claims that Eumathes was manumitted by Epigenes through a judicial procedure (en dikastêriô), but the claim was disputed by the heir to Epigenes’ property.36 The speaker states that he had rendered services to Eumathes on a former occasion as well, and explains his willingness to support a freedman as a result of Eumathes’ honest behaviour as a banker: As a result of this conduct, when I returned safe home, I became still more intimate with him, and, when he established his bank, I provided him with capital, and afterwards, when Dionysius tried to enslave him,

32  Fisher, Aeschines, 357–62. 33  Fisher “Independent slaves,” 12. 34  See also the comments of Todd, Shape of Athenian Law, 192–194. 35  Zelnick-Abramovitz, Not Wholly Free, 293–94. 36  In another fragmentary speech of Isaeus it is stated that an unnamed person cast Hermocrates into prison, alleging he was a freedman, and did not release him until he extracted 30 drachmas: Isaeus, fragment 13.

Slavery, Freedom and Citizenship in Classical Athens

343

I asserted his liberty, being well aware that he had been liberated by Epigenes in court.37 But asserting the freedom of a slave could have serious consequences, as another speech included in the Demosthenic corpus shows: Theocrines here owes five hundred drachmas, as his father did not pay the supplementary fine he owed for claiming that Cephisodorus’ slave was a free woman.… Read, please, the law which declares that anyone who is adjudged to have wrongfully asserted the freedom of a slave shall pay half the sum assessed into the public treasury.38 This example shows conclusively that it was not merely the owner of a slave who could argue that his slave was actually free when it suited him; in fact, anybody could present such a claim for the slave of another person. Unfortunately, we are not given a motive for Theocrines’ father making such a claim, but this case is important in establishing that a slave could count in certain circumstances on a citizen’s help for disputing his/her status as slave. It is possible for example that there was a dispute between Theocrines’ father and Cephisodotus, and that he tried to harm Cephisodotus by claiming that the slave was actually a free woman; or we can assume that there was a dispute between the woman and Cephisodotus, and the woman took advantage of an enmity between Theocrines’ father and Cephisodotus to ask for his help in claiming her freedom; or we might even imagine that there was a relationship between the woman and Theocrines’ father which induced him to claim that she was free. Whatever was the reason, it must have been strong enough, for Theocrines’ father initiated a risky procedure that in the end cost him dear.39 But let us now move in the opposite direction and look at the issue from a different perspective: that of slaves and freedmen as active agents following strategies to enhance their position. In his masterful collection of characters, Theophrastus presents the portrait of the slanderer: The slanderer is one who, when asked who so-and-so is, will reply, in the style of genealogists, ‘I will begin with his parentage. This person’s father was originally called Sosias; in the army he came to rank as Sosistratus; and, when he was enrolled in his deme’ as Sosidemus. His mother, I may 37  Isaeus, fragment 16. 38  Demosthenes, Against Theocrines, 19–21. 39  See also Lysias, fragment 437 (Carey).

344

Vlassopoulos

add, is a Thracian of good family. At all events, she is called < >, and they say that such ladies are esteemed noble in their own country. Our friend himself, as might be expected from his parentage, is naturally a criminal with a tattoo.’40 According to the slanderer it was possible for a former slave to assume the role of the citizen; and he presents the former slave as following a conscious strategy to obliterate the stigma of slavery and enhance his position in society. But for the time being let us explore whether the calumniations of the slanderer were a figment of Theophrastus’ imagination, or whether they illustrate a widespread phenomenon. In fact, ancient sources offer a number of real cases of such challenges and pronouncements. That these challenges were taken seriously and were followed closely can be seen by a speech which concerns an inheritance dispute. At the centre of the dispute is the question of whether a certain unnamed woman had a legitimate marriage with Pyrrhus, the deceased, and whether Phile, the offspring of this relationship, was a legitimate child that could properly inherit his property. The speaker, Pyrrhus’ nephew, denies the claim arguing that on the basis of probability it would be impossible that his uncle would have married such a woman. One of his main arguments is that the woman came from a family of disputable origin: Further, would our uncle have thought of marrying the sister of a man, who, when he was accused of usurping the rights of citizenship by a member of the deme to which he claimed to belong, obtained those rights by a majority of only four votes?41 We should start by explaining the procedure that Isaeus refers to. Every Athenian was a member of a deme, one of the approximately 140 local wards into which Attica was subdivided. Membership in a deme was hereditary, and depended not on the actual place where a citizen resided, but on where one’s ancestors were resident when the initial registration took place with the reforms of Cleisthenes in 508 bc. On reaching adulthood, every Athenian had to be accepted by the assembly of the deme to which his father belonged in order to become a citizen.42 The deme assembly had thus the right to reject a candidate, though the rejected candidate could then appeal to a popular court. In addition, the Athenians periodically held diapsêphiseis, a procedure in which the 40   Characters, 28.2–3; translation from Diggle, Characters. 41  Isaeus, On the Estate of Pyrrhus, 37. 42  Whitehead, Demes, 97–109.

Slavery, Freedom and Citizenship in Classical Athens

345

deme assembly voted whether to accept or reject the citizenship claim of every single deme member. The individual in the speech above had his citizenship confirmed, even if just barely, but others were not so lucky. Aeschines narrates one more shameless act of Timarchus, whom we have already encountered: In consequence of this experience so great became his contempt for you that immediately, on the occasion of the revision of the citizen lists, he gathered in two thousand drachmas. For he asserted that Philotades of Cydathenaeon, a citizen, was a former slave of his own, and he persuaded the members of the deme to disfranchise him. He took charge of the prosecution in court … and though he swore by the usual gods of oaths and called down destruction on his own head, yet it has been proved that he received twenty minas from Leuconides, the brother-in-law of Philotades, through Philemon the actor…. And so he broke his oath and abandoned the case.43 Thus, Philotades’ claim to citizenship was rejected in one such diapsêphisis, after the proposal of Timarchus.44 But how exactly could that happen? This story could be interpreted in a cynical manner as one more example of the incessant feuding among Athenian citizens. If we believe Aeschines, Philotades was indeed a citizen, and Timarchus denounced him as his own freedman either because of enmity and/or because he wanted to profit financially from the act. But how could he persuade the members of the deme that one of their own deme members was actually a former slave? For certainly, given the dominant assumptions about the division between the free and the slave in modern scholarship, this attitude is difficult to grasp. The best way to account for it is to accept that there were indeed freedmen or slaves who tried to or succeeded in passing off as citizens. There are indeed many instances of citizens accused of being slaves or freedmen. Lysias 30 is a speech against a certain Nicomachus. Nicomachus became a commissioner for the transcription of the Athenian laws in the last years of the Peloponnesian War and in the immediate aftermath of Athens’ defeat. At that point he must have been a citizen. The speaker argues that Nicomachus’ father was a public slave and accuses him of becoming ‘citizen instead of a slave, rich instead of beggar, lawgiver instead of under-clerk’.45 A similar accusation was levelled against Agasicles in a fragmentary speech by Dinarchus. Agasicles was 43  Aeschines, Against Timarchus, 114–15. 44  For problems with this story, see Fisher, Aeschines, 253–55. 45  Lysias, Against Nicomachus, 2; Jacob, Esclaves publics, 180–3.

346

Vlassopoulos

charged with bribing the citizens of the deme of Halimous to enrol him in the citizen register. According to the speaker, ‘he was the son of a Scythian measurer, born among public slaves and was himself a measurer up till now in the market, and you regularly receive the corn from him’.46 Another speech by Lysias concerns a denunciation (endeixis) against Agoratus for the murder of Dionysodorus.47 The accusation was that Agoratus had testified that Dionysodorus and other important citizens, who were opposed to the Athenian surrender to Sparta, were conspiring against Athenian interests; their conviction and execution led the way to the assumption of full power by the Thirty Tyrants and the subversion of democracy in 404 bc. According to the speaker, Agoratus was ‘a slave and a son of slaves’ (§18), the son of Eumares, who was the property of Nicocles and Anticles (§64). Consequently, he was liable to be tortured to exact his testimony (§27). On the other hand, the speaker admitted that Agoratus had exercised citizen rights by bringing public prosecutions and speaking in the assembly (§§65–6). How did he become a citizen? According to the speaker, Agoratus claimed that he had killed Phrynichus, one of the leaders of the oligarchic revolution of 411, and was subsequently rewarded with citizenship by the Athenian people (§70).48 The speaker denied strongly Agoratus’s involvement in the murder of Phrynichus, claiming that it was only Thrasybulus of Calydon and Apollodorus of Megara who participated in the act and got Athenian citizenship as a reward; consequently, he accused Agoratus of falsely assuming Athenian citizenship (§§71–3). By a rare piece of luck the decree for the killers of Phrynichus that Lysias cited is still extant.49 It appears that Apollodorus and Thrasybulus were considered responsible for the act and were granted citizenship and other privileges. Agoratus is indeed mentioned among other benefactors (euergetai) in the same inscription, which seems to imply that he was connected in some way to the act, but not closely enough to merit the same honours. Agoratus and the other benefactors were granted some fiscal privileges, but certainly not citizenship. Thus, the speaker seems justified in accusing Agoratus of falsely assuming Athenian citizenship. It is of course possible that Agoratus gained citizenship after the extant decree was passed, either because the Athenians later awarded citizenship to the other benefactors as well, or 46  Dinarchus, fragment 16.4. See Jacob, Esclaves publics, 119–21. 47  Lysias, Against Agoratus. 48  See the article by Marc Kleijwegt in this issue for the Roman connection between political services, manumission and citizenship. 49  For the text see Osborne, Naturalization I, 28–30.

Slavery, Freedom and Citizenship in Classical Athens

347

because he gained it in a subsequent circumstance, for example through his participation in the democratic forces who overthrew the Thirty Tyrants.50 We thus reach one more case where it is impossible to decide what the precise status of the individual involved was. In the same speech the orator also mentions Aristophanes of the deme of Cholleidai, who was arrested in connection with the same events and whose fate was debated in the assembly: Some, indeed, desired that Aristophanes should be put to the torture, as one who was not of pure Athenian stock, and they prevailed on the people to pass the following decree…. Well, after that the persons who then had control of affairs came to Aristophanes and appealed to him to save himself by a denunciation, and not to run the risk of the extreme penalty by standing his trial on the count of alien birth (xenias). But he said ‘Never!’51 So Aristophanes was one more citizen who was accused of falsely assuming citizenship. In another Demosthenic speech52 the orator accuses Aristogeiton, a citizen with active involvement in public affairs, of not being a freeman (§79) and relates how his mother was sold when she was convicted of defrauding her emancipator (§65). Demosthenes also accuses the politician Androtion of denouncing citizens in such a manner: Then do you suppose that all these men are his inveterate enemies merely because he collected this money from them? Is it not rather because he said of one of them, in the hearing of all of you in the Assembly, that he was a slave and born of slaves and ought by rights to pay the contribution of one–sixth with the resident aliens; and of another that he had children by a harlot; of this man that his father had prostituted himself; of that man that his mother had been on the streets …—slandering them all in turn?53 But Androtion’s calumnies were not a rare occurrence. In fact, one of the most impressive facts about Athenian politics is that many Athenian politicians of the first rank were accused of falsely assuming citizenship and of being slaves, 50  Osborne, Naturalization II, 16–21. 51  Lysias, Against Agoratus, 59–60. 52  Demosthenes, Against Aristogeiton I. 53  Demosthenes, Against Andorion, 61.

348

Vlassopoulos

or children of slaves. In a fragmentary speech, the orator Andocides makes the following claim about Hyperbolus, one of the most important politicians in the final years of the fifth century bc: ‘Hyperbolus I blush to mention. His father, a branded slave, still works at the public mint; while he himself, being a foreigner and a barbarian makes lamps for a living’.54 Demosthenes made similar accusations about his chief political opponent, Aeschines: I am at no loss for information about you and your family; but I am at a loss where to begin. Shall I relate how your father Tromes was a slave in the house of Elpias, who kept an elementary school near the Temple of Theseus, and how he wore shackles on his legs and a timber collar round his neck?… Only recently—recently, do I say? Why it was only the day before yesterday when he became simultaneously an Athenian and an orator, and, by the addition of two syllables, transformed his father from Tromes to Atrometus.55 Aeschines replied with the same coin, calling Demosthenes ‘more slave than freeman, all but branded as a runaway’56 and claiming that Demosthenes’ mother was of Scythian origin57 and thus Demosthenes did not qualify to be a citizen, since both his parents needed to be of Athenian parentage. In the case of the fourth-century politician Pytheas we have evidence that all these accusations were not mere idle talk. Pytheas was prosecuted for illegally assuming citizenship: For who does not know that this man, when, under the obligation to serve you, he was entering upon public life, was being hounded as a slave and was under indictment as an alien usurping the rights of a citizen and came near being sold by these men, whose servant he now is?58 It is very unlikely that the accusations against Aeschines, Demosthenes or Hyperbolus had any factual basis. We know enough about the families of these politicians to be able to say that they were established for quite some time.59 Aeschines’ attack on Demosthenes is the best example of how such an 54  Andocides, fragment 5; See Brun, “Hyperbolos.” 55  Demosthenes, On the Crown, 129–30. 56  Aeschines, On the Embassy, 79. 57  Aeschines, Against Ctesiphon, 171. 58  Demosthenes, Letter III, 29–30. 59  See Davies, Athenian Propertied Families.

Slavery, Freedom and Citizenship in Classical Athens

349

accusation could be used for rhetorical ends. For Aeschines loosely calls Demosthenes ‘almost a slave’ and stresses the foreign origin of his mother. Since the issue at hand in the speech is neither the status of Demosthenes nor his right to exercise citizenship, the accusation must play a different role. As Josiah Ober has shown, the main purpose of such accusations was to dispute the intentions of the public speaker and the appropriateness of his giving advice to the community: ‘a slave was a natural enemy of the political order, and he could be presumed to hate the system that maintained his servile status’.60 Thus a slave could not be expected to have gratitude towards the Athenians and his advice could not be trusted. Accordingly, we can see how the accusation of falsely assuming citizenship could be used to attack a fellow politician, even if there was little factual basis in it. But the habit of attacking political leaders as slaves or of servile origin would have never emerged, if such accusations were not possible about the rest of the citizens. We have to interpret the accusations of politicians in the light of the wider phenomenon that we examined. All of these accusations assume a pattern in which a slave, a freedman or the child of former slaves tried to assume the identity of citizen or even actively exercise citizenship rights, often by changing his own name or the name of his parents and thus hiding his links to the past. Let me summarise the argument so far. The distinction between citizen and non-citizen, free and slave, was far from clear and indisputable; claims to individual status and identity in Athens could be and were challenged; even the most important politicians could be accused of being foreigners, slaves or children of slaves. I have presented both speeches which argue that an accusation is false and speeches which strongly support such accusations. We do not have to, and in most cases we cannot, determine whether any of these accusations and challenges was indeed true. It is obvious that accusations and challenges could be used in the service of private and public enmities and feuds with little factual support or even as outright calumnies. It is also clear that the precarious state of Athenian bureaucracy and the rudimentary nature of Athenian documentation of such important acts as marriage, adoption or manumission made it relatively easy to advance all sorts of spurious claims.61 But the phenomenon is widespread enough to require a more general explanation. Unless a significant number of freedmen, free foreigners and even slaves were able to hide their status, claim they were citizens, or even exercise citizenship rights

60  Ober, Mass and Elite, 270–1. 61  Scafuro, “Witnessing,” 173–4.

350

Vlassopoulos

actively, it is impossible to understand otherwise how such claims could be so easily and so often presented or challenged. Once we have reached this conclusion, a more difficult question emerges: how was this ever possible? How could freedmen, foreigners or slaves ever assume the role of citizen? There are in my view two answers to this question which are closely connected. The one is economic and social, the other is political. In contrast to many other societies based on slavery, Athenian society did not make distinctions based on status, when it came down to labour and professions. In other words, in almost every single profession one could find citizens, free foreigners, freedmen or slaves exercising it.62 It was not even the case that slaves tended to exercise the less skilled professions, while freemen or indeed citizens reserved for themselves the more lucrative ones. It is of course true that the professions with the worst working conditions were often exercised by slaves. Mining, probably the profession with the most appalling working conditions, was almost exclusively reserved for slaves.63 But slaves could be found in all lucrative professions, and some of them, for example banking, were almost exclusively exercised by slaves or freedmen.64 Judging from their names, many slaves or freedmen must have worked as painters of the exquisite Athenian vases; yet, despite the fact that these artefacts are evidence of a very high level of craftsmanship and expertise, it is impossible to differentiate between the work of slaves and the work of free painters. This fact had an important repercussion. Thousands of slaves were not serving their masters in their households, but were treated as a source of income.65 These slaves, usually skilled, were rented out or worked on their own, thus providing their masters with a share of their profits and with a standard income. Many slaves worked and lived on their own, and their only contact with their master was the weekly or monthly meeting in which they would pay a fixed amount as a return. Also the fact that the slaves exercised all sorts of professions meant that they needed a free hand to deal with the issues and problems arising from it.66 To give only a single example, the three slaves of Athenogenes presented in a speech of Hypereides ran a perfume workshop on such a scale that they had incurred debts of five talents.67 The requirements created by 62  Randall, “Erechtheum workmen”; Schumacher, Sklaverei, 91–238. 63  Lauffer, Bergwerkssklaven. 64  For the important role of slaves in banking, see Cohen, Athenian Economy and Society, 61–110. 65  Perotti, “Esclaves choris oikountes.” 66  Cohen, Athenian Nation, 130–54. 67  Hypereides, Against Athenogenes, 9–10.

Slavery, Freedom and Citizenship in Classical Athens

351

such social arrangements are nicely illustrated by a fifth-century author traditionally called the Old Oligarch: If anyone is also startled by the fact that they let the slaves live luxuriously there and some of them sumptuously, it would be clear that even this they do for a reason. For where there is a naval power, it is necessary from financial considerations to be slaves to the slaves in order to take a portion of their earnings, and it is then necessary to let them go free. And where there are rich slaves, it is no longer profitable in such a place for my slave to fear you. In Sparta my slave would fear you; but if your slave fears me, there will be the chance that he will give over his money so as not to have to worry anymore. For this reason we have set up equality of speech between slaves and free men, and between metics and citizens.68 Thus, the requirements of the Athenian economy created a substantial mass of slaves, freedmen and free foreigners (metics), who exercised the same professions with the lower classes of the citizen population, lived in the same areas, and in general could be distinguished with difficulty from them.69 The Old Oligarch provides an eloquent description of this phenomenon: Now the slaves and metics at Athens are most licentious; you can’t hit them there, and a slave will not stand aside for you. I shall point out why this is their native practice: if it were customary for a slave (or metic or freedman) to be struck by one who is free, you would often hit an Athenian citizen by mistake on the assumption that he was a slave. For the people there are no better dressed than the slaves and metics, nor are they any more handsome.70 The key to understanding the above passage is the position of the lower classes among the citizens. Athenian citizenship could be exercised without any exclusion based on wealth, profession or appearance. The Athenian citizen body comprised people who were rich, comfortably well-off, struggling to make ends meet, poor, even destitute; it comprised the aristocrat and the landowner together with the wage labourer, the artisan, the shopkeeper and the merchant. The upper-class reactionary could not maltreat lower-class people randomly, because those of the lower classes who were citizens had political power and 68   Constitution of the Athenians, 1.11–2; Klees, Sklavenleben, 355–78. 69  Klees, Sklavenleben, 64–74. 70   Constitution of the Athenians, 1.10. See Cataldi, “Akolasia.”

352

Vlassopoulos

would not accept maltreatment stoically. But since it was impossible to differentiate who exactly was a citizen and who was a foreigner, a freedman or a slave, it was prudent to abstain from aggressive behaviour in general. Thus, the unintentional effect of democratic citizenship was that slaves, freedmen and foreigners were often protected by association, simply because it was impossible to differentiate them from the citizens. On the other hand, this was precisely the condition that made it possible for certain foreigners, freedmen or even slaves to assume the identity of the citizen and even exercise it actively. That this was no idle rhetoric can be shown through two examples. The first shows in detail how the difficulty in distinguishing between slaves, freedmen and poorer citizens could lead to challenges to one’s claim to citizenship. The second shows how a slave or a freedman could take advantage of this lack of distinction in order to avoid mistreatment, enhance his position and claim a title to citizenship. Euxitheus, the speaker of Demosthenes 57, was disfranchised by his deme in a diapsêphisis and was trying to convince the jury to reinstate him as a citizen.71 Euxitheus was by then in his middle years and he had previously served as dêmarchos, chief magistrate of his deme (§63), and phratriacrhos, chief of his phratry (§23). He was thus no stranger to his fellow citizens. In fact, Euxitheus attributed his difficulties to a family enmity with Eubulides (§8), the citizen who proposed his disfranchisement, to the illegal manner and the irregular conditions under which the vote was taken (§9–14), and to enmities which he incurred during his service as dêmarchos for trying to collect debts owed to the deme (§63). Thus, if we are to believe Euxitheus’ viewpoint, we should adopt the cynical model and view the accusations of falsely assuming citizenship as the result of feuds and enmities among the citizens. But the situation is much more complicated, as the evidence presented by Euxitheus himself shows. Although never explicitly stated, it seems that the accusation against him was that he was either a child of slaves and thus not eligible for citizenship, or even the illegitimate child of parents with doubtful credentials.72 The accusations concerned both his father and his mother. As regards the father: They have maliciously asserted that my father spoke with a foreign accent. But that he was taken prisoner by the enemy in the course of the Decelean war and was sold into slavery and taken to Leucas, and that he there fell in with Cleandrus, the actor, and was brought back here to his 71  Demosthenes, Against Eubulides; Lacey, “The family of Euxitheus.” 72  Humphreys, “Kinship patterns,” 60–2.

Slavery, Freedom and Citizenship in Classical Athens

353

kinsfolk after a long lapse of time—all this they have omitted to state … they have made his foreign accent the basis of a charge against him.73 But the accusations against his mother were even more serious.74 Not only was she accused of selling ribbons in the marketplace (§31), which was not considered a proper employment for an Athenian woman, but even more she had worked as a wet nurse, the profession par excellence of a slave woman:75 Some time after this, when by now two children had been born to her, she was compelled … to take Cleinias, the son of Cleidicus, to nurse. This act of hers was, heaven knows, none too fortunate with reference to the danger which has now come upon me (for it was from this nursing that all the slander about us has arisen).76 Thus, the speaker concludes by stressing the difficulty of distinguishing between citizens and slaves based on profession or labour: Many are the servile acts which free men are compelled by poverty to perform, and for these they should be pitied, men of Athens, rather than be brought also to utter ruin. For, as I am informed, many women have become nurses and labourers at the loom or in the vineyards owing to the misfortunes of the city in those days, women of civic birth, too; and many who were poor then are now rich.77 We cannot of course know if Euxitheus is telling the truth, since we do not even have the speech of the opposing side. But if he is, it is indeed the lack of distinction between the lives and employments of his parents and those of slaves and freedmen, which gave place to calumnies that eventually secured his disfranchisement. There were thousands of citizens who led lives similar to those of Euxitheus’ parents.78 Obviously, it was Euxitheus’ success in political life which brought the question of the true legal status of his parents to the centre of attention. For the majority of ordinary citizens this would have rarely become an issue. 73  Demosthenes, Against Eubulides, 18. 74  See also Cohen, “Women, property and status,” 58–61. 75  Bäbler, Fleissige Thrakerinnen, 37–43. 76  Demosthenes, Against Eubulides, 42. 77  Demosthenes, Against Eubulides, 45. 78  See also Silver, “Slaves versus free hired workers.”

354

Vlassopoulos

Let us now look at this phenomenon from the opposite direction: how could slaves or freedmen exploit this lack of distinction for their own benefit? The speaker in Lysias 2379 explains that having certain grievances against an individual called Pancleon he decided to prosecute him. Believing that Pancleon was a metic, he went to the fuller’s, where Pancleon was working, and summoned him to the court of the Polemarch, who was the magistrate in charge of suits involving metics. Pancleon countered that he was indeed a Plataean by origin, and therefore, since the Athenians had granted citizenship to all Plataeans, a citizen (§2). Therefore, he could not be summoned to the Polemarch’s court. The speaker then relates how he tried to ascertain whether Pancleon was indeed a Plataean, only to find out that no Plataean had ever heard of him. The only exception was a Plataean who claimed that he had a slave by the name of Pancleon (§§7–8). The next thing narrated is an event in which Pancleon is confronted by a man who claims that Pancleon was indeed his slave. What saved Pancleon in these circumstances was the people whom the speaker loosely describes as ‘those present’ (§§9–11). These bystanders stood surety for Pancleon claiming that he had a citizen brother who would vindicate his freedom through the process of aphairesis eis eleutherian. Next day in the court no brother of course appeared, while a woman came forward to claim Pancleon as her own slave; but according to the speaker Pancleon’s supporters just carried him off by force and went away. Finally, in a last piece of investigation, the speaker claims to have found out that Pancleon lived for a certain time as a metic in Thebes. Given that Thebes was the city that had destroyed Plataea, it was highly unlikely that Pancleon would have lived there if he was truly a Plataean (§15).80 Here we have an individual whom his opponent takes as a metic, others claim as their slave, while he insists that he is a citizen. It is of course impossible to be certain today what was actually the truth, though the speaker seems to have a good case that Pancleon was a metic, if not a slave or former slave. Even more, when challenged to declare his deme, Pancleon stated that he belonged to the deme of Decelea. Given the fact that Decelea was profoundly disturbed during the Peloponnesian War a few years earlier, and the problematic relationship between membership in the deme of Decelea and membership in the oikos of the Deceleans, there is a possibility that Pancleon deliberately chose this particular deme in order to make detection more difficult.81 How could a person like Pancleon find out about this blind spot of Athenian insti79  Lysias, Against Pancleon. 80  Zelnick-Abramovitz, Not Wholly Free, 296–98. 81  Rhodes and Osborne, Greek Historical Inscriptions, 38–9; Jones, Associations, 83–6.

Slavery, Freedom and Citizenship in Classical Athens

355

tutional history? Can we suppose that he acquired this knowledge from his citizen friends, who were ready to rescue him? Thus, it may not just have been the lack of distinction between slaves, freedmen and foreigners on the one hand, and the lower-class citizens on the other, which facilitated the blurring of identities in classical Athens. It was also the case that the shared experiences of working and living together allowed slaves and freedmen to muddy the waters about their status and even assume citizen identity.82 The orator Isaeus gives us a final piece of evidence to consider regarding the connection between subaltern communities and the blurring of identities is in place: Next consider, in the first place, what motive our father could have for lying and for having adopted Euphiletus as his son, if he were not really so. You will find that all those who do such things either have no legitimate children of their own, or else are constrained by poverty to adopt aliens in order that they might receive some assistance from them, because they are indebted to them for their Athenian citizenship.83 To return to Finley’s initial pronouncement: precisely where the distinction between slave and free was most clearly made, it also seemed very difficult to decide the status of many individuals. This was the result of the interlinking of two different factors. One factor was the social content of Athenian democratic citizenship: the fact that the Athenian lower classes, the peasants, the artisans and the traders, possessed significant political rights. The other factor was the economic structure of Athens and the lack of distinction between free and slave in most professions. As the Old Oligarch commented with fury, one had better avoid assaulting a badly dressed person for fear of assaulting a lower class citizen rather than a slave or freedman. It was the combination of the political rights of lower-class citizens and the impossibility of distinguishing them from many freedmen and slaves which created this blurring of identities. Many slaves and freedmen like Pancleon took advantage of the impossibility of distinguishing between free, citizen, slave and freedman on the basis of profession or living conditions, in order to enhance their position, avoid detection and even escape slavery. On the other hand, this blurring of identities could be variously used to challenge somebody’s citizenship or freedom, to attack a political opponent, or to acquire money through bribery or extortion. Slavery, freedom and democracy were linked together in more ways than Finley supposed. 82  Vlassopoulos, “Free spaces,” 39–47. 83  Isaeus, On behalf of Euphiletus, 1–2.

356

Vlassopoulos

Acknowledgements The author would like to thank Rachel Zelnick-Abramovitz, Niall McKeown and Virginia Hunter for their help. Bibliography Bäbler, Balbina. Fleissige Thrakerinnen und wehrhafte Skythen. Nichtgriechen im klassischen Athen und ihre archäologische Hinterlassenschaft. Stuttgart and Leipzig: B. G. Teubner, 1998. Brun, Patrice. “Hyperbolos, la création d’une ‘légende noire’.” Dialogues d’histoire ancienne 13 (1987): 183–98. Cairns, Francis. “The civic status of Theodotus in Lysias III.” Emerita 70 (2002): 197–204. Cataldi, Silvio. “Akolasia e isegoria di meteci e schiavi nell’ Atene dello PseudoSenofonte: una riflessione socio-economica.” In L’opposizione nel mondo antico, edited by M. Sordi. Milan: Vita e Pensiero, 2000. Cohen, David. “Women, property and status in Demosthenes 41 and 57.” Dike 1 (1998): 53–61. Cohen, Edward E. Athenian Economy and Society: A Banking Perspective. Princeton, NJ: Princeton University Press, 1992. Cohen, Edward E. The Athenian Nation. Princeton, NJ: Princeton University Press, 2000. Davies, John K. Athenian Propertied Families, 600–300 B.C. Oxford: Oxford University Press, 1971. Diggle, James. Theophrastus: Characters. Cambridge: Cambridge University Press, 2004. Finley, Moses I. Ancient Slavery and Modern Ideology. London: Chatto & Windus, 1980. Finley, Moses I. Economy and Society in Ancient Greece. London: Chatto & Windus, 1981. Fisher, Nick R. E. Aeschines: Against Timarchus. Oxford: Oxford University Press, 2001. Fisher, Nick R. E. “‘Independent’ slaves in classical Athens and the ideology of slavery.” In From Captivity to Freedom: Themes in Ancient and Modern Slavery, edited by C. Katsari and E. Dal Lago. Leicester: University of Leicester Press, 2008. Gagarin, Michael. “The torture of slaves in Athenian law.” Classical Philology 91 (1996): 1–18. Harris, Edward M. “Did Solon abolish debt-bondage?.” Classical Quarterly 52 (2002): 415–30. Harrison, A. R. W. The Law of Athens, Vol. I: The Family and Property. Oxford: Oxford University Press, 1968. Humphreys, Sally C. “Kinship patterns in the Athenian courts.” Greek, Roman and Byzantine Studies 27 (1986): 57–91.

Slavery, Freedom and Citizenship in Classical Athens

357

Hunter, Virginia J. Policing Athens: Social Control in the Attic Lawsuits, 420–320 B.C. Princeton, NJ: Princeton University Press, 1994. Hunter, Virginia J. “Introduction: Status distinctions in Athenian law.” In Law and Social Status in Classical Athens, edited by V. Hunter and J. Edmondson. Oxford: Oxford University Press, 2000. Hunter, Virginia J. “Pittalacus and Eucles: Slaves in the public service of Athens.” Mouseion 6 (2006): 1–13. Jacob, Oscar. Les esclaves publics à Athènes. New York: Arno Press, 1979. Johnstone, Steven. Disputes and Democracy: The Consequences of Litigation in Ancient Athens. Austin: University of Texas Press, 1999. Jones, Nicholas F. The Associations of Classical Athens: The Response to Democracy. New York and Oxford: Oxford University Press, 1999. Kapparis, Kostas. “The Athenian decree for the naturalization of the Plataeans.” Greek, Roman and Byzantine Studies 36 (1995): 359–78. Klees, Hans. Sklavenleben im klassischen Griechenland. Stuttgart: Franz Steiner, 1998. Lacey, William K. “The family of Euxitheus (Demosthenes LVII).” Classical Quarterly 30 (1980): 57–61. Lauffer, Siegfried. Die Bergwerkssklaven von Laureion. Wiesbaden: Franz Steiner, 1955/6. Mirhady, David C. “Torture and rhetoric in Athens.” Journal of Hellenic Studies 116 (1996): 119–31. Morris Silver. “Slaves versus free hired workers in ancient Greece.” Historia 55 (2006): 257–63. Ober, Josiah. Mass and Elite in Democratic Athens: Rhetoric, Ideology and the Power of the People. Princeton, NJ: Princeton University Press, 1989. Osborne, Michael J. Naturalization in Athens. Vol. I, Brussels: Palais der Academiën, 1981. Osborne, Michael J. Naturalization in Athens. Vol. II, Brussels: Palais der Academiën, 1982. Perotti, Elena. “Esclaves choris oikountes.” In Actes du Colloque 1972 sur l’esclavage. Paris: Les Belles Lettres, 1974. Randall, Richard H. “The Erechtheum workmen.” American Journal of Archaeology 57 (1953): 199–210. Rhodes, Peter J., and Robin Osborne. Greek Historical Inscriptions 404–323 BC. Oxford: Oxford University Press, 2003. Rihll, Tracey. “The origin and establishment of Ancient Greek slavery.” In Serfdom and Slavery: Studies in Legal Bondage, edited by M. L. Bush. London and New York: Longman, 1996. Scafuro, Adele C. “Witnessing and false witnessing: Proving citizenship and kin identity in fourth-century Athens.” In Athenian Identity and Civic Ideology, edited by

358

Vlassopoulos

A. L. Boegehold and A. C. Scafuro. Baltimore and London: Johns Hopkins University Press, 1994. Schumacher, Leonard. Sklaverei in der Antike: Alltag und Schicksal der Unfreien. Munich: C. H. Beck, 2001. Thür, Gerhard. “Komplexe Prozessführung. Dargestellt am Beispiel des Trapezitikos (Isokr. 17).” In Symposion 1971. Vorträge zur griechischen und hellenistischen Rechtsgeschichte, edited by H. J. Wolff. Cologne and Vienna: Böhlau Verlag, 1975. Thür, Gerhard. Beweisführung vor den Schwurgerichtshöfen Athens: die Proklesis zur Basanos. Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 1977. Thür, Gerhard. “Der Streit über den Status des Werkstättenleiters Milyas (Dem., or. 29).” In Demosthenes. Wege der Forschung, edited by U. Schindel. Darmstadt: Wissenschaftliche Buchgesellschaft, 1987. Todd, Stephen C. The Shape of Athenian Law. Oxford: Oxford University Press, 1993. Todd, Stephen C. A Commentary on Lysias, Speeches 1–11. Oxford: Oxford University Press, 2007. Trevett, Jeremy. Apollodorus, the Son of Pasion. Oxford: Oxford University Press, 1992. Vlassopoulos, Kostas. “Free spaces: identity, experience and democracy in classical Athens.” Classical Quarterly 57 (2007): 33–52. Whitehead, David. “The mysterious ‘Cyreneans’ in [Demosthenes] 59.9.” Classical Quarterly 56 (2006): 317–21. Worthington, Ian. A Historical Commentary on Dinarchus: Rhetoric and Conspiracy in later Fourth-century Athens. Ann Arbor: University of Michigan Press, 1992. Zelnick-Abramovitz, Rachel. Not Wholly Free: The Concept of Manumission and the Status of Manumitted Slaves in the Ancient Greek World. Leiden: Brill, 2005.

Justifications: Barbarians and Natural Slaves N. R. E. Fisher Slavery, in one form or another, was universal in the ancient cultures known to the Greeks; they grew up with it and accepted it, together with its inherent cruelty and contradictions, largely without question. Slight hints, however, of some intellectual unease are to be found from the late fifth century on, as part of the ferment of radical questioning brought to the Greek cities by the Sophists. Their theorising, and their debates with Socrates and his varied pupils, seemed to raise some doubts about the standard view that slavery was natural and just. But the dominance of this view was not seriously affected, particularly if the slaves were not Greek.

Slavery and the Greek/Barbarian Distinction

Originally, enslavement may been been accepted as a stroke of ill fortune which might happen to anyone, and which would then have a catastrophic effect on one’s personality as well as on one’s way of life. The view that foreigners (‘barbarians’) deserved to be enslaved, whereas Greeks deserved to rule, may have begun to spread during the sixth century as increasingly slaves were being imported from the neighbouring nonGreek areas. But it was undoubtedly the Greek victories in the great Persian Wars (490–480 bc), and the subsequent extension of Athenian sea-power to areas previously ruled by Persia throughout the Aegean and the coast-line of Asia Minor, which together gave the view overwhelming force. The bravery, military skill, and political freedom and innovations of the citizens of the Greek poleis made them greatly superior, they felt, both to the strong and tough, but politically disorganised, peoples to the northeast of Greece (such as Thracians and Scythians), and equally to the many soft and weak Asiatic peoples mostly ruled by the Persian empire. All these peoples who enjoyed no political freedom or control over their lives at home were thought fit to be ‘enslaved’ in Greece, and even to benefit from their contacts with the superior Greek culture. This belief, applied to the subjects of the Persian Empire, may have been strengthened by the Greeks’ awareness that Persian Kings routinely referred to all their subjects, including their high Source: Fisher, N. R. E., “Justifications: Barbarians and Natural Slaves,” in N. R. E. Fisher, Slavery in Classical Greece, London: Bristol Classical Press, 1993, 86–98. © N.R.E. Fisher, 1993, Slavery in Classical Greece, and Bristol Classical Press, an imprint of Bloomsbury Publishing Plc. © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_015

360

Fisher

officials such as satraps, as their ‘slaves’; for example the Greek version, preserved on stone of a letter from King Darius to a satrap in Ionia uses the word doulos of the satrap (see Fornara, Archaic Times no. 12). This may also, along with the sharpening of the idea of the free citizen, have helped to make doulos from the early fifth century on, the standard and commonest Greek term for a slave. Ideas and stereotypes of this sort can be found throughout Greek drama and the speeches of the orators, usually crudely asserted but occasionally (especially in some plays of Euripides) more subtly open to question. The connected purposes of Herodotus’ Histories are to present the history and cultures of the peoples to the East with whom the Greeks came into contact (mostly the Persians and the people whom the Persians conquered or tried to conquer), and to give an account of the Great Persian Wars. It is built on very complex patterns of contrasts between Greek political systems and culture and those of the variously different non-Greek peoples he considers. Overall, the work suggests that the Greeks were able to defeat the much larger forces of the Persians because of their much greater commitment to fight for their polis-based conceptions of political freedom and because of their preparedness, in a crisis, to unite against the common threat. Such ideas also play a significant part in an important work of fifth-century Greek ‘scientific’ theory, the Airs, Waters, Places, one of the collection of medical texts associated with the great founder of scientific medicine, Hippocrates, and written some time during the later fifth century bc The unknown author offers two different types of explanation for the characteristics of the various peoples he considers. The first is the regularity or irregularity of climatic conditions; the second is the contribution made by the political and social institutions. Hence the Asiatic peoples are said to be less courageous, less warlike and weaker than Europeans, first because of climatic conditions, namely that the seasons are uniform and equable and without sudden changes of temperature which might stir up wild passions and harden resolve. But another, political, reason for these characteristics is then given, namely their nomoi, that is their customs, laws and political institutions. Since they are mostly ruled by kings they have much less incentive to be warlike, because the military service demanded of them ‘necessarily involves suffering and death on behalf of their masters, far from their children, wives and other friends’ (ch. 16). Hence the author holds that men who are naturally brave and courageous can be ruined by such nomoi (for example, perhaps, Greeks living under Persian rule), whereas those peoples in Asia (Greek and non-Greek) who live independently of despotism are the bravest in all Asia, Similarly, the Scythians’ peculiarities are explained both by the (highly exaggerated) coldness and wetness of

Justifications: Barbarians And Natural Slaves

361

Figure 11 On Athenian vases, ornamental tattoos on necks, arms and legs (as opposed to tattooed or branded marks imposed as punishment on runaway slaves) are often found, and indicate ‘barbarian’ women, especially those from Thrace. So these girls fetching water from a fountain are identified as Thracian slaves. An Attic water-vessel (hydria) by the Aegisthus Painter, c. 470 BC.

their climate, and by their nomadic and excessively equestrian habits. And other Europeans (he has Greeks above all in mind) are varied in physique because of the different climates: in character, they are often fierce, anti-social and spirited because of the changeableness of the climate, but it is also their political independence which explains why they are so brave and so prepared to risk death. Overall, while the work shows an interest in making precise observations and seeking ‘scientific’ explanations, it can be seen grossly to exaggerate both the climatic variations and the effects of political systems in order to reach predictable conclusions about the differences between Asiatics, Greeks and Scythians. These conclusions, conveniently, served two purposes. First, they

362

Fisher

explained, as had Herodotus, the military superiority of Greek soldiers so evident—to Greek eyes—in the wars of the fifth century; and second, they justified the use of such inferior ‘barbarians’ as slaves in the Greek world. But of course the more these justifications were seen as fundamental, the more problematic became the continued enslavement in the Greek world of other Greeks.

Hints of a Critique of Slavery?

Similar climatic and political explanations of the inferiority of barbarians are found in the fourth century philosophies of Plato and Aristotle. But before considering their treatment of the issues, the very limited evidence for some intellectual challenges to the justification of slavery as ‘natural’ should be considered. The debate was largely conducted in terms of a number of basic ‘polarities’, contrasts between opposed terms, of which the most important were slave/free, Greek/non-Greek, and the major analytical tool of the Sophists, the distinction between physis (nature) and norms (custom, law, habit or convention). Until lately, it was held that Antiphon, one of the late fifth-century sophists, expressed radical views that attacked the validity or ‘naturalness’ of the distinctions between high and low birth and between Greek and barbarian, and hence was probably also questioning (in the name of the ‘unity of mankind’) the justification of the enslavement of non-Greeks. The evidence comes in fragments from his sophistic dialogue ‘On Truth’, found on a papyrus in Egypt. But a small extra fragment discovered recently has changed our view of his argument. Antiphon now seems to have been arguing that all men who approve of the laws (nomoi) of those around them, and disapprove of those further away, ‘have been made barbarians with regard to each other’; but in fact men are all alike in their basic physiological natures and needs. It seems, then, that this relativist argument suggested that the Greeks were not alone— or necessarily right—in despising the laws and habits of other peoples, and treating them as ‘barbarians’; and it was probably part of Antiphon’s more general, rather amoral, position, which was to allow all men the right to pursue their own ‘necessary’ desires, which is the main point that is made elsewhere in this fragment. So there seems no reason now to think that Antiphon either attacked the distinction between high and low birth, or that he explicitly developed his point about the shared physical characteristics of all humans into a criticism of slavery.

Justifications: Barbarians And Natural Slaves

363

The only certain fifth-century evidence for criticism of the justice of slavery is to be found in a number of statements made by various characters in the surviving plays of Euripides, or in similar fragments from his plays which are now lost. Of course, since they are statements made by characters in plays one cannot easily assume that these statements, or indeed the many opposing statements in his plays which assert the justice of the enslavement of foreigners by Greeks, represent the views of Euripides himself. But the plays do provide evidence that theoretical and provocative statements criticising some fundamental Greek attitudes and ideologies were made on the Athenian stage in the late fifth century, and it is reasonable to suspect that, as in other cases of radical ideas in Euripides, these theatrical debates were echoing ideas expressed by some of the more radical of the sophists. A few such opinions can be quoted here. Two central points are made more than once. First, after scenes in which individual slaves show qualities of loyalty, intelligence or bravery equal, or superior, to those shown by free people, characters in the plays often conclude that the institution of slavery often operates unjustly, enslaving and degrading those who do not deserve it. For example, in the Ion a loyal slave, prepared to risk his life for his mistress, claims that One tiling alone brings shame to the slaves, the name; apart from all of that, a slave is no worse than free men in anything, if he is good. Euripides, Ion 855–7

One should perhaps also note that in this passage (and in a similar one in Helen 726–33) the noble slave shows qualities of loyalty and devotion to the interests of his master which involve acceptance of his status as a slave. The same is probably true of another such statement, for which we have no context: For many slaves the name is shameful, but their minds have more freedom than the minds of those who are not slaves. fr. 833

Yet more radically, in his Alexandros (a play about the early life of Paris), it seems that Paris, who has been brought up as a slave, succeeds in his claim to be permitted to compete with the free men in heroic games at Troy. Interpretation of the fragments is not easy, but it looks as if the following views are expressed: all men originally had the same appearance; noble birth, free birth and race do

364

Fisher

not guarantee good moral qualities; and while some slaves are enslaved only in name, for others it is an appropriate fate. Second, in plays focusing on the consequences of war for the defeated victims, great pity is aroused at the enslavement of the women and children, for example, of the house of Troy in Andromache, Trojan Women and Hecuba. These may have had the effect of causing Greeks some anxiety about the justice of enslaving those defeated in war. Further, since there is considerable emphasis, above all in the Trojan Women, on the injustice and callous cruelty of the victorious Greeks, and considerable sympathy for the terrible sufferings and humiliations of the Trojan women who are represented as ‘barbarians’ by the Greeks, serious questioning of the Greek-barbarian contrast and the orthodox view of the moral superiority of the Greeks is also a significant part of the effects of these plays. We cannot now identify the sophistic sources of the ideas expressed in these plays. Aristotle started his discussion of the issues of slavery with an admission that there were those who argued that the whole institution was unnatural because the distinction between slave and free was a matter of nomas (agreed convention), not of physis (nature), and was also unjust because it was based on force (Politics 1253b15–23). But he gives no hint whether such ideas were expressed in the late fifth century or closer to his own time in the second half of the fourth century. One such statement of which Aristotle was aware was made by Alcidamas, a teacher of rhetoric and a rival to Isocrates as a pupil of the sophist Gorgias. Aristotle refers to a statement in Alcidamas’ pamphlet written in defence of the Messenian ex-helots who had set up their new state in 370 bc, which proclaimed a universal precept about justice(Aristotle, Rhetoric 1373b19); and the relevant sentence is quoted by the ancient commentators (Scholia) on the passage; ‘God left all men free; nature made no man a slave’. This far-reaching assertion was almost certainly used to condemn the longstanding injustice of the helot-system in Sparta. Moreover, it seems likely that sophistic use of the nature/convention distinction in relation to slaves and free men sharpened many Greeks’ distaste for the enslavement of their fellow Greeks, and hence increased interest in the origins of the helot-system and sympathy for the Messenians and other Greeks enslaved in wars or by piracy. But we do not know which thinkers applied, as Aristotle says some did, these ideas to chattel slavery of non-Greeks, nor how far they may have pressed their conclusions. Did they admit that it was ‘against nature’, and ‘unjust’, but leave it there? Did they perhaps claim that though this was so, it was necessary and had to be accepted (as many philosophers and lawyers later, for example under the Roman Empire, were to argue)? Or did they go further and suggest that it should, however gradually, be abolished? We have to state firmly that on existing

Justifications: Barbarians And Natural Slaves

365

evidence we cannot be sure that anyone in classical Greece took that more radical step and proposed the abolition of chattel slavery. What we can be certain is that there was nothing remotely resembling an ‘abolitionist movement’.

Slavery in Plato’s Thought

Plato did not offer an extended account and defence of slavery; but, as we saw in chapter 5, he offered some guidance on the treatment of slaves. He also made passing comments on the naturalness and propriety of slavery throughout his work. Furthermore, a number of his major ideas of the proper rule and hierarchy in society, and in the soul of the individual, are developed precisely in terms of the idea of the justified rule of masters over slaves. First, it is perfectly clear that in both Plato’s ‘ideal states’, the startlingly radical, if not fully detailed, account of the Republic (written perhaps during the 370s bc) and the more realistic and much more detailed Laws (his last work, worked on perhaps up to his death in 347 bc), slaves are assumed to be essential to the management of the good state. Indeed, in the Laws, as was suggested in chapter 5, Plato’s proposed laws affecting slaves, and his advice on their treatment, are somewhat stricter than those current in the Athens of his time. Second, it seems from a number of passages that Plato would justify slavery on the basis of the slave’s fundamental mental inferiority. For example, he claims that there are, properly, two types of medical practice (Laws 720a–e). Doctors treat their free clients by constructing a history of the disease through detailed communication with the patients and their friends, and give careful instruction to them in the disease and the treatment to be followed; but they treat slaves swiftly, giving them no account of the disease but prescribing treatment peremptorily, ‘like a tyrant’. He also suggests that by and large free doctors treated free patients, and slave assistant-doctors treated slave patients. However, it is interesting that the detailed records of treatments contained in the Hippocratic works do not bear out such a division of their work. Plato’s reason for the practice he recommends is that slaves lack the mental capacity to understand principles or give reasons for their beliefs; they cannot be left to decide anything for themselves, but must always be told what to do, and compelled by force if necessary. There are clear signs that Plato took such views because he accepted up to a point the basic ethnographic orthodoxy. For example, in the Republic (435d– 6a) he deploys the accepted stereotypes of the Thracians, Scythians and northern peoples as especially spirited, the Greeks as rational and intelligent and the Phoenicians and Egyptians as keen on money-making, while in another

366

Fisher

work (Statesman 262d) he criticises the over-simple lumping together of all non-Greeks as sharing precisely the same inferior characteristics. Plato’s preparedness, which he never fully defended, to assume that at least some barbarian peoples were mentally inferior to the Greeks was strongly reinforced by his keen desire to promote harmony between the Greek states and to encourage them to fight wars against ‘barbarians’ rather than against each other, to treat only the ‘barbarians’ as natural enemies, and to restrict the enslaving of the defeated enemy (as also the serious destruction of land and buildings) to wars against foreigners, not those between Greeks (Republic 469b–471c). The slaves in Plato’s state would be predominately non-Greek, and considered ‘naturally’ inferior. At times, however, he writes as if slavery would be an appropriate punishment for any in a community who deserved it because they were hopelessly sunk in ‘ignorance and feebleness’ (Statesman 309a). Thirdly, Plato’s view that power should be held by the intellectual elite (the ‘philosopher Kings’ in the Republic), who alone have true knowledge and understanding of the nature of the world and of justice, is often expressed in ways that suggest that only the rulers have ‘true’ freedom to determine their own lives and the direction of the state; all the rest, he suggests, whether notionally free workers, soldiers, or actual slaves, are best placed in a relationship of subjection or slavery (douleia) to the rulers. That is to say, that while the distinction between slave and free in Plato’s ideal communities remains fixed and firm, most of their free populations should obey their intellectual betters in a manner comparable to the way good slaves obey their masters. In a similar fashion, Plato divided the ‘soul’ of the individual into three parts. The best part, which should rule, is the capacity to reason and to understand true principles; the other two inferior parts, which should be controlled and obey the best part, are, first, the ‘spirited’, the seat of courage and the sense of personal honour, and, second, the seat of the desires. The soul should obey the same hierarchical rules as should the state, and hence the two worse parts should obey, like slaves, the rational part. It is those who find this difficult because they lack the developed powers of reason who must be compelled to obey the laws of society developed by those in whom reason does reign supreme. This seems to be the case even for Greeks, in whom the reasoning power is especially well developed. In these senses, then, the principle of justified slavery can be seen to hold a central place in some of the most important areas of Plato’s thought.

Aristotle’s Defence of the Theory of Natural Slavery

It was Aristotle, Plato’s greatest pupil, who confronted directly the issue of the moral justification of slavery. His main account comes in the first book of his

Justifications: Barbarians And Natural Slaves

367

Politics (1253b–1255b, also available in GARS 2), written during the period 335– 323 bc He views slavery as a natural and necessary element in the city-state, which he saw as the social system in which men could achieve the best form of civilization and individually attain the good life. As we saw, he was aware that a few thinkers had challenged the naturalness of slavery. Nonetheless he defended the institution, and saw its growth as part of the necessary progress from loose collections of households or villages to the cohesive city-state. His defence, complex and thoughtful though it is, is flawed by fundamental contradictions and illogicalities, as well as by the standard Greek-centred contempt for foreigners and insensitivity to slaves. This is in fact only to be expected in an attempt rationally to defend what seems to us to be patently indefensible. It is a testimony to the force of the ideological need to defend a basic social institution that such a profound and often fairminded thinker was content with such weak arguments. Unlike Plato, Aristotle did not see the rule of master over slave as the model for other types of rule or hierarchy. He carefully distinguished the rule of masters over slaves from other types of rule such as husbands over wives, parents over children, and various types of political authority. Rule over slaves is ‘despotic rule’, operating essentially in the interest of the masters. His project in defending such ‘rule’ involved at least four distinct elements: first, to define the characteristics and deficiencies that made some people suited only to be ‘natural slaves’; second, to establish that sufficient numbers of identifiable people existed who fitted the role; third, to establish that these were the people who, mostly, were in fact employed as slaves in Greece; and fourth, to establish what form of slave-management would be effective and beneficial to both members of the master-slave relationship. In commenting briefly on how well Aristotle carried out this project, I shall bring out how his conception of the nature of the slave changes in the course of the arguments. The contradictions we have already noted in the images which other Greeks had of slaves, seeing them as tools, as animals, as perpetual children, or as almost like free humans, can be shown to cause decisive difficulties in Aristotle’s account too. Aristotle starts his account by saying that slaves are human beings who belong completely to their masters. He sees them as living pieces of property, who operate in ways comparable to the inanimate tools the master also possesses, but suggests that since slaves help us in many of our activities, they should not be defined as tools for production, but rather as tools for action. Slaves, like other humans—but also like animals—have a body and a mind (or soul). Hence slaves naturally suited for slavery should possess both a body suitable for their tasks—i.e. a body suited for indoor, bending-down activities (but less suited for military, farming or community activities), and also have the mental capacity to ‘share in reason sufficiently to understand it, but not to possess it

368

Fisher

themselves’. Hence, like Plato, Aristotle holds that natural slaves were seriously deficient in the mental powers to make their own decisions and develop their own arguments, and so benefited by being permanently controlled and commanded by their masters. But what is involved in this idea of ‘lacking reason’ is not always clear, as there are significant variations in his images of the slave. Sometimes natural slaves seem to him to be little better than tame animals, or to relate to their masters as the body relates to the mind, as beings able only to respond to simple commands and happiest when simply obeying their masters. In other parts of his argument he thinks of them as closer to children (which was also a common Greek view), capable both of logically understanding the reasons for doing things, and of being trained to feel the correct emotions. The failure to sort out these differing conceptions has fatal results for his argument. As for the second element in the project, Aristotle’s theory requires that ‘nature’ should provide an abundant supply of easily identifiable people whose bodies were fit for slavish work, and whose minds lacked full reasoning powers. He recognises, however, that nature very often fails to achieve both of these aims; for example, he notes that many with the minds of free men have less than ideal bodies. In deciding who should be slaves, he puts a greater weight on the mental criterion; but he does note the difficulty that this mental deficiency is less easily observable than are physical weaknesses. Where, then, does he think mentally inferior, natural slaves are to be found? He is less explicit than he might be on this, but it is in fact clear that he agrees with the standard classical Greek view that it is the ‘barbarian’ peoples who provide them. One of the major problems he admits—which gives strength to the view that slavery is only ‘conventional’, not ‘natural’—is that slaves are often created as a result of wars, and that this often leads to the ‘wrong’ people being enslaved. Even if one takes the view (about which he has in any case strong doubts) that victory in war suggests the excellence of the victors (which might or might not include their justice), many of those thus enslaved would not be so clearly inferior as to be natural slaves. Such people, Aristotle agrees, can be seen as only slaves by convention (nomos), not by nature (physis). The true natural slaves are evidently various peoples of the non-Greek world, although which peoples he had in mind are not made clear in this passage. Elsewhere in the Politics, however, he presents ethnographic views very like those given by the Hippocratic author and by Plato quoted above (see pp. 87ff. and 92ff.). The non-Greek Europeans and those living in cold areas are spirited, but lack intelligence, so that they live as free peoples but without political organisation; Asiatics have intelligence and craft skills, but lack spirit and live under a form of rule that is called ‘slavery’; the Greeks—in the middle—have the virtues

Justifications: Barbarians And Natural Slaves

369

both of spirit and intelligence, and hence ‘the Greek race lives in freedom, has the best political organisation, and would be able to rule all others, if it could attain a single political unity’ (1327b18–33). And, in another place, the barbarians generally are said to be ‘more slave-like’ than the Greeks, and among barbarians the Asiatics more so than the Europeans, because they endure despotic rule without discontent (1285a19–24). The ‘natural slaves’ seem thus to be identified, and it is their political weakness or anarchy at home which justifies their enslavement. However there is a clear contradiction here: the supposed general deficiency in the capacity to reason mentioned in Book 1 does not cohere with the failure in spirit, though not in intelligence, that makes the Asiatics apparently those foreigners most suited to slavery. On the third element, it is evident that Aristotle holds that for the most part in Greece those enslaved are the proper ones. This amounts to the view that it is in general wrong for Greeks to be enslaved, but proper for non-Greeks. In this, as we saw, he is at one with Plato and with a growing feeling in Greece during the fourth century. Whether he thought that there were some Greeks so deficient in reasoning capacities (and/or in spirit) that they too deserved to be slaves, seems not to emerge clearly from the text. Overall, then, Aristotle did accept, pretty uncritically, the blanket condemnation of non-Greeks and the extremely crude stereotyping of different non-Greek races common in his world. This attitude can equally be seen in his famous advice to his pupil, Alexander the Great, that he should treat the barbarians he was engaged in conquering as slaves and animals. In Aristotle’s treatment of the management of slaves—the fourth element in the project—the inconsistences and contradictions in his view emerge even more clearly. Only if ‘natural slaves’ have throughout their lives extremely limited mental capacities can it conceivably be justified to treat them in a uniquely ‘despotic’ way. In such cases, one would, presumably, simply give them orders, and there could be no question of any ‘friendship’ existing. In places Aristotle does take this view (for example, in Nicomachean Ethics 1161a32–3): because there is no justice or community in the relationship, there can be no friendship. In the passage in Book 1 of the Politics, however, he allows that there can be some sort of friendship if the slave is a ‘natural slave’, and, presumably, believes that the relationship benefits him or her as well as the master. But in many places Aristotle implicitly recognises that there were many aspects of master– slave relationships, and advantageous ones, which were based on a different conception of slave capacities. He suggests that slaves need to have some small amount of appropriate ‘virtue’ developed by their masters, and hence be given reasons for their instructions, and not merely told what to do. He notes that some slaves act as foremen or overseers, with independence and some ‘rule’

370

Fisher

of their own (1255b35–6); he asserts in the passage of the Nicomachean Ethics that one may develop friendship with a slave not inasmuch as he is a slave, but inasmuch as he is a man, which brings out the fundamental contradiction very clearly; and as we saw above, he advocates the use of manumission as a incentive for all slaves (Politics 1330a32–3), which is of course the single most glaring contradiction—and one firmly based in the institution—in the whole theory of the natural slave. If the slave was benefited by being a slave, freeing him would be unjust; if he had the potential to be free, he should be treated from the start more like a child needing training and development than like an animal. Thus we see that in these latter arguments Aristotle is operating with quite different models, in which the slave is potentially at least fully human, and his potential can be developed to the point where he can enjoy a limited degree of liberty as a freedman. But these models, while it may reflect aspects of the treatment offered to some privileged slaves, cannot justify the despotic rule which remained the basis of the whole slave system. The many contradictions the account reveals are, in fact, those inherent in the practical workings of the slave system in Greece (and in many other slaveowning states). Aristotle’s attempt to defend the indefensible usefully brings them out into the open. The evidence of his will, preserved, as are many others, in Diogenes Laertius’ Lives of the Philosophers, is interesting, if it is genuine; though some scholars think that all the philosophers’ wills thus preserved are later compilations. As it stands, it suggests that as an individual master he treated some at least of his slaves as more than ‘natural slaves’, and did (as he advised at Politics 1330a32–3) offer all his slaves the incentive of manumission. Some of his thirteen slaves are to be freed immediately, and set up securely, with slaves of their own; others are to be set free similarly on his daughter’s marriage; and the rest are not to be sold, but to be still employed, and ‘set free when they reach the right age if they deserve it’ (5.11–16 = GARS 95).

Novel Evidence for Roman Slavery* Keith Hopkins This is an article about Roman slavery, and an experiment with method. Substantively my objective is to sketch the slave’s experience of slavery, and the fears and anxieties which slavery evoked in Roman masters. I am interested here in thoughts and feelings. Methodologically I am trying a new tack by experimenting at length with lies, whereas most Roman history is purportedly aimed at the discovery of truths, through establishing facts or describing events which are known to have occurred.1 This article is built around a single source, which is an inventive fiction, a pack of lies, an anonymous accretive novella, composed and revised, as I suspect, over centuries, as a vehicle for comedy and manners. It is the biography of a slave, the only full-length biography of a slave surviving from classical antiquity; a text which as far as I know has never yet been used as the basis of historical reconstruction, probably because it is so obviously a fiction.2 But Source: Hopkins, Keith, “Novel Evidence for Roman Slavery,” Past and Present, 138 (1993): 3–27. By permission of Oxford University Press. * I should like to thank Mary Beard, Simon Goldhill, John Henderson and Richard Sailer for conversations and criticisms, and audiences at Berkeley, Cambridge, Copenhagen, Eton, Los Angeles, Oxford, Princeton and Victoria for indicating where I failed to be persuasive. In some, the text produced the same reactions as its hero, Aesop; out of respect, I have changed it little. 1  The dominant convention that history is primarily concerned with “truth” about events which have occurred hardly needs authentication. See, for example, Paul Veyne, Writing History, trans. M. Moore-Rinvolucri (Manchester, 1971), pp. 10–11: “history is a body of facts … History is the relating of true events”. The tactic of experimenting with conscious fictions as a source of socio-cultural history is not unknown, but still rare in ancient history. It is well known, even fashionable, in early modern history. See, for example, Stephen Greenblatt, Shakespearian Negotiations (Oxford, 1988). 2  The text is edited with all its known variants in Aesopica, ed. Ben E. Perry (Urbana, 1952). The main manuscripts are very well translated by Lloyd W. Daly, Aesop without Morals (New York, 1961). I owe a huge amount to Daly’s translation, though the translations in this article are my own, except where noted. The Life of Aesop is not dealt with, as far as I know, in any modern discussion of ancient slavery; it is not even mentioned, for example by W. L. Westermann, The Slave Systems of Greek and Roman Antiquity (Philadelphia, 1955); I. Biezunska-Malowist, L’esclavage dans l’Egypte gréco-romaine, 2 vols. (Wroclaw, 1974–7); Keith Hopkins, Conquerors and Slaves (Cambridge, 1978); or by M. I. Finley in, for example, Ancient Slavery and Modern Ideology (London, 1980), Economy and Society in Ancient Greece © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_016

372

Hopkins

before I get down to discussing this biography in detail, I want to dwell a little on a small number of anecdotes, more or less familiar to modern historians of the Roman world, drawn from formal histories and from a medical doctor’s observations. My objective here is to illustrate my underlying contention that the social history which can be squeezed from “real histories” and from fiction may be broadly similar, and that, for the interpretation of culture, there is little justification for privileging, one above the other.3 First, a historian—Tacitus. In AD 61, the senatorial mayor (praefectus urbi) of Rome was murdered in his own home by one of his slaves. According to Tacitus, the motive was disputed: was it because the master had promised his slave freedom, and had even agreed the price, only to welch on the deal? Or was it because of some homosexual rivalry between master and slave over a shared lover? Whatever the cause, under strict Roman law, if a slave killed a master in his own house, then all of his slaves living in the household were to be crucified. The murdered mayor was a rich noble, a former consul; in his town house alone, he had four hundred slaves. Their imminent execution caused a huge stir. There was a heated debate in the Roman senate. Some senators were for softening the traditional harshness of the law; they pleaded for mercy for the large number of slaves, including women and children, who were undeniably innocent of any complicity in the crime. But a majority of senators voted to uphold the law as it stood. How else, the traditionalists argued, could a solitary master sleep soundly among a whole gang of slaves, unless it was in the interest of each to protect him against any murderous conspirator? Foreign slaves, worshipping foreign gods or none, could be controlled only by fear. After all the traditional custom of punishing a cowardly legion by choosing by lot one soldier in ten, and then having him cudgelled to death by his former companions, sometimes involved murdering the bravest men. Making an example benefited the whole community, even if it sacrificed some individuals unjustly. This was the gist of a speech by the triumphant conservative lawyer. The Roman senate was persuaded. (London, 1981), or Classical Slavery (London, 1987). Keith R. Bradley, Slaves and Masters in the Roman Empire (Brussels, 1984), pp. 150–3, discusses the fables of Phaedrus, but not Aesop or his Life. Lawrence W. Levine, Black Culture and Black Consciousness (New York, 1977), esp. pp. 81 ff., discusses the meaning of slaves’ stories—but in an intellectual style quite different from that deployed here. 3  In calling fiction “lies”, I am colluding with conventional dichotomies: facts/fictions, truth/ lies, evidence/discardable opinions, even though the basic argument of this article is that fiction provides a usable and trustworthy representation of Roman culture. My intention here is not to squeeze fiction for facts, but to interpret fiction as a mirror of Roman thinking and feelings.

Novel Evidence For Roman Slavery

373

The populace was less impressed. A crowd wielding firebrands and stones tried to stop the mass executions, but the emperor Nero stood firm, had the route lined with soldiers, and all four hundred slaves were crucified.4 Slavery was a cruel and repressive institution, enforced by hatred and fear. “All slaves are enemies”, stated a Roman proverb. The hostility of Roman slaveowners to their slaves, and of slaves to their owners, lay just below the surface of Roman civilization like an unexploded volcano. To be sure, some slaveowners were probably kind and considerate to some of their slaves, whether out of affection, generosity or the self-interest which was prompted by a desire to safeguard valuable assets. Sometimes owners’ kindliness was projected into an idealized image of the grateful and loyal slave, who in an emergency even sacrificed his or her life to save the slave-owner.5 That favourable image was one side of the coin. The other, and dirtier side of slavery rested in slaves’ legal powerlessness and in their more or less complete dependence on their owner’s favour, which, as far as the slave knew, could turn at any moment from love to hate. It is not my concern here to describe the varied behaviour of Roman slaves and of their owners, nor to draw up a tentative balance-sheet of what was typical. Given the fragmentary state of surviving evidence, that would be impracticable. But the realities of Roman slavery, its scale and pervasiveness throughout the Mediterranean basin, its legal basis and its economic implications, necessarily frame any modern analysis, as they framed all ancient experiences, of slavery.6 Roman slavery was in significant respects distinct from 4  Tacitus, Annals, xiv.42 ff. His account of the senatorial debate may not be accurate, and may reflect later political concerns. This is the argument of Joseph G. Wolf, “Das Senatusconsultum Silanianum und die Senatsrede des C. Cassius Longinus aus dem Jahre 61 n. Chr.”, Sitzungsberichte der Heidelberger Akademie der Wissenschaften, Philhist. Klasse (Heidelberg, 1988). 5  For the proverb, see Seneca, Moral Letters, xlvii.5; Macrobius, Saturnalia, i.11.13. On the loyal slave, see Joseph Vogt, “Sklaverei und Humanität”, Historia Einzelschrift, viii (1972), pp. 83–96; Joseph Vogt, Ancient Slavery and the Ideal of Man (Oxford, 1974), pp. 129 ff.; for sharp criticism, see Finley, Ancient Slavery and Modern Ideology, pp. 93 ff. For an equally idealized, but more nuanced view of slave/owner relations within complex households, see Paul Veyne, La societé romaine (Paris, 1991), pp. 16 ff. 6  See Keith Hopkins, Conquerors and Slaves (Cambridge, 1978), pp. 99 ff., for a quasi-objective account of Roman slavery, complementary in intellectual style to this subjective study of relations between masters and slaves, or the other works on slavery cited in n. 2. We have only very sketchy knowledge about how the practice and experience of slavery varied by period and subculture in the diverse and vast Roman empire. I use the term “Roman slavery” rather crudely to refer to slavery in the Roman empire, without implying either uniformity

374

Hopkins

slavery in the American South. For example, Roman slaves were not distinguished by the colour of their skin. Some Roman slaves were educated, sometimes better educated than their masters. Some slaves occupied positions of trust and responsibility, for example, as secretaries, clerks, teachers, physicians and architects, or as administrative agents in business or the management of agricultural estates. Many of these skilled slaves could save up out of profits or pocket money in the hope of buying their freedom. Many thousands of Roman slaves achieved freedom either through purchase or by the master’s generosity (or both—the master “generously” granting the slave the right to buy his freedom). And any male slave, freed in due legal form by a Roman citizen master, himself became a Roman citizen; and a similarly freed female slave could bear free citizen children by a citizen father. The Roman slave system therefore faced a recurrent problem on a scale unknown in the American South: the problem of the clever, talented, educated slave occupying a position of responsibility, who had a realistic prospect of freedom and the constant image before his or her eyes of other slaves who had themselves achieved freedom. At the same time, such slaves were in their owners’ power, and at the mercy of their whims. The visibility and social prospects of powerful and clever slaves, whose ability owners wanted and needed to harness, intensified the implicit struggle between masters and slaves and the oppressive cruelty of the Roman slave system. One objective of this essay is to investigate the seamier side of slavery, by analysing some of the stories which slave-owners (and perhaps even slaves) told, at least partly to help themselves delineate and negotiate the boundaries of appropriate behaviour between masters and slaves. These ancient stories are not about normal behaviour, any more than a modern popular newspaper typically reports news of normal conformist behaviour. Stories and news, and perhaps social history, concentrate on the boundaries of morality, or what one can call the penumbra of moral ambiguity, about which even the same reader, let alone different readers, will have had ambivalent feelings and reactions. In some respects what I am doing in this article is like trying to understand conditions in British factories during the depression in the 1930s from an English intellectual’s eclectic library, the one surviving copy of Charlie Chaplin’s Modern Times and a few fragments from British parliamentary papers and The Times. The interpretation of comic caricature is necessarily individualistic and errorprone, but it would be dangerous to assume that serious newspapers or the

or consistency—just as one might use the term “bread” to cover bread in France, England, Russia and the U.S.A., in spite of the known variations.

Novel Evidence For Roman Slavery

375

minutes of the Roman senate are more secure guides to the partial recovery of a lost reality than creative fiction. Roman literature abounds with stories of incidental cruelty to individual domestic slaves, who in the owner’s opinion had betrayed a trust, done something wrong, or simply had not done it fast enough. Slaves were there to be impatient with. Domestic slaves stood in the front line. They were more privileged and pampered than the tens of thousands of slaves who laboured without realistic hope of freedom in the fields or in the mines. Some domestic slaves, as I have said, could save up in the hope of buying liberty.7 But domestic slaves also had more contact with their owners, and were more often subjected to their despotic whimsy. Even “good” masters could be brutish. The emperor Augustus, for example, who by and large has had a favourable press in history books both ancient and modern, is said to have had the legs of a trusted slave broken because he had taken a bribe and revealed the contents of a letter; and Augustus is also said to have had a middle-ranking palace official, a freed slave, nailed to a ship’s mast because he had eaten a prize fighting-quail. The court physician Galen records that the emperor Hadrian once in anger stabbed a slave in the eye with a pen. He later felt remorse, and asked the slave to choose a gift in recompense. The slave did not reply. The emperor pressed him for a response. The slave said that all he wanted was his eye back.8 Individual stories are difficult to interpret. Some were recorded incidentally, almost casually; or so it appears. Others, perhaps the majority, were told because they marked the frontiers of acceptability. We have to be wary of constructing an image of Roman normality out of what even Romans regarded as exceptional. And yet these individual stories do help us map out Roman 7  In the one area of Roman Greece for which testimony is available, a high proportion (63 per cent) of freed slaves were female; the price of bought freedom was significantly high (on average about 3 tonnes of wheat equivalent for an adult slave, enough to provide minimum subsistence for a family of four persons for three years); and the conditions imposed on the newly freed slaves were often onerous, including (in a nearby area) the provision of replacement slave children: ibid., pp. 133 ff. For insubstantial quibbles about one of these conclusions, see Richard P. Duncan-Jones, “Problems of the Delphic Manumission Payments”, Zeitschrift für Papyrologie und Epigraphik, lvii (1984), pp. 202–9. 8  Suetonius, Augustus, 67; Plutarch, 207B; Galen, v.17. References to Galen are to the old and still the fullest edition, Claudii Galeni opera omnia, ed. C G. Kühn, 20 vols. (Leipzig, 1821–3). The use of anecdotes here is not to be confused with attempts to use them as authentication of particular actions by individual Romans, a practice rightly criticized by Richard P. Sailer, “Anecdotes as Historical Evidence”, Greece and Rome, cxxvii (1980), pp. 169–84. He shows how stories did the rounds and were amended and then successively attributed to various actors.

376

Hopkins

attitudes to morality, just as they helped Romans themselves to sort out what was allowable, legitimate and praiseworthy. Individual stories do not tell us directly what was normal, but they do indicate which abnormalities met with overt disapproval, and for these purposes, I should stress, it does not matter so much whether these stories were true. It matters more that they were told and retold. One striking story dramatized the confrontation of cruel rich man and emperor. An extremely rich man, called Vedius Pollio, himself an ex-slave, but now risen to the status of a knight, had a fish-pond stocked with huge lampreys which he fattened on the flesh of slaves who offended him in any way. Once, when Augustus was dining with him, a young slave dropped a precious crystal bowl. His master immediately ordered him to be seized and thrown alive to the lampreys. The boy slipped from his captors’ grasp and threw himself at the emperor’s feet “to ask only that he be allowed to die in some other way, not as human bait”. Augustus was so shocked at Vedius’ cruelty that he pardoned the boy, and ordered that all Vedius’ crystal bowls be smashed there and then, and that the fish-pond be filled in. Again it does not matter here whether the story is true, or how much it has been embroidered in the telling. It matters that it was told by Romans to each other as a moral story overtly celebrating the evil corruption of non-hereditary riches and the benefits of a benevolent monarch reasserting antique values. At least, that is one interpretation, although ancient readers of the story, or listeners, probably had their own divergent interpretations and sympathies. Such stories circulate, I imagine, partly because their moral messages are problematic and ambiguous (for example, did all slave-owners think that the emperor was right?). However that may be, this moral story has a further twist. When Vedius died, he left most of his property to the emperor, and Augustus had his huge palazzo destroyed, converting the site into a public pleasure garden in honour of his mother.9 Out of vice came virtue and pleasure. Emperors set the tone, or reflected broadly held values. Galen, for example, also tells us that his mother used to bite her slave maids in fits of rage. He hated her for it. By contrast, he admired and tried to imitate his father’s self-control. Neither father nor son, he claimed, had ever whipped one of their slaves with their own hands. But many of his father’s friends used to hit their slaves in the face; Galen’s father calmly said that these friends deserved to get a dose of boils and seizures and die (rather as fervent early Christians preaching a 9  Dio, liv.23; elaborated by Seneca, On Anger, iii.40. See Ronald Syme, “Who was Vedius Pollio?”, in his Roman Papers, ii (Oxford, 1979), pp. 518 ff.; Paul Zanker, The Power of Images in the Age of Augustus (Ann Arbor, 1988), p. 137.

Novel Evidence For Roman Slavery

377

gospel of peace wished on all heretics and sinners the torments of Hell). After all, Galen’s father went on, they could have waited a while, and then inflicted as many blows as they wanted with cane or whip, after due consideration.10 It was not the physical punishment of slaves to which Galen objected, but impulsive, irrational behaviour. Galen recounts another story, which he tells us he often told.11 He was travelling from Rome to Athens with a friend, who was a simple and friendly fellow, good-natured and open-handed in sharing their daily expenses, but irascible and inclined to strike his slaves with his hands, sometimes even kick them, and quite often hit them with a whip or whatever piece of wood came to hand. When they landed at Corinth, they had sent most of the baggage and slaves on by ship and themselves took the shorter overland route with a carriage they had hired. On the way, Galen’s friend asked the slaves who were walking behind for a particular small bag; they replied that they had not brought it with them. “He was enraged, and because he did not have anything else to beat the young lads with, took his sword in its sheath, and hit each of them on the head, not with the flat of the sword, which would not have been serious [!], but with its cutting edge”. The friend hit each of the slaves twice, and the sword burst its scabbard, so inflicting serious head wounds. When he saw the huge streams of blood, the friend fled in the direction of Athens, because he did not want to be on the spot when one of the slaves died. Killing a slave under Roman law could count as murder unless death occurred in the course of reasonable punishment.12 But the excellent doctor Galen looked after the slaves, ensured their recuperation (that, doubtless, was one of the main points of the story), and got them safely to Athens. Galen’s friend was overcome with remorse. One day soon afterwards, he took Galen into a house, stripped off his clothes, handed Galen a whip, went down on his knees and asked to be flogged. The more Galen tried to laugh it off, the more persistent his friend became. Eventually Galen consented, but only if the friend would submit to certain conditions. Willingly the penitent friend agreed. So Galen made him promise to listen to what he had to say, and then lectured him on the virtues of self-control, and on the possibility of controlling slaves by methods other than the whip. This story would probably evoke different responses in different readers, ancient and modern. In my view, four points emerge: first, beating a slave on the head with the flat of a sword was considered 10  See Galen, v.41. 11  See ibid., v.17. 12   Digest, i.6; Gaius, Institutes, i.53; Theodosian Code, ix.12; W. W. Buckland, The Roman Lato of Slavery (Cambridge, 1970 edn.), pp. 36–8.

378

Hopkins

regrettable, but not serious; secondly, even remorse took the form of violence; thirdly, violence and self-control represented an axis of moral strain within polite Roman society—the controlling élite was expected to be self-controlled; and finally, telling stories constituted an instrument of social control, outlining to listeners and subjects the limits and penalties of transgression.13 I too want to tell a story, derived from the only biography of a slave to survive from the ancient world. This slave biography is a satirical fiction, the so-called Life of Aesop, written in its finished form in the first century AD in Roman Egypt.14 It exists in several slightly different versions and in several copies. Most of the surviving copies are in Greek, but a Latin version also survives. The work has no known author; it is a composite text, which was probably frequently revised, anonymously, in order to incorporate the various popular stories circulating in the Graeco-Roman world about masters and slaves. The multiple versions and surviving copies of the Life of Aesop (several surviving on papyri written in Greek in Roman Egypt from the second to the fifth century AD) indicate its forgotten popularity, and dramatically increase its utility for us; 13  I want to stress again that ancient interpretations of such stories were probably pluralistic and conflicting. Even the same person reacts to symbolic messages with a whole range of sometimes conflicting emotions. The story is told again and again, partly just because it strikes a transgressive nerve, and precisely because the outcome of a struggle in the imagination between right and wrong is not assured. For an exciting discussion of these issues, see Peter Stallybrass and Allon White, The Politics and Poetics of Transgression (Ithaca, N.Y., 1986). 14  For the best edition and translation of the Life of Aesop, see above, n. 2. The two principal manuscripts are denoted by the letters G and W; I mostly follow G, occasionally W, and sometimes amalgamate both (for the texts, see Aesopica, ed. Perry). Four papyri fragments of the Life found in Egypt and dating from the second to the fifth century ad are listed by Roger A. Pack, The Greek and Latin Literary Texts from Greco-Roman Egypt, 2nd edn. (Ann Arbor, 1967), nos. 2072–5. In the opinion of F. R. Andrados, “The Life of Aesop”, Quademi Urbinati di cultura classica, xxx (1979), pp. 93 ff., the Life reached its surviving form by the first century ad. This seems plausible, but I suspect it involves simply working back from the earliest known papyrus fragment. Several passages in the surviving text firmly suggest an Egyptian context for the tale (see below, n. 20), but then the mention of Roman money, denarii (Aesop, G 24, 27), the death scene at Delphi (G 125 ff.), and Aesop’s visit to the emperor at Babylon (G 101), often equated in apocalyptic literature with Rome, and to the legendary king of Egypt at Memphis, whom he also outwits (G 112 ff.) suggest multiple sites, origins and fantasies. There seems mercifully little written in modern times about the Life of Aesop, but see Ben E. Perry, Studies in the Text History of the Life and Fables of Aesop (Haverford, 1936); A. Wiechers, Aesop in Delphi (Meisenheim, 1961); Heinrich Zeitz, “Der Aesoproman und seine Geschichte”, Aegyptus, xvi (1936), pp. 225 ff.

Novel Evidence For Roman Slavery

379

what we have is not a single author’s idiosyncratic vision, but a collective, composite work incorporating many different stories told about slaves, and attached here to Aesop. But we can have little or no idea about the audience at whom it was aimed, or by whom it was read or heard. Its survival in Egypt and the occasional adaptation of the text to Egyptian conditions should not mislead us into thinking that it relates primarily to Egypt, any more than its being written first in Greek should make us think that it related primarily to Greek slavery. In my view, the Life of Aesop is a generic work, related generally, but not specifically, to slavery in the whole Roman world. To be sure, the Life makes jokes about academic pedantry and the respect due from children to their professors, which may indicate its origins or circulation among students.15 But its simple prose style and unaffected humour suggest that it had a broader appeal among social strata well below the literary and power élites. Serious historians of the ancient world have often undervalued fiction, if only, as I have said, because by convention history is concerned principally with the recovery of truth about the past.16 But for social history—for the history of culture, for the history of people’s understanding of their own society— fiction occupies a privileged position. We have here not the “true” history of a single exceptional slave, who would not be typical.17 Instead we have an 15  Perry, Studies, p. 1, thought the Life of Aesop to be “a naïve, popular and anonymous book, composed for the entertainment and edification of the common people rather than for educated men”. I think his assumptions about the sophisticated tastes of the educated in the Roman world and the reading capacity of the common people are questionable. Besides, some elements of the intended audience can perhaps be seen towards the end of the novella: Aesop, after he has been freed, becomes tutor at the court in Babylon, and the surprising mouthpiece of almost conventional wisdom. He advises his adopted son: “Above all revere the divine; honour the emperor; since you are human, remember your humanity; God brings the wicked to justice. Honour your professor equally with your parents” (Aesop, G and W 109). It seems easy to spot the academic origins of such advice. So also “When drunk, do not show off your learning by discussing technical points of literature. Someone sometime is bound to outwit you, and you’ll make a fool of yourself” (G 109). 16  Of course, ancient historians have combed ancient fictions for historical “facts”. See, for example, the encyclopaedic and to my mind misconceived efforts of Jean C. Dumont, Servus (Paris, 1987), pp. 309 ff., seeking to discover facts and figures about Roman slavery in the comedies of Plautus and Terence. Both do indeed touch often on the issues dealt with at greater length in the Life of Aesop, and so corroborate the Life’s relevance to Roman slavery. See also the more sensitive yet still positivist rendering of Apuleius’ novel by Fergus Millar, “The World of the Golden Ass”, Jl. Roman Studies, lxxi (1981), pp. 63–75. 17  I am thinking here, enviously, of the difficulties of interpretation posed for the history of slavery in the American South by slave autobiographies: for example, Frederick Douglass,

380

Hopkins

invented, generalized caricature of a slave, whose relations with his master, and his master’s wife, guests, friends, other slaves and fellow citizens, reflect the central tensions in the relations between masters and slaves. We see the conflicts of interest, the pressures towards unthinking and therefore unquestioning obedience, towards loyalty and sabotage, seduction and infidelity, hard work, violent punishment and shirking. This fictional biography runs along the raw nerves of slavery, the same raw nerves which Roman comedies repeatedly touched, but never had the time or interest thoroughly to explore. The hero of the story is the historical Aesop, a slave who, according to tradition, lived in the sixth century bc and who constructed the lively animal fables which are still popular (such as the race between the hare and the tortoise, or the greedy peasant who killed the goose which laid the golden eggs, just to see if its insides were also made of gold). None of these tales survives in its original form. The earliest surviving collection was written in Latin verse by a freed slave, Phaedrus, who claimed that animal fables were invented by slaves as a secret way of talking to each other without being understood by their masters—evidently not with uniform success, since Phaedrus was himself punished by Sejanus, the powerful prefect of the palace guard under the emperor Tiberius, because of the offence which his fables had caused in court. Perhaps the underlying cause was their pervasive subversiveness. Consider what the ass said to the old shepherd. An anxious old man was grazing an ass in a meadow. Suddenly, he was terrified by the noise of enemy soldiers approaching. He tried to persuade the ass to flee, so that they would not be captured. But the ass, unmoved, asked “Do you think that the victorious soldiers will load me with two packs?”. The old man said no. “Then”, said the ass, “what difference does it make to me, so long as I carry only one load?”. The moral of the fable was clear and explicit: “A change of emperor brings a change of master for the poor, but not a change in prosperity”.18 The fables of Aesop were not politically innocent, nor were they written in these versions exclusively or primarily for children. The real hero or anti-hero of our tale is an Aesop of the imagination: to be sure, a slave, but what a slave! Pot-bellied, weasel-armed, hunchbacked, a squalid, squinting, swarthy midget with crooked legs; a walking disaster, deaf, and what is worse, dumb (Aesop, G and W 1). The grotesque repulsiveness of this My Bondage and My Freedom (New York, 1855); those collected in G. Osofsky (ed.), Puttin’ On Ole Massa (New York, 1969); also of the systematizing attempts to overcome the perils of individual reminiscence in G. P. Rawick (ed.), The American Slave: A Composite Autobiography, 22 vols. (Westport, 1972–81), based on faded memories of slavery collected in the 1930s. 18  Phaedrus, i.15; for the alleged origins of animal fables, see ibid., iii.pr.33–44.

Novel Evidence For Roman Slavery

381

misshapen body, “a turnip with teeth” (G 24) is consciously deceptive; it hides virtuous generosity, as well as a supremely astute and cunning mind, which is not above teaching his master a sharp lesson. Much of the story turns on the contrast between superficial appearance and unseen capacity, and on the social inversions in which the despised slave repeatedly humbles, even humiliates, his master. His master, Xanthus, is a philosopher, trapped by his status, his academic ambitions and his social pretensions into being both pedantic and reasonable.19 The slave is both simpler and cleverer. Aesop represents all that a master might despise and fear in a slave. Physically he is a caricature of the alien other, and like all slaves an irremediable outsider. And yet the master, in order to exploit his slave to the hilt, has to ask Aesop inside, where he can see all that happens in the house, and judge all that he sees. Ideally it would have been more comfortable if the slave had been, or could have been, treated as a machine; just as in Roman tracts on estate management the slave was defined as a talking tool (instrumentum vocale), though there was more hope than realism in this persuasive and derogatory definition.20 In reality, as the master fully realizes, the slave does have intelligence, and in order to exploit his domestic slaves thoroughly, the master has to trust the slaves to use that intelligence. He also has to leave the slaves occasionally to act alone, and to be left alone with the master’s wife. Hence anxiety, even fear, oppression, cruelty and, perhaps, half-unconscious jealousy. The master can beat the slave, cheat him of his promised freedom, but only at the cost of showing up his own moral inferiority. And in the end, almost inevitably, since this is a comic satire, the master who teaches his pupils not to trust women unwittingly encourages his own cuckoldry, and is then saved from even deeper, public humiliation in front of his friends, by the personal humiliation (which he shares with the reader) of privately depending on the slave Aesop’s help. Small wonder then that the story ends with Aesop’s liberation, and murder by a collectivity of free citizens. Aesop is simply too resourceful and too troubling to be left alive. The irony is that once freed Aesop no longer 19  The story, inevitably, had many resonances for the ancient listener and reader, which now deserve comment. For example, Xanthus was a common slave name; Aesop, the ugly slave, was also presumably an inversion of the ideal, beautiful slave (servus formosus), on whose acquisition rich Roman slave-owners sometimes spent a small fortune. 20  See Varro, On Agriculture, i.17; presumably, it was convenient for owners of large slavegangs to think of slaves as non-human. In Roman law, the slave was repeatedly defined as a thing (res), though of course the slave’s utility to his/her owner depended largely on not being a thing. See Digest, 1.6.2; Gaius, Institutes, i.53, for qualifications; see also the classic Buckland, Roman Law of Slavery, in which part 1 is divided into eight chapters, called “The Slave as Res [Thing]” (chs. 2–3), and “The Slave as Man” (sic, chs. 4–9).

382

Hopkins

has the protection of his master. Freedom exposes the ex-slave to fresh dangers.21 Slavery here is not just a tool for telling a story, but a category for thinking with, a way of understanding the tensions of Roman society. The appeal of the story for Romans depended, I think, on its nightmarish understanding of all that might go wrong in a master’s control over his slaves. But let us get down to the story. The Life of Aesop begins, as Greek novels often did, with the religious. Aesop shares his simple food and shows the way to a lost priestess of Isis, and in response to her prayer, Isis appears to Aesop while he is asleep, exhausted from his work in the fields, and gives him back the gift of speech, and the capacity to weave tales in Greek.22 Aesop immediately uses his new-found tongue to rebuke the overseer, himself a slave, for his cruelty, and unsettles him to such an extent that he rushes to their master in the town and reports Aesop’s recovery of speech, not as a miracle, but as an ill omen. The overseer receives permission, because Aesop is so ugly, “like a dogfaced baboon”, to give him away, or if that fails, to flay him alive (G 11). Masters, in real life in the Roman empire as well as in fiction, had effective power of life and death over slaves. Physical punishment, the fact of it and the threat of it, threads through the whole story, like the shouts of small children at play; even silence is ominous.23 21  The idea that freedom would be problematic for some slaves is touched on by Epictetus (Discourses, iv.1.35: the freed slave has to earn a living, and has no one who automatically feeds him), and in Phaedrus’ version of Aesop’s tales (Phaedrus, Fables, appendix, 20). Here an old slave complains that he has earned his freedom. His hair has turned grey, he gets beaten too much, and is given too little food. When the master dines at home, he stands by all night. When his master is invited out, the slave waits outside the host’s house in the street till dawn. He decides to run away. The wise slave reminds him that then his life would be much worse. To be sure, the issue here may be an amalgamation of the risks of freedom and of running away. But it is interesting that even a modest slave attendant was thought to have some chance of being freed. 22  Here is one hint among several that the Life of Aesop in its surviving form was adapted to an Egyptian context, where Isis (but not Greek) was thoroughly at home. Later on (G 65), the local district magistrate is called sirategos, a term used in this sense as far as I know only in Egypt under Greek and Roman rule. For the texts of Greek novels in English, see conveniently Collected Ancient Greek Novels, ed. Brian P. Reardon (Berkeley, 1989); for recent bibliography and discussion, see Shadi Bartsch, Decoding the Ancient Novel (Princeton, 1989); see also the imaginative Arthur Heisermann, The Novel before the Novel (Chicago, 1977). 23  Slavery involved being whipped, either in actuality or in fearful anxiety. That is the image which emerges repeatedly from ancient authors. The working conditions for slaves in mines and factories were particularly harsh. Apuleius, again in a novel, (written in Latin in the second century ad) about a man turned into an ass, which is also a symbol for a

Novel Evidence For Roman Slavery

383

As often happens in short stories, a slave-trader is passing by. The overseer sees his chance of a tidy profit, and has Aesop summoned from the fields. Aesop complains bitterly about the hardships of serving and obeying another slave, instead of his true master—this recalls the fiction of the good king, which resurfaces in the tales of Robin Hood: if only the good King Richard knew, he would stop the misdeeds of the wicked sheriff of Nottingham. Aesop gives a brief picture of what he is required to do, from laying the table, to feeding the livestock, drawing the water and heating the bath, as well as working in the fields. Whatever Roman lawyers said, this is a salutary warning against dividing slaves too readily into separate categories, agricultural and domestic.24 Many slaves performed a whole range of jobs and services. When the slave-trader, who is a specialist in fancy boys, sees “what a piece of human garbage” (G 14) Aesop appears to be, he at first refuses to buy. But Aesop appeals to Isis and to the trader’s sense of profit, and is soon proving his worth to his new master by his ingenuity, much to the indignation of the other slaves; one of them says “That fellow deserves to be crucified” (G 19). The symbol of the cross and the terrifying threat of crucifixion overshadowed the lives of slaves for centuries, before Christians converted it into a sign of religious devotion and eternal salvation.25 At last the slave-trader is left with only three slaves to sell, Aesop and two quite good-looking, skilled slaves, one a musician, the other a teacher. He stands them on the slave block as a trio. The philosopher’s wife passing by in her litter hears the herald advertising the sale. She goes home and persuades slave, describes the physical conditions of slaves in a bakehouse with a surprising sympathy: “Their skins were seamed all over with the marks of old floggings, as you could easily see through the holes in their ragged shirts, which shaded rather than covered their scarred backs; some wore only loincloths. They had letters tattooed on their foreheads, and their heads were half-shaved, and they had irons on their legs. Their complexions were frighteningly yellow. Their eyelids were caked with the smoke of their baking-ovens, their eyes so bleary and inflamed they could hardly see out of them. Like athletes in the arena, they were powdered, not with dust but with dirty flour”: Apuleius, Golden Ass, ix.12. 24   Digest, xxxiii.7. Dividing slaves into categories, and into hierarchical sets, commanded by slaves and by freed slaves, was a device for controlling slaves. 25  Based on an emendation of the text in G to “staurou”. The image of the cross, a symbol of punishment for slaves and bandits, reminds us of early Christianity’s appeal to the oppressed, and its promotion of a heaven which would be a world upside down: the poor would be privileged and the rich would be refused entrance. Needless to say, this radical vision receded as Christianity triumphed and forged alliances with the state. For other ancient restricted Utopian visions, see Diodorus, xxxiv.2; Athenaeus, vi.265 ff.; Vogt, Sklaverei, pp. 26–8.

384

Hopkins

her husband to buy a nice-looking male slave. He goes to the slave market, is too mean to buy one of the expensive slaves and buys Aesop instead. When he gets home, he tells Aesop to wait outside, and goes inside the house to tell his wife the good news. She immediately thanks Aphrodite, the goddess of love, for realizing her dreams, and starts fantasizing about the new slave’s charms. The female slaves of the household all start squabbling, and even come to blows, each imagining to herself that the new slave will be her husband.26 “The master bought him for me”, says one; “No, he’s mine”, says another, “because I saw him in my dream”. “Where is this idyll?”, asks the mistress. “Outside”, replies the philosopher Xanthus, “waiting to be asked inside” (G 20–30). And so the stage is set for master and mistress, and the rest of the household, to welcome inside the house their own undoing. Comedy takes on the trappings of tragedy. Once inside the philosopher’s house, the new slave Aesop succeeds in turning everything upside down.27 Unlike a real slave, Aesop speaks his mind, and so uncovers what is in the mind of others. The comedy lies in the shocking recognition of the truth behind the pretensions; the comedy’s cruelty lies in the revelation of what everyone would prefer to keep hidden. The husband loves his wife, but the wife, as so often in Greek thought, is portrayed as a “sex-crazed slut”, on the lascivious look-out for a “young, handsome, athletic, good-looking blond slave”, desired, as Aesop says openly to the wife, in order to shame her: “so that this fine slave can escort you to the baths, and there take your clothes from you, and when you come out of the bath, the same good-looking slave will put your wrap around you, and will kneel at your feet and put your sandals on, and then he’ll play games with you and look into your eyes as though you were one of his slave playmates who he had just taken a fancy to, and you’d smile back, and try to look young and feel excited, and [then later] you’d call him into 26  The wife’s litter, presumably carried by male slaves, and the quarrelling female slaves suggest a considerable slave establishment in the philosopher’s household. That was what the story demanded; it is difficult to know how typical that was for philosophy professors in the Roman provinces. I suspect it depended on how much teaching philosophy was the primary means of support and how much the philosopher could rely on inherited wealth. Assistants to a famous professor of rhetoric at Antioch in the fourth century ad (Libanius) had only 2–3 slaves each, or not even that: see Libanius, Opera, ed. R. Foerster, 12 vols. (Leipzig, 1903–27), iii, p. 129 (Speech, xxxi.11). 27  The polarity upright world/world upside down is to be sure only one of several polarities played against each other in the novel: slave/free, male/female, insider/outsider, educated philosopher/uneducated peasant. The reader can shuttle his identification from one to another as speedily as a modern filmgoer. See John Henderson, “Satire Writes Woman: Gendersong”, Proc. Cambridge Philol. Soc., xxxv (1989), pp. 50 ff., for a speedy introduction to multiple meanings.

Novel Evidence For Roman Slavery

385

your bedroom to have your feet rubbed and in a fit of excitement, you would pull him towards you, and kiss him, and do with him what suits your shameful impudence” (G 32). The baths, cleanliness, heat and lust were a heady mixture; and the close association between powerful female mistresses and their male slave attendants in public and in private stimulated the anxieties of husbands, the malicious gossip of envious observers and of later Christian moralizers.28 And besides, Aesop himself, who begins by severely rebuking his new mistress for immorality, ends up by seducing her, in revenge for his master’s ingratitude and broken promises. Admittedly it is a dangerous game to have sex with a tempestuous mistress, but Aesop’s revenge is all the sweeter, because when the lovers quarrel, as lovers inevitably do, Aesop inventively appeals to his master for arbitration and so gets him unwittingly to foster his own wife’s further infidelity (W 74–6). Some of the original manuscript has been lost in antiquity, so that we cannot follow the story through all its ramifications. But at one stage, clearly, both master and mistress have been completely infuriated by Aesop’s wilful misunderstanding of their instructions. I imagine it was a common enough problem with slaves, as later with servants.29 In the modern industrial world, this sense of impatient outrage is perhaps best parallelled by the consumer’s frustration or fury at not being able to understand or follow the instructions which accompany a self-assembly kit for a piece of household equipment. I suggest the comparison because I want to stress that the conflict between Aesop and his owner is structural and symptomatic, not personal and idiosyncratic. To be sure, the respected professor is constrained by his philosophy to be, as I have said, both reasonable and pedantic. He is the Goliath of the story, but more a figure of fun, at least outside the academic profession, where the obsession with formal definitions and verbal precision has rarely caught on. Almost against his will, the philosopher is trapped by his profession into being stolidly reasonable, into not punishing his slave without good reason. Driven to despair, he tries to control his slave by giving very precise instructions: “do nothing more or less than you are told”, under threat of a beating (G 38). But the slave, as of course the plot demands, determines to teach his master a lesson. He says to himself: 28  For example, Clement of Alexandria, The Educator (Paidagogos), iii.5, launched a viciously colourful attack on the loose morals of women who stripped naked in front of male slaves in the baths. 29  For the corrupting dependency of masters on their slaves, see Seneca, Moral Letters, 47, who argues that human nature transcends slave or free status. Did awareness of his view help slaves tolerate slavery better?

386

Hopkins

“Masters who are overdemanding with regard to the services they want have only themselves to blame for what goes wrong” (G 38). What is so remarkable surely about this story, designed I must stress to be read in a slave society (and perhaps even, as often happened, designed to be read aloud by a slave reader to his listening master and family), is that we are asked and expected in a slave society to side with the slave against the master.30 If the midget Aesop is David to Xanthus’ Goliath, it is as though the story was destined to be told not among Jews, but among Philistines. Humour camouflages, but does not fully hide, the lines of battle in the unceasing guerilla warfare between master and slave. The master says: “Take the oil flask in your hands, and the towels, and let’s go to the baths”. So Aesop picks them up, and without putting any oil in the flask, follows his master. The comedy lies in the predictability of conflict, in our knowledge of whom the fictional victor will be, in the unrealistic patience of the master, who is by inversion the fall-guy. “Xanthus gets undressed, hands his clothes to Aesop, and says: ‘Give me the oil flask’. Aesop gave it to him. Xanthus took the oil flask, turned it upside down, found that there was nothing in it, and said: ‘Aesop, where is the oil?’. Aesop said: ‘At home’. Xanthus said: ‘Why?’. And Aesop said: ‘Because you said to me “Bring the oil flask and the towels”, but you didn’t say anything about oil. You had told me not to do anything more than I was told, or else, if I broke that rule, I was likely to be beaten’. And then he kept his peace” (G 38).31 End of Round 1. 30  Various Roman slave-owners are known to have had their slaves read to them, during meals, in the bath, or while they went to sleep; the mere fact that some slaves were explicitly called “reader” (lector, anagnostes) strongly suggests that such functions were common in grand households. See J. Marquardt, Das Privatleben der Römer (Leipzig, 1879), p. 148, for collected references. The implicit picture of a master, mistress, children and slaves all hearing the Life of Aesop read aloud reinforces my claim that hearers interpreted what they heard diversely. 31  Plutarch, a Greek commentator on Roman life, writing mostly towards the end of the first century ad, noted that Roman slave-owners were trapped by the very strict discipline which they themselves imposed on their slaves. He tells a story nowhere else recorded, about a Roman consul of 61 bc, M. Pupius Piso. He invited another magistrates Clodius, to dinner and prepared a sumptuous feast. The other guests came, but not Clodius. So he sent the same slave who had carried the original invitation to see if Clodius was coming. After another long wait, Pupius asked the slave if he had really invited Clodius. “Yes”, said the slave. Master: “Why then has he not arrived?”. Slave: “Because he did not accept the invitation”. Master: “Why did you not tell me immediately?”. Slave: “Because you did not ask me that” (Plutarch, On Talkativeness, 18 = 511 DE). The story has the same point, or so it seems, as this section of the Life of Aesop: that is reassuring. What the Life of Aesop tells us is not so much new, as unparalleled in its richness and in its perspective.

Novel Evidence For Roman Slavery

387

In the next round, Xanthus tries to be even more careful and precise. He meets some friends at the bath and plans to invite them to dinner, so he sends Aesop back home: “Aesop, go on back home … and cook lentil [sic]32 for us. Put it in the pot, put some water in with it, set it on the cooking range, put wood under [the pot], light it; if it goes out, blow on it. Now, do exactly as I say”. Aesop says: “I will”. And he goes back to the house, into the kitchen, tosses one lentil into the cooking pot, and cooks it. Meanwhile back at the baths, Xanthus (as befits a stereotypical philosopher) is making a parade of his simplicity, and asks his friends back home to eat a modest snack. They accept. Xanthus brings his friends to the house and says to Aesop: “Give us something to drink for men straight from the bath”. So Aesop runs to the baths with a pitcher and fills it up from the outlet, handing it with a flourish to Xanthus, who is overcome by the stench of it and asks “Phew, what’s this?”. And Aesop says “Something to drink straight from the bath”. (G and W 34–40; translated by Daly). End of Round 2. Eventually, when Xanthus and his friends have been drinking wine for a time, Xanthus asks Aesop whether the “lentil” is cooked. Aesop replies that it is ready. Xanthus, just to make sure, asks to try it. So Aesop brings him the one lentil on a spoon. Xanthus eats the lentil, and says: “That’s fine; it’s well cooked. Bring in the food and serve it”. So Aesop gets a large dish, pours in the boiling water, and says “Bon appetit; it’s well cooked”. When his master expostulates, Aesop explains that he has eaten the one cooked lentil, and that he had originally ordered “lentil”, not “lentils”; “the one is singular, the other plural” (G 41). So the pedantry is volleyed back, the master is mastered. The slave relents and says: “You shouldn’t have laid down the law so literally, or I would have served you as best I could. But don’t mind, master … you’ve learnt not to make similar errors in lectures”. And since Xanthus can find no proper reason for flogging Aesop, he says nothing (G 43). The hard-nosed realist may well object that all this is mere comic fantasy. In “real life”, no slave had such licence or such a self-controlled, philosophical master. But real life consists of fantasies and feelings as well as of external realities. We think, we laugh and dream, as well as behave; our thoughts and feelings shape our behaviour, and mould its meanings. And each of us thinks, both idiosyncratically and by orchestrating a standard repertoire of culturally specific stereotypes. The slave of Aesop’s story, as a stereotypical outsider, invited inside, was good to think with; his story served as a social mirror which helped insiders to define themselves, and their relations with each other.

32  The Greek word for “lentil” used here, phakos, like the English “sheep”, has unusually the same form for singular and plural.

388

Hopkins

Stories about au pairs in modern British families serve similar functions; so perhaps do stories about summer camps or college in the U.S.A., where children and post-adolescents learn about being insiders, by being exported en masse, of course “for their own good”, into the outside. What matters here is not just the master’s experience of the slave, or the slave’s experience of slavery, or the child’s experience of summer camp, or the parents’ ideas about what a college education achieves. The experience and ideas about the experience are orchestrated and enhanced partly by the stories which we tell about our experiences with outsiders. And what is fascinating is that these stories repeatedly reveal the dependency of the insiders on outsiders for the delineation of their identity. So too Roman masters needed slaves in order to be masters, and they needed stories about slaves in order to work through and recreate some of the problems which their own social superiority inevitably caused.33 In these stories, the anti-hero, the ugly hunchback dwarf slave, rather like Rumpelstiltskin, is a representative of the other, who is allowed to express what cannot easily be attributed to a free insider. He is socially despised, but symbolically central. He is created from a repertoire of social images, whose appeal lies in the tension they generate between underpinning and undermining the dominant culture.34 Aesop is a projection of repressed emotions; he is potent, cunning and vengeful. He is what we desire, envy, and fear. The appropriate question is not “Did slaves really talk like that to their masters?”, but rather “Did many slave-owners in some way fear that they might?”. The passing recognition of that common fear, set in a disconcerting, but reassuringly untrue story, bound masters together as masters. The story allowed repressed fears and erotic attributions to rise briefly to the surface, gave fantasy a short airing, and then blocked off the imaginary transgressions (for example, the slave’s wit, wisdom and seductive virility) by mocking them away as comic fictions. To be sure this is only one interpretation of a story. I cannot be sure that it is right. But interpretative history flourishes on precisely those

33  The stories were not merely reflections of problems, or part of their easy resolution, but became themselves part of the process by which the multiplicity of meanings which slavery generated were negotiated by masters and slaves—just as stories about footballers and football hooligans are part of the experience of football. 34  The ideas of interdependence between the despised and the despiser, and the common repertoire of images which they may share, are fruitfully explored in Stallybrass and White, Politics and Poetics of Transgression.

Novel Evidence For Roman Slavery

389

ambiguities which the substantivist (conservative) historian of objective truth finds least comfortable, or most objectionable.35 The hostility between master and slave runs like a sore throughout the story.36 One pungent scene illustrates this. The master after a hot bath and a few drinks, feels the “call of nature” and goes to the privy, with Aesop in attendance holding a towel and a pail of water; they discuss why men after defecating often inspect their shit (G 67). This episode reminds us of what our own cultural prudery (not about looking, but perhaps about telling) tempts us to forget. Roman domestic slaves, especially in small “bourgeois” households, were body slaves. To their repeated degradation, they knew their master inside out. They knew him in all his vulnerability, with his trousers down—rather as a wife does, but without the tender feelings or the commitment of a lifetime. The master understandably hated the slaves who knew his weaknesses, and projected the anxiety which that hatred aroused on to the slaves themselves. At least, that is my interpretation of why the master repeatedly in the Life of Aesop calls the slave “refuse” (katharma). It reflects, by denial, the master’s fear that the slave is really human. Indeed each of the other common ways of addressing slaves—“boy”, “whipling”, “runaway”—also reflected by denial a different aspect of the master’s anxieties.37 Calling him “boy”, for example, both expressed and repressed the fear that the slave was a responsible adult, just like the master; “whipling” expressed the idea that the master was fully entitled to beat the slave as he wished, but repressed the idea that the beaten slave might take revenge if or unless beaten; “runaway” tempted, teased and dared

35  In ancient history, at least, the lines of this particular intellectual battle are not clearly drawn. The protagonists, conservative positivists and interpretative pluralists, prefer, I suspect, to ignore or to undervalue each other’s work, rather than to discuss each other’s methodological advantages and shortcomings. There may be something to be said for divergent practice over naïve epistemology, as in Veyne, Writing History, or Geza Alföldy, Die römische Gesellschaft (Stuttgart, 1986), p. 133: “History as created by scholars is not identical with the actual past”(!). 36  I think the polarity master/slave is dominant in the text, but as I have said (see above, n. 25) other polarities, male/female, educated/uneducated, transect and enrich the comedy of the conflict between master and slaves. 37  Slaves were variously called boy (puer, pais), whipling (verbero) which was particularly common in Latin comedies, and runaway (drapeta). Such derogatory forms of address were matched by the common Greek word for a slave, andropedon, which is neuter rather than masculine or feminine in form, and by the practice in Greece and Rome of denying that a slave could have a legal father or a legal spouse. Slaves were effectively rightless, powerless and neutered.

390

Hopkins

the oppressed slave to do just that, and so leave the master with one slave less to exploit. On both sides, hatred was based in fear. The slaves knew of the discrepancy between private weakness and public show. The comedy in many of the scenes in the Life of Aesop turns on Aesop’s audacious willingness to show his master up in public. The master wants an intelligent slave, but not that intelligent; he therefore is repeatedly torn between a desire to exploit his slave’s intelligence and the desire to flog him into senseless submission. The slave reciprocally is torn between the desire to serve loyally by identification with the master, the hope of freedom, dumb insolence and revenge. The novel explores the whole range of these options; and makes its explorations of conflict between master and slave, seen from the slave’s point of view, tolerable by cloaking them in comedy. Only modern historians need to take them seriously. In the remainder of the story, Aesop works through all the permutations of relationships inside and outside the household. He takes revenge on his mistress by driving a wedge between her and her husband. He demonstrates that the bitch (dog) loves her master more than his wife does, and when the wife, understandably indignant, withdraws in dudgeon to her parents’ house, Aesop devises a simple stratagem for getting her back home (G 44–50). Here, as elsewhere, the novella is pointedly misogynist. Next Aesop, by outwitting the students, as well as the professor in front of his students, to say nothing of his scholarly friends, illustrates, at least to the reader, the superiority of native intelligence over learned folly (G 48, 51–4, W 50a, 77b). The same point, the superiority of simplicity over contrived sophistication, is driven home in the next scene. The master makes a desperate attempt with his wife’s help to find a good reason for thrashing Aesop. The term used for thrashing here in the Greek literally means “teach him a lesson”; but of course, it is the master and the reader who are taught the lesson. Aesop is challenged to find a dinner guest who really can mind his own business. He duly succeeds in finding, not a sophisticated, self-controlled philosopher, but a taciturn peasant, who simply tolerates all the provocations put in his way by his host and hostess, who humiliate only themselves by their offensively unconventional behaviour. In the end, the professor, Xanthus, is inevitably forced to admit defeat to his own slave. But the topsy-turvy world is put partly to rights when Aesop in return reaffirms his loyalty to his master (G 64). The real lesson is that good service from a high-quality slave cannot be based on explicit instructions, or fear of punishment. It has to be based on trust. In this new phase of loyal service, Aesop tries to save his master from a drunken folly. He is rebuffed. Even so, with typical cunning resourcefulness, he

Novel Evidence For Roman Slavery

391

saves his master from the social death of bankruptcy and from public humiliation in front of his fellow citizens.38 In return, Aesop asks for his freedom; his master ungratefully refuses. Aesop swears revenge, and sets about seducing his mistress.39 One morning, Aesop stands outside the house holding his cock in his hand “like a shepherd” to stimulate it, and to give himself comfort from the cold. When Xanthus’ wife came outside and suddenly sees how long and thick it is, her lust is aroused, and she says to Aesop: “If you do what I want, I’ll give you more pleasure than your master gets”. Aesop demurs, like a shy young maiden feigning reluctance, and reminds her of the risks which he runs, if he is found out. But she, in hot pursuit, says: “Have sex with me ten times, and I’ll give you a shirt”. He says, “Swear on it”. She is so excited that she swears the oath. Aesop takes her word. He wants vengeance on his master. He has sex with her nine times. And then he says, “Lady, I can’t do it any more”. She is burning with desire, and says: “If you don’t do it ten times, then you’ll get nothing from me”. So he tries a tenth time, but fails to satisfy her. He says, “Give me the shirt, or I’ll appeal to the master”. She says, “Do it once more and you can have the shirt”. He cannot, but her petulance hardens his resolve. So naturally, when the master comes home, Aesop goes up to him and says: “Master judge between me and my mistress”. And as is to be expected, he tells Xanthus the whole story, in the form of a fable, which Xanthus the philosopher fails to understand. So the master adjudges Aesop the winner, but, to make the peace, says with the wisdom of Solomon, that when they have been to the market-place, Aesop should try the tenth time again. So the wife is satisfied and promises once more to give Aesop the shirt (W 75–6). Eventually, to cut a longish story short, Aesop is flogged, saves his master from suicide, and finally tricks him into giving him freedom irrevocably in a public assembly. He then outwits the collective citizens, the governor (here called strategos, which is a good clue that these versions derive in part from Egypt) and various kings, but is finally forced to die by an unholy alliance between a jealous God (Apollo), local dignitaries and the free citizens. Aesop, the wise man, like Socrates, is put to death, because he is too free with his thinking. 38  The humiliation and dishonour of slaves in different slave cultures is a central argument of Orlando Patterson, Slavery and Social Death (Harvard, 1982), pp. 11 ff. 39  The following section also owes something to the Latin version of the Life of Aesop (Aesopica, ed. Perry, p. 127). The episode reminds us that we should not domesticate the Romans to be cosy, idealized mirror-images of our civilized selves. They told and presumably some of them enjoyed stories which are more vulgar than most modern histories of Rome. But should Roman history be as nice as we might like to be ourselves?

392

Hopkins

The murder of the scapegoat is not just a convenient end to the story. It also reflects, I think, the endemic hostility to the clever slave in Roman society. From the masters’ point of view, Aesop is too clever by half and finally, like the villain in a western, gets what he richly deserves; the hostility of the god Apollo helps legitimate his execution. This novel about Aesop is a rebellion of the mind, and none the less potent for that. Like much comedy, it works by the inversion of normality, and by the suspension of probability. But the rebellion is soon put down. And in the end, the anti-hero, the dwarf slave, the fall-guy, is forced to his death and jumps over a cliff. Society has exacted its revenge, and normality has been restored. Indeed the plot may have been only a ruse, just as authoritarian governments, from ancient Sparta to modern China, sometimes encourage displays of initiative, and so foster the hope and the illusion of freedom among the oppressed, all the better subsequently to reinforce repression. The very tactic of inversion implies its opposite. Disorder implies order. I shall finish, as I began, by commenting very briefly on method. Inevitably, when I write a piece like this, I wonder what type of history it is, how I can show that I am right, how anyone else can show that I am wrong. I suppose the first point to be made is that the categories right/wrong are not so much irrelevant here, as broadly unhelpful when applied to this type of history. So what type of history is it? In this article I have tried to create an evocative history out of fiction, not by describing a single reality, but by delineating one set of realities out of the many probably available and used by different Roman slaves and masters. I have tried by the conservative method of reading a text and by empathetic imagination to reconstruct how a master and a slave, and by implication how many masters and many slaves, struggled to negotiate their competing interests.40 The Life of Aesop taught me to perceive the relationships between masters and slaves not just as a hierarchy governed by the external structures of law and economy. To be sure, these external structures visibly impinged in the stories I have told and set broad limits to what a master and slave typically did; but they left each master and each slave some degree of freedom to find their own paths within the broad avenues of convention. Each master and slave had some freedom of manoeuvre to act in accord with his or her 40  Anthony Giddens, The Constitution of Society: Outline of the Theory of Structuration (Cambridge, 1984), has created the concept structuration, awkwardly named, to capture the idea that actors are not simply the passive victims of external pressures, but themselves repeatedly reinterpret conventional social values and mores, and so by their actions reproduce the social order. Action here includes both behaviour and thoughts.

Novel Evidence For Roman Slavery

393

capacities, opportunities and interests. By implication, I have argued that this type of social history, which deals with what we could call “the considered experience”, deserves more care and attention than it has received—at least when compared with the public time-ordered history of external structures and élite politics. By implication, I think I have also argued that empathy, rhetoric, psychological insight and story-telling are appropriate techniques for a historian trying to recapture, not all the evidence, but some of the past. Put less grandly, the Life of Aesop has taught me to understand aspects of Roman slavery more fully than I ever had, and it is that experience which in this article I have tried to share.

Resisting Slavery Keith Bradley In the reign of Caligula a slave named Androcles was set free at Rome in rather unusual circumstances. The story is told by A. Gellius (5.14). Having been sentenced to death by exposure to wild animals in the amphitheatre (a standard penalty for slaves guilty of capital offences), Androcles was one of a group of prisoners who appeared in the Circus Maximus on a day when the emperor himself happened to be in the audience. Expectations of a fine spectacle were high owing to the exceptional size and ferocity of one of the waiting lions, but instead of a bloody battle between man and beast, the crowd witnessed an altogether different sight: for the ferocious lion recognised Androcles as an old companion and to the amazement of all turned the slave’s terror to joy by refusing to attack him. By popular demand Androcles was delivered from his punishment, set free and given custody of the lion.1 When summoned to account for the animal’s extraordinary behaviour, Androcles informed the emperor that in Africa years before he had once removed a huge splinter from the lion’s paw, having unwittingly taken refuge in the lion’s lair while running away from his master, who was then serving as the province’s governor. The grateful lion had subsequently shared his quarters with Androcles and helped him to survive, but Androcles was eventually captured by Roman troops and restored to his owner (now back in Rome), who had him condemned to death for having run away. Only in the Circus did Androcles discover that the lion too had been captured and brought to Rome—a discovery that meant his life was saved. The liberation of Androcles is a marvellous tale, and perhaps no more than that. But the story goes back to a writer who claimed to have been an eyewitness of the event, Apion of Alexandria, and as told by Gellius it is full of authentic details that lend the whole episode a certain credibility. It is of particular interest, therefore, that in Gellius’ version (5.14.7) the story contains a very rare example of a Roman slave’s own explanation of his behaviour, for Androcles accounted for his original act of flight in this way: Source: Bradley, Keith, “Resisting Slavery,” in Keith Bradley, Slavery and Society at Rome, Cambridge: Cambridge University Press, 1994, 107–131. © Cambridge University Press, reproduced with permission. 1  Androcles: cf. Sen. Ben. 2.19,1; Aelian, de Natura Animalium 7.48. Standard penalty: Garnsey 1970: 129. © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_017

Resisting Slavery

395

When my owner was governor of Africa, I was driven to run away by the unjustified beatings I received from him every day (‘iniquis eius et cotidianis verberibus ad fugam sum coactus’). And to make my hiding places all the more safe fro6m him, the master of that whole region, I took refuge in isolated plains and deserts, intending if I should not find food somehow to kill myself … From time to time statements of servile motivation appear, or are implicit, in conventional Roman literary sources. Tacitus (Ann. 14.42.3), for example, attributes the murder of the city prefect L. Pedanius Secundus in ad 61 to either Pedanius’ reneging on a manumission agreement made with the slave who killed him—and presumably to outrage on the slave’s part—or to the slave’s inability to tolerate his master in a homosexual liaison. And Ammianus Marcellinus (28.1.49) tells of a slave named Sapaudulus who in ad 369 informed against his mistress and the lover she was concealing from prosecution because the slave’s wife had earlier been flogged: revenge was clearly at play in the act of informing. It was possible, therefore, for proslavery writers sometimes to acknowledge that slaves were human beings with human emotions, and to recognise that bruised feelings could lead to violence and danger to themselves. But there could never be any open encouragement or support of such reactions. Any injury by a slave to a slaveowner was a crime, given the moral standards that prevailed in Roman society, a transgression that as in the Pedanius Secundus affair was attributable to some moral weakness such as servile treachery; and as with Androcles, the offending slave had to be punished, and punished in an appropriate manner.2 The personal evidence of Androcles, however, suggests that what established society saw as crimes meriting punishment (in Androcles’ case, the ‘crime’ of flight) was perceived completely differently by the slave exposed to the grim realities of oppression. From his own point of view the flight of Androcles was not a crime at all but an act of self-preservation and survival that despite its attendant risks (which, paradoxically, included the risk of selfdestruction) was preferable to a life of continued servitude. It was an act, that is to say, of resistance to slavery and to the power of the slaveowner based on the slave’s standards of moral comportment, not on those of the master. What authors such as Tacitus and Ammianus report pejoratively as servile misbehaviour is in fact evidence of servile opposition to slavery, and it is to this subject of slave resistance at Rome at large that I now turn.

2  Acknowledge: cf. Cic. Fam. 11.28.3. Appropriate manner: cf. Galen, de Placitis Hippocratis et Piatonis 6.8.82.

396

Bradley

The likelihood that Roman slaves attempted from time to time to reduce the rigours of servitude or to extricate themselves permanently from their condition may be readily admitted in simple terms of human nature, especially in view of the already documented fact that prisoners-of-war in Roman antiquity often preferred to inflict death upon themselves than to submit to the horrors of capture. The evidence of revolt is decisive. Thus in 73 bc, to take the most celebrated instance, the gladiator Spartacus led a revolt of some seventy slaves from a gladiatorial training school in Capua and for almost two years roamed throughout Italy with an army formed from the tens of thousands of slaves who flocked to join him, defeating a series of Roman legionary forces in the process and at one point posing a grave threat to Rome the city. Spartacus was eventually defeated. But the insurrection he headed is enough in and of itself to demonstrate slaves’ willingness to take positive action against their enslavement once conditions were conducive, as also to account for slaveowners’ perpetual fears and suspicions of their slaves.3 Still, whether on the grand scale or at the less intensive level of a slave conspiracy organised in southern Italy in AD 24 by a former member of the praetorian guard, slave revolts were infrequent after Spartacus, and this has led some scholars to conclude consequently that there was little cause for slaves to offer resistance at all. The chief difficulty with this idea, however, is the false assumption that revolt was the only avenue of resistance available to slaves and that in its absence all was calm. In the New World slave revolts were particularly virulent in the Caribbean, but in Brazil and the United States, just as at Rome, they were relatively infrequent. Indeed, an insurrection on the scale of that led by Spartacus did not occur again in the history of slavery until the turn of the nineteenth century when the modern state of Haiti emerged from the slave movement headed in St Domingue by Toussaint L’Ouverture. Once the difficulties of organising revolt, the risks of betrayal and detection, and slaves’ fears of punishment and reprisal are brought to attention, this is comprehensible enough. But in spite of the low incidence of revolt in some regions, there was no lack of rebelliousness among New World slaves generally; rather, as historians of modern slavery have repeatedly shown, resistance to slavery was endemic but took forms less spectacular, and less obvious, than that of open revolt.4

3  See in general, Bradley 1989. 4  A D 24: Tac. Ann. 4.27, with ILS 961 and Alföldy 1969: 149–53. Some scholars: Alföldy 1988: 70; Dyson 1992: 38–40; see, however, Cartledge 1985, and cf. Brunt 1990: 112, on the lack of provincial uprisings against Roman rule.

Resisting Slavery

397

Those forms ranged from violent acts, such as suicide or murderous assaults on slaveowners (often provoked by excessively brutal treatment), to the far less extreme actions of lying, cheating and stealing, of pretending to be sick or working at a calculatedly slow pace, of resorting in fact to any form of petty sabotage thinkable in order to indicate that slaves would not cooperate with their masters on a day to day basis, that they would cause their owners constant annoyance and frustration, and that they would take for themselves whatever relief from oppression was possible. In between was the ubiquitous practice of running away, either to gain temporary respite from slavery or in the hope of escaping servitude for ever, which New World slaves sometimes achieved, especially in Jamaica, Brazil and Surinam, through the formation of independent rebel communities in geographically remote locations. For the most part resistance was non-revolutionary, in the sense that most slaves most of the time were not impelled by ideological imperatives to effect radical political and social change, but were usually attempting only to protest against, and to diminish, their personal sufferings and to take a measure of revenge against their owners. (The Haitian revolution, which was not without antecedents or subsequent influence, was exceptional in that political aims were involved from the outset.) Generally speaking, slave resistance concerned itself with individuals and small groups, not with global situations, and slaves knew very well what they were doing, for resistance was a term in their own vocabulary, serving as an antidote to the telling labels of ‘rascality’ and ‘roguery’—the sheer troublesomeness—with which their masters constantly characterised, or rather mischaracterised, their behaviour. From the servile perspective, therefore, the moral regime of the slaveowners did not apply. If for survival’s sake slaves had to lie and deceive, then lies and deception were morally defensible. ‘I always thought it right’, said Thomas Hedgebeth, a fugitive from North Carolina, ‘for a slave to take and eat as much as he wanted where he labored’. More eloquently, and emphatically, still, the runaway Harriet Jacobs declared in her autobiography that ‘the condition of a slave confuses all principles of morality, and, in fact, renders the practice of them impossible’.5 5  Forms: see variously Bauer and Bauer 1942; Stampp 1956: 86–140; Genovese 1972: 585–660; Blassingame 1979: 192–222; Price 1979; Craton 1982; Gaspar 1985; Craton 1986; Quierós Mattoso 1986: 125–49; Karasch 1987: 302–34; Fogel 1989: 155–62. Troublesomeness: see the primary evidence, for example, of Col. Landon Carter of Sabine Hall, Virginia in 1770, in Rose 1976: 259–61; of the attorney Walter Tullideph in Antigua in 1748, in Gaspar 1985: 190; and of an Alabama slavedriver writing to his master in 1847, in Miller 1990: 157. Own vocabulary: Gates 1987: 380, 382 (Harriet Jacobs), with Yellin 1987. Thomas Hedgebeth: Drew 1856: 278. Harriet Jacobs: Gates 1987: 385.

398

Bradley

The various forms of resistance behaviour other than revolt known from the history of modern slavery naturally invite the question of whether anything comparable is visible in the Roman record. Quite straightforwardly, the answer is yes. The situation of war captives apart, there are several indications that Roman slaves were at times driven to a point of desperation from which the only relief was self-destruction. In 195 bc when the elder Cato was on campaign in Spain, one of his slave attendants, a man named Paccius, is said by Plutarch (Cato 10.5) to have rashly and improperly purchased from a batch of public prisoners three slave boys for his own use, and then to have killed himself rather than face his master, a strict disciplinarian and a model of personal restraint. It seems that Paccius was so afraid of Cato and his powers of correction that forestalling certain punishment by the act of suicide was all that he could do. The same is implied by an inscription from Moguntiacum (CIL 13.7070) which commemorates the freedman and pecuarius M. Terentius Jucundus, who was killed by one of his slaves whose name is unfortunately not recorded. After killing his master, however, the slave killed himself by jumping into a river and drowning, apparently because he too was so conscious of his powerlessness as a slave to defend what he had done that suicide was his sole recourse. Seneca (Ep. 4.4) acknowledged that excessive fear could drive slaves wishing to escape dyspeptic owners and runaways who could not stand the prospect of recapture to take their own lives; but in a passage remarkable for its utter failure to comprehend the world of the slave, Seneca calls such reasons for suicide frivolous, comparing the case of a lover who hangs himself in front of a girlfriend’s door.6 Roman lawmakers regarded attempts at self-destruction on the part of slaves as commonplace to judge from information in the Digest. When a slave was sold the aedilician edict required the seller to declare whether the slave had ever tried to kill himself (Dig. 21.1.1.1), and like the requirements on disease and defect and place of origin, this suggests an objective reality of slave life that could be determined with some accuracy. According to Ulpian (Dig. 21.1.23.3), there were grounds for cancelling sale if the slave in question were disposed to suicide because of the ‘badness’ of character thus revealed. Moreover, when discussing the definition of a fugitive Vivianus (Dig. 21.1.17.4) said that a slave who threw himself from a height was not to be called a runaway if all he wished to do was to end his life, and Caelius (Dig. 21.1.17.6) said the same for a slave who threw himself into the Tiber or jumped from a bridge. The issue was 6  For further details, see Grisé 1982: 276–81; Bradley 1986a; van Hoof 1990: 171–2, 184–5. For comparable modern material, see especially Stampp 1956: 129; Genovese 1972: 639–41; Blassingame 1979: 7, 297; Conrad 1983: 59, 124; Karasch 1987: 316–20.

Resisting Slavery

399

clearly one of moral intent, and Paul (Dig, 21.1.43.4) could consequently define slave suicide as follows: ‘A slave acts to commit suicide when he seeks death out of wickedness or evil ways or because of some crime that he has committed, but not when he is able no longer to bear his bodily pain.’7 It appears, therefore, that slaves always understood that they had the option of ending their enslavement by ending their own lives, and perhaps over time a considerable number—how many it is impossible to say—did so. They will have been slaves to whom the ultimate human sacrifice was preferable to the burdens of servitude. Accounts of assault by Roman slaves against their owners are rarely found, but two detailed episodes of murderous violence against prominent men are known from the reigns of Nero and Trajan. The first victim was L. Pedanius Secundus, the city prefect who was killed in AD 61 for reasons already noted, while the second was Larcius Macedo, a senator of praetorian rank who died in AD 108 because, according to Pliny (Ep. 3.14), he had been unusually cruel in his treatment of slaves. Another victim was Hostius Quadra, who was murdered when Augustus was emperor, perhaps for reasons that included sexual abuse. Not even a slaveowner who was himself a former slave was immune from attack, as the case of M. Terentius Jucundus shows.8 Assault against slaveowners was probably much more common than these specific instances suggest. In the record of Pedanius Secundus’ murder, Tacitus (Ann. 14.42.2) alludes to an ancient Roman custom of torturing all the members of a murdered slaveowner’s familia, that is, all the slaves ‘under the same roof’. The rationale was that the slaves concerned had an obligation to come to the aid of their master at the time he was under assault and that they deserved to be executed had they not done so. The custom was the subject of legislation in the time of the dictator Sulla and on four other occasions from Augustus to Marcus Aurelius, knowledge of which depends chiefly on evidence from the Digest (29.5). Legislators regularly perceived a need, so it seems, to modify or reimplement the law because incidents of slaves attacking their masters also

7  Cf. also Dig. 14.2.2.5; 15.1.9.7; 47.2.36 Pr. 8  L. Pedanius Secundus: Tac. Ann. 14.42–5 (cf. Vita Aesopi 74, 79–80 for broken promises). Hostius Quadra: Sen. Naturales Quaestiones 1.16. See also Plin. Ep. 8.14 on Afranius Dexter, another senator who may have been killed by his slaves; and cf. Cic. Brut. 85; Artem. Onir. 5.25. For modern material, see especially Stampp 1956: 131; Genovese 1972: 616–17; Rose 1976: 193, 230; Conrad 1983: 251; Gaspar 1985: 193–5; McLaurin 1991.

400

Bradley

took place regularly. Assault at least was one of the obvious possibilities that occurred to Martial (Spectacula 7) when accounting for a slave’s grisly execution.9 Whether investigation of the whole household occurred only when an owner had been killed by a slave, or in cases of murder by non-slaves as well, is not entirely clear. Ulpian (Dig. 29.5.1 pr.) refers to assaults both by domestici, household slaves (slaveowners’ natural enemies in the view of Tertullian (Apol. 7.3)), and by extranei, outsiders, but not necessarily slaves. Tacitus, however, recording the passage of the senatus consultum Claudianum in AD 57 (Ann. 13.32.1), speaks unequivocally of slave murders. The jurists who are resumed in the Digest certainly anticipated some instances of assault by slaves, so the force of the simple word ‘often’ in Ulpian’s discussion (Dig. 29.5.1.27) of the meaning of the phrase ‘under the same roof’ should not be underestimated for what it implies about the frequency of slave attacks: Does this mean within the same walls, or, beyond that, within the same living room or bedroom or the same house or the same park or the whole country house? And Sextus says that it has often been decided by judges that whoever were in such a position that they could hear a cry are to be punished as having been under the same roof … Callistratus (Dig. 48.19.28.11) similarly remarked that slaves who had conspired against the safety of their masters were generally (‘plerumque’) burned alive in punishment, implying again that conspiracy of this sort was not uncommon. The victims the jurists spoke of, whether real or putative, included adults and children, husbands and wives, people killed in the city or in the country or travelling from one to the other. They might have died from having had their throats cut, being strangled or thrown down, or because they had been struck with rocks or clubs or stones.10 The emperor Hadrian once wrote in a rescript that a slave woman (ancilla) who had not cried out for help when her mistress was being attacked, even though the assailant had threatened to kill her if she raised the alarm, ought in any case to be put to death, ‘so that all other slaves may not think that when their masters are in danger each should look after himself’ (Dig. 29.5.1.28). Reasoning of this kind imposed impossible demands on slaves, both male and female, and must presumably have fostered divisions among them. The severe 9  Rationale: Dig. 29.5.19. Evidence: see Buckland 1908: 94–5 for the minor sources. See in general, Griffin 1976: 271–3. 10  Anticipated: Dig. 29.5.16; 29.5.21; 29.5.22; 29.5.25. Victims: Dig. 29.5.1.7; 29.5.1.14–17; 29.5.1.27; 29.5.1.30–1. Poisoning was regarded as a separate type of killing: Dig. 29.5.1.18; cf. 29.5.1.22.

Resisting Slavery

401

dictates of the law, moreover, were certainly met, for in the sequel to Pedanius Secundus’ murder some four hundred slaves were executed even though most could have had nothing to do with the killing. L. Trebius Germanus, moreover, the consul of AD 125, put to death, in his capacity as provincial governor, a slave boy who had been sleeping at his master’s feet but who had not cried out when the master was attacked and killed. Despite the unprivileged legal circumstances in which they found themselves, however, slaves were evidently prepared at times to respond to the violence inherent in slavery as an institution by using violence directly against their owners.11 Roman slaveowners were very sensitive to what they perceived as the frittering away of time or resources on the part of their slaves, as the legal definition of the truant slave (erro) at once makes clear: ‘one who … frequently indulges in aimless roving and, after wasting time on trivialities, returns home at a late hour’ (Dig. 21.1.17.4). Complaints from proslavery writers imply that owners had to contend with a constant undercurrent of day to day difficulties and harassment. Thus, in a notorious passage, Columella (1.7.6–7) cited a long list of the kinds of damage agricultural slaves could cause on the farm: hiring out the oxen without permission, improperly pasturing the cattle or ploughing, claiming to have sown more seed than was really the case, neglecting what had been sown, diminishing the yield of the harvest through theft or incompetence, and failing to keep proper records. He also recommended (12.3.7) that the uilica should make a daily check for malingerers who were trying to avoid work, as if the problem were widely familiar to his readers.12 Columella’s evidence is far from unique. Cato (Agr. 67) knew that oil was likely to disappear from the pressing-room or the storeroom. Cicero (in Verrem 3.119) found the image of the fraudulent vilicus selling off his master’s livestock and farm equipment realistic enough to exploit for rhetorical effect in an important trial. Seneca (Tranq. 8.8) numbered among the bugbears of slaveowning the fact that owners had to rely on people who were always breaking down in tears. The elder Pliny (Nat. 33.26) pontificated about pilfering slaves— hordes of them, there were—while Juvenal (11.191–2) thought it natural to associate domestic slaves with destruction and loss of property. Lucian observed (Merc. Cond. 23) that the newly enslaved slave, in contrast to the slave born into slavery, was not such a good investment because the memory of freedom encouraged poor performance from him at work; but there was also the opposite view, known to Tacitus (Ann 1.31), that it was homeborn slaves who were 11  Sequel: Tac. Ann. 14.44–5. L. Trebius Germanus: Dig. 29.5.14; see Syme 1988: 472, 557, 593. 12  Definition: the erro was not thought culpable to the same degree as the runaway proper: Dig. 49.16.4.4.

402

Bradley

the poor workers—and they were insolent at that. Tertullian (Apol 27.5) asserted that slaves generally took pleasure in offsetting their fears of their owners by deliberately causing them harm, while a bandit in Apuleius (Met. 4.8) reproached his less adventurous colleagues for being able to steal only on the miserable level of slaves with the words, ‘But you, honest robbers, with your petty, slavish pilferings [‘inter furta parva atque servilia’], are just junk-dealers creeping timidly through baths and old ladies’ apartments.’ The jurists took it as given that slaves stole in all manner of circumstances. A slave leased to work in a shop could be expected to steal from the person hiring him; a slave working on board ship or in an inn would rob the passengers or guests; a slave undertaker could be expected to steal from the corpse he was preparing for burial. The objects stolen were on the small scale: cash, silver, plate, furniture, clothing, perishables—household items in the main that could be disposed of quickly. The jurists, recalling Apuleius’ brigand, used the phrase ‘domestic thefts’ (‘domestica furta’) to refer to this sort of pilfering, misdeeds that were too trivial to justify prosecution, but which must have been deleterious to slaveowners in their overall effects.13 As for damage and destruction: breaking into storehouses and ransacking the contents, cutting down trees on a rural estate without permission, deliberately setting fire to property (the lavishly appointed Tusculan villa of M. Aemilius Scaurus, stepson of Sulla and praetor in 56 bc, for example)—the catalogue of slave crimes known to the jurists was endless. Slaves even caused physical damage to themselves, inflicting wounds for instance to make it seem that they had helped a slaveowner under assault in case they should be punished under the terms of the senatus consultum Silanianum. Literate slaves were able to falsify records and documents to the disadvantage of their owners. The late Republican jurist Alfenus Varus wrote (Dig. 11.3.16) of a slave accountant (dispensator) who was found, when his accounts were examined upon manumission, to have spent his master’s money on a disreputable woman (‘apud quandam mulierculam’); another ‘had not entered in the accounts certain moneys collected from [the owners’] tenants’ (Dig. 40.7.40 pr). Tampering with records could become habitual, a ‘consuetudo peccandi’ (Dig. 11.3.11.1), so

13  Jurists: Dig. 19.2.45 pr.-1; 47.5.1.5; 4.9.3.3; 14.3.5.8. Objects: Dig. 19.1.30 p1.; 47.2.57.5; 41.1.37.2; 30.48 pr.; 40.12.43; 46.3.19; 13.6.21.1; 40.4.22; 47.2.52.9; 40.7.40 pr. Cf. the list of items stolen by the slaves of Rio de Janeiro, subsumed under the heading of ‘petty thievery’: ‘all articles of clothing, a French church history, parrots, oxen, horses, other slaves, watches, purses, clocks, weapons, bottles, glasses, a door, objects from churches, and so on’ (Karasch 1987: 330). ‘Domestic thefts’: Dig. 48.19.11.1. See for further details, Bradley 1990: 141–2.

Resisting Slavery

403

that the slave who drew up his owner’s will could be expected to insert a clause guaranteeing, in due course, his own manumission (Dig. 48.10.22.9).14 As for shamming sickness, an episode from Xenophon of Ephesus’ story (Ephes. 5.7) illustrates what was possible, when Anthia is sold as a prostitute to a brothel-keeper in Tarentum. Put on display, she attracts a host of clients, but to safeguard her chastity she collapses, pretending to be epileptic, and subsequently concocts a tale to explain to her owner how she became ill. The episode is not all fancy: doctors in antiquity were not always able to tell whether their patients were sick or simply pretending to be sick when they examined them.15 Truancy, dilatoriness, lying, dissembling, stealing, causing damage, feigning sickness—at the strictly factual level these types of slave behaviour are all well in evidence. The sources report them in a prejudiced way, however, constantly speaking of slaves’ misdeeds, but seen from the servile point of view they are better understood as ways by which slaves could deliberately cause their owners vexation and thereby provide themselves with a measure of relief from their position of subordination and inferiority. To secure relief from adversity by physically separating oneself from its cause is a universal human strategy of survival. Thus before Constantine, Christians regularly evaded persecution by running away from their tormentors, taking refuge either in cities where they could literally lose themselves in a crowd, or else in remote rural areas, woods and deserts, for instance, where they could avoid detection. The tactic was so common that theologians contrived a spiritual justification for it based on Jesus’ withdrawal to Gethsemane before the Passion. For Christians, consequently, flight was a commendable act of self-preservation which allowed individuals to deliver themselves to the providence of God, not an act of weakness or cowardice. Likewise in Roman Egypt, peasants, and men of higher status too, frequently abandoned their homes when unable to satisfy the demands of the tax-collectors or to meet the liturgical obligations imposed upon them, preferring again to immerse themselves in large cities like Alexandria or to take up with itinerant bands of outlaws. It is not surprising, therefore, to find that Roman slaves ran away 14  Damage and destruction: Dig. 19.2.55 pr.; 47.7.7.5; 18.6.12; 9.2.27.9;19.2.30.4; 9.2.27.11; 19.2.11.4. M. Aemilius Scaurus: Plin. Nat. 36.115. Wounds: Dig. 29.5.1.37; cf. 15.1.9.7. See Drew 1856: 178 for evidence of an American slave who cut off the fingers of his left hand with an axe to prevent being sold away. Falsify records: cf. Dig. 9.2.23.4; ro.2.18 pr.; 30.67 pr., and see Drew 1856: 185 for an American slave who in 1854 wrote himself three passes to aid his escape from Mississippi to Canada. Dig. 11.3.16: quoted in chapter 2. 15  Quint. Inst. 2.17.39.

404

Bradley

from their owners. What is surprising is the amount of attention flight by slaves receives in the ancient record.16 From Roman Egypt documentary papyri give very vivid impressions of what running away involved—of where fugitives made for and the means their owners used to recover them. A letter (P. Turner 41) from the late third century shows a woman named Aurelia Sarapias seeking the help of a district governor (strategos) in tracking down her runaway slave Sarapion. The slave had once belonged to her father; both for that reason and because Sarapion had held a position of responsibility in her household, Sarapias had never expected trouble from him; but now, she said, under the influence of an unknown party he had abandoned his duties and, forsaking the material comfort of the household, had secretly run away. Worse yet, he had taken clothing and other items at his disposal with him. Sarapias seems from her letter to have had some idea of where her trusted slave was, but whether she ever recovered him is unknown. Another document (P. Oxy. 1643), of the same period, shows a certain Aurelius Sarapammon, a citizen of both Oxyrhynchus and Athens, appointing an agent to go to Alexandria to find a runaway for him, a thirty-five-year-old man; ‘and when you find him’, Sarapammon wrote sternly, ‘you are to deliver him up, having the same powers as I should have myself, if I were present, to … imprison him, chastise him, and to make an accusation before the proper authorities against those who harboured him, and demand satisfaction’. Similarly one Flavius Ammonas, an official on the staff of a fourth-century prefect of Egypt, wrote to a colleague authorising him to arrest his slave Magnus (P. Oxy. 1423): the man was a runaway and a thief and was understood to be living in Hermopolis.17 The fugitives in these examples were slaves who ran away by themselves, but the papyri also provide cases of men and women running away in small groups of three or four or five. They show too that even at the village level lists of runaways might be kept, and that the fugitives, as in Sarapammon’s case, could sometimes count on the support of third parties in their flight: in one example from Oxyrhynchus (P. Oxy. 1422) a person accused of harbouring a runaway is said to have disappeared himself. The papyri also attest the practice slaveowners adopted of advertising for their runaways in public places such 16  Christians: Nicholson 1989. Egypt: Lewis 1983: 163–5, 183–4. Flight by slaves: see in general, Bellen 1971. 17  Sarapias: cf. the similar experience of the Alabama slaveowner Sarah Gayle (Fox-Genovese 1988: 23–4), who owned a slave, Hampton, who had previously belonged to her father and who had known his mistress since her childhood; despite the long association, she found herself constantly complaining of his ‘insolence and contrary disposition’.

Resisting Slavery

405

as temples by posting notices of their slaves’ physical appearance and offering a reward for their return. One document (P. Oxy. 3616) refers to a runaway named Philippus, who was about fourteen years old, had pale skin and spoke badly, and who, when he ran away, was ‘wearing a thick (?) woollen tunic and a used shoulder belt’; the owner promised his finder a reward. Another example (P. Oxy. 3617) concerns an anonymous Egyptian slave, a thirty-two year old man who was a weaver by trade and who could speak no Greek; he was ‘tall, lean (?), smooth-shorn, with a slight (?) wound on the left side of his head, honey-complexioned, somewhat pale, with a scanty beard—(or rather) with no hair at all to his beard, smooth-skinned, narrow in the jaws, long-nosed’. The owner’s indignation at his loss seems to emerge in the final details: ‘he walks around as if he were somebody important, chattering in a shrill voice’.18 The hard record of papyri is not available of course for the Roman world at large, but the evidence of literature and the law indicates that the patterns of behaviour visible in Egypt were in fact visible everywhere. Advertising for runaways, for instance, emerges as a widespread convention from Ulpian (Dig. 11.4.18a), as also from the parody in Lucian’s dialogue, The Runaway Slaves, where the action proceeds from the story line that Hermes, Philosophy and various other deities are assisting three slaveowners in their quest for three fugitive slaves in Philippopolis in Thrace on the express orders of Zeus, who had received a complaint from Philosophy that runaways impersonating itinerant philosophers are bringing her into disrepute. To facilitate the search, Hermes at one juncture issues the following proclamation (Fugitivi 27)—the advertisement is oral rather than written—which is remarkably similar in form and content to the genuine notices from Egypt: If anyone has seen a Paphlagonian slave, one of those barbarians from Sinope, with a name of the kind that has ‘rich’ in it, sallow, close-cropped, wearing a long beard, with a wallet slung from his shoulder and a short cloak about him, quick-tempered, uneducated, harsh-voiced, and abusive, let him give information for the stipulated reward. As a literary device, the description of the runaway slave can be traced to the Hellenistic poets Moschus and Meleager; but in Latin literature it appears in Apuleius’ story of Psyche and Cupid (Met. 6.8), where the informant, directed 18  Groups: Biëzuńska-Małowist 1977: 141–2. Lists: Biëzuńska-Małowist 1977: 140. Practice: this was also the habit of New World slaveowners, who placed similarly detailed descriptions of runaways, and offered rewards for them, in newspapers; see for examples, Rose 1976: 57–8; Conrad 1983: 362–6.

406

Bradley

to proceed to the temple of Mercury close to the Circus at Rome, is promised as a reward for information on the runaway Psyche ‘seven delicious kisses plus one more, deeply sweetened by the touch of her caressing tongue’ from Venus.19 Runaway slaves, or slaves about to run away, are indeed ubiquitous in imaginative literature. In Petronius’ Satyricon (107.4), Encolpius at one point offers the plausible view that repentance might lead runaways back to their owners; Photis in Apuleius’ Metamorphoses (3.16), afraid of being beaten by her mistress, tells Lucius that only her desire to be with him has prevented her from absconding; Daphnis in Longus’ romance, afraid, it may be recalled, of being handed over to a favourite of his master, sees flight with Chloe as a preferable fate; and in Apuleius once more there is the whole community of pastores who flee together when they learn of their owner’s death, alarmed apparently at the prospect of a change of ownership. In Chariton’s novel (4.2), the wrongfully enslaved Chaereas is seen at one stage working on a chaingang in Caria: one night sixteen of his fellow slaves break out of their chains, kill the overseer, and try to escape, but they are caught almost at once and crucified the next day.20 In legal sources references to runaway slaves are also abundant, and one section of the Digest (11.4), though relatively brief, is devoted entirely to them. It suggests that fugitive slaves, like fugitive Christians, might often make for secluded rural areas in which to hide, or that they could commonly be found in ports, presumably trying to board ship. It confirms, too, that fugitives were often given help while on the run or in hiding, by farm bailiffs among others, men who were frequently slaves or ex-slaves themselves but whose chief loyalties were supposed to lie with their owners. The law at least speaks of harbourers acting out of ‘humanity or compassion’ towards runaways.21 The law recognised that it was largely the owner’s responsibility to effect recovery of his property (as seen in reality in the Egyptian evidence and fictionally in Lucian), but the owner could usually count on certain mechanisms of support, troops or civic officials or the provincial governor, as appropriate; the geographical scope of the phenomenon was obviously global. Callistratus 19  Poets: see Gow 1953: 126–7; Gow and Page 1965: 2.628–9; LeGrand 1967: 134–6. Temple: cf. Robertson and Vallette 1965: 2.77 n.3. See also Petr. Sat. 97.2, where a herald gives a description of the ‘runaway’ Giton. The herald’s job was disreputable according to Dio Chrys. 7.123. 20   Pastores (Ap. Met. 8.15–23): note the parallel with accounts from eighteenth-century Russia (Kolchin 1987: 279–81), of whole communities of serfs abandoning their farms and villages in resisting absentee landlords. See also Vita Aesopi 26: natural for a potential purchaser to ask the slave on the block if he intends to run away. 21  Rural areas: Dig. 11.4.1.1; 11,4.1.2; 11.4.1,3; cf. 21.1.17.8. Ports: Dig. 11.4.4. Help: Dig. 11.4.1.1; 11.4.3; cf. 11.3.1 Pr.; 11.3.1.2; 11.3.5 Pr. ‘Humanity’: Dig. 11.3.5 pr.

Resisting Slavery

407

wrote (Dig. 11.4.2) that captured runaways should normally be returned to their owners but that they should be punished ‘more severely’ if they had tried to pass as free while on the loose. His opinion is a reminder that, unlike New World slavery, Roman slavery had no association with skin colour (it was available to all), and it is likely therefore that ‘passing’ was a tactic fugitives commonly attempted. In defining what constituted a runaway, the jurists emphasised intent. If a slave ran away from a brigand or a fire or a collapsing building and was clearly trying to save his life from danger, he was not considered a fugitive; it was only if he planned to escape from his master that he became a runaway. Ulpian’s view (Dig, 11.4.1.5) that the offspring of a runaway female was not a fugitive is a further indication that women as well as men were likely to abscond if they could.22 The ways in which slaves in the Roman world responded to slavery seem to bear a striking resemblance, then, to the ways in which slaves in the New World resisted their servitude. Modern resistance strategies can be paralleled in almost every respect, making plausible the conclusion that the behaviour of Roman slaves illustrated so far represents not simply a set of reactions to slavery, but a determined and at times very conscious demonstration of defiance of oppression. This conclusion is incapable of proof, however, because it depends on evidence that offers no statements of motivation deriving directly from slaves themselves; but to reject its inherent probability is to be unduly sceptical. Admittedly, in some circumstances slaves’ behaviour, while in theory fitting the resistance pattern, might have had nothing at all to do with opposition to slavery. Evidence from the Roman jurists shows that slaves sometimes committed misdeeds because they were coerced by criminous slaveowners—they colluded with their masters, that is, in acts of theft, piracy and even murder. And with the strategy of suicide a variety of motives might have obtained, because like any free person a slave could commit suicide to put an end to the suffering of sickness, or else the demands of loyalty to the master might again come into play: in Alexandria in 30 bc the defeated triumvir M. Antonius ordered a faithful slave named Eros, specially trained for the task, to draw a sword and strike him down. Eros, however, appears to have been so devoted to his master that he thought it better to take his own life than to serve as the instrument of Antonius’ death, for at the critical moment he turned his sword on himself and so inspired his owner, according to Plutarch (Ant. 76), to a similarly heroic end. Again, some instances of suicide may have had no rational explanation at all: 22  Responsibility: Dig. 11.4,1.2; 11.4.1.4; 11.4.3; cf. 18.1.35.3; 21.1.17.9. Mechanisms: Dig. 11.4.1–2; 11.4.4. ‘Passing’: e.g. Dig. 47.2.52.15; 47.10.15.45. Intent: Dig. 21.1.17 pr.–3.

408

Bradley

on the first day of ad 38, a slave named Machaon entered the temple of Jupiter Capitolinus at Rome and began to utter strange prophecies; he next killed a dog which he had brought with him and then killed himself. Whether he was demented, a religious fanatic, both or neither, is unknown.23 The reasons why slaves behaved as they did can seldom have been uncomplicated. Yet the manner in which slaveowners judged both them and their actions was very narrow indeed, consisting of a bland moralistic division into good and bad. Pilfering could be forgiven if a slave had special gifts which endeared him to an owner, but the slave’s tendency to commit wicked deeds was regarded, by Cicero (Att. 7.2.8) among others, as normal, so that when a newly manumitted slave betrayed the trust set in him, abandoned an important assignment and absconded, that was the unpardonable but typical behaviour of a reprobate that left Cicero no alternative but to revoke the grant of freedom. For Quintilian (Inst. 4.2.69), it was a given that slaves should always be trying to explain their shortcomings (‘peccata’), an attitude like that of Cicero and countless others that was predicated on standards of undeviating obedience which slaveowners insisted upon from their slaves. When the standards were met, slaves were good; when not, they were bad.24 In the minds of the Roman lawyers all slaves were corruptible beings, capable of being persuaded to undertake criminal acts or to behave in other morally reprehensible ways. According to Ulpian (Dig. 11.3.1.5), any slave might be led to steal, to damage property, falsify accounts, mismanage his peculium, run away and waste time at the public shows, to cause sedition, behave promiscuously, be insolent to his owner, corrupt other slaves. The lawyers believed that gambling, drunkenness, truancy, excessive attention to the games or to works of art were ‘defects of the mind’ (‘animi vitia’) in slaves, and that their misbehaviour was attributable to the generic characteristics of cunning and effrontery (calliditas, protervitas). According to Gaius (Dig. 21.1.18 pr.), the good slave was loyal, industrious, diligent and thrifty, while the bad slave was fickle, wanton, slothful, sluggish, idle, tardy, a wastrel. The language is reminiscent of the ‘rascality’ and ‘roguery’ that New World slaveowners saw in their slaves, and it is scarcely surprising that Ulpian (Dig. 17.1.18.4) should have described the recalcitrant Roman slave as a ‘servus onerosus’, a troublesome property.25

23  Coerced: Dig. 44.7.20; 48.18.1.5; 47.2.52.23; 25.2.21.1–2; 47.2.35.1. Eros: cf. Dio 51.10.7, and see also Dio 69.22.2, on Hadrian and Mastor. Machaon: Dio 59.9.3 (cf. van Hooff 1990: 97). 24  Standards: see Bradley 1987a: 21–45. 25   Dig. 11.3.1,5: quoted in chapter 2. ‘Defects’: Dig. 2.1.25.6; 21.1.65 pr. Characteristics: Dig. 15.3.3.9; 47.4.1 pr.-1. See further, Bradley 1990: 144–6.

Resisting Slavery

409

However, when representatives of the Roman slaveowning establishment referred to good and bad slaves—when, recall, they wrote of slaves killing themselves out of ‘wickedness’ or because of ‘evil ways’—it did not occur to them to question the appropriateness of their moral categories or to imagine that slaves might not have shared their values. Yet when slaves committed suicide because they were afraid of their owners, or assaulted them because of broken promises or cruel punishment, or ran away to avoid being beaten, or procrastinated at work to gain revenge, they were responding to conditions that for them made the morality of the established order meaningless and unacceptable, just as it was for New World slaves like Thomas Hedgebeth and Harriet Jacobs. ‘Who can blame slaves for being cunning?’ asked Jacobs; ‘They are constantly compelled to resort to it. It is the only weapon of the weak and oppressed against the strength of their tyrants.’ Slave behaviour could only be construed as misbehaviour by those it inconvenienced. Slaves themselves, Roman as much as any other, had to find every means possible to cope with adversity and they constructed in the process a morality that could be diametrically opposed to that of their owners.26 Roman slaveowners acknowledged from time to time that their slaves behaved ‘badly’ because of the way they were treated and not because of inherent character flaws. It was conceded that the master’s threatening words could force the slave to run away, that the dispensator might embezzle because he needed food, that the slave might lie to avoid torture, that apparent greed might have something to do with the slave’s hunger. Cruelty, fear, deprivation—these are recurring elements in the record of master-slave relations in the Roman world, and practical slaveowners could see that injustice caused resentment. Yet owners could never condone their slaves’ misdeeds. The common sense argument was understood, as seen earlier, that a slave who ran away from a collapsing building or a fire in order to save his life was not strictly speaking a fugitive. But the next logical step was not taken: that slaves who ran away to save themselves from physical abuse or some other harsh consequence of slavery were acting in appropriate and defensible ways as victimised human beings. It was this failure to comprehend and to sympathise with the slave mentality that controls the moralistic and often pejorative presentation of slave behaviour in the historical record. It also necessarily obscures the true purpose and meaning of that behaviour.27 26  Harriet Jacobs: Gates 1987: 426. On stealing as a cause of moral dilemma, see McFeeley 1991: 43–4. 27  Conceded: Plut. Mor. 459A; cf. Sen. de Ira 3.5.4; Quint. Inst. 6.3.93; Salvian, de Gubernatione Dei 4.3.16; cf. Philo, de Fuga et Inventione 1.3, and for the jurists’ recognition that flight

410

Bradley

Once the moralistic camouflage is stripped away, however, the plain record of fact remains that Roman slaves reacted to slavery with a variety of strategies that can only be regarded as strategies of resistance intended, to one degree or another, to ameliorate their lives and to reduce the hardships of servitude. As in the slave societies of Brazil, the Caribbean and the American South, acts that slaveowners regarded as crimes were in reality acts that gave ‘outward expression to inward rebelliousness’ in slaves. The owner’s power was absolute; but when power was exercised to the limit of human endurance, slaves responded with assertions of their human capacity that gave rise to struggles of the will and mind that their owners did not always win. Representatives of the slaveowning order wrote prescriptively of the ideal relations they thought should exist between masters and slaves, enjoining upon the latter the obligation to endure their condition with loyalty and obedience. Read uncritically, those writings suggest a willing acquiescence in their lot on the part of slaves and an uninterrupted stability in their dealings with their owners. Beneath the surface calm which elitist writings evince, however, there was a constant ferment of defiant activity as slaves, of every description, ran away, stole, cheated, damaged property and shirked work, or as they directed violence against themselves or their owners, all in an effort to withstand the cruelty and deprivation slavery heaped upon them. At no point in the central period of Roman history can it be said that passive acceptance of subjection was characteristic of the entire slave population.28 To stress the extent of resistance which Roman slaves displayed, across time and place, is to provide a better appreciation of what it was like to live in slavery, since attention has then to be forced on the risks and costs resistance naturally entailed. Any act which questioned the established order exposed the slave to potential danger and so demanded a certain determination, if not courage, before it could be attempted. To consider the hazards involved in flight, in particular, makes the point. Think first of travelling in Italy in safety, taking as a control Horace’s account (Serm. 1.5) of the journey he made, with various companions, from Rome through Campania and Apulia to the port of Brundisium in 37 bc. The distance covered was about three hundred and fifty miles, the time spent on the trip some two weeks, the usual mode of operations to travel leisurely by day from one town to another, with nights spent in lodgings easily paid for, or, could be a response to cruelty (saevitia), see Dig. 21.1.23.1; 21.1.17.3; 21.1.17.4; 21.1.17.12. Resentment: Gel. 10.3.17, on the elder Cato. 28  Quotation: Queirós Mattoso 1986: 133.

Resisting Slavery

411

better, in the villas of friends. Carriages and carts were available for transportation, mules for carrying luggage. The open road (the Via Appia to begin with) took the party from stage to stage of the route. There was always enough to eat, the journey was comfortable and convivial, and the travellers could stay up late into the night to carouse if the fancy took them. There were some discomforts. At an early point Horace took the evening barge from Forum Appii to Terracina and was kept awake all night by biting mosquitoes and croaking frogs. In the marshlands his eyes suffered, a problem later aggravated by smoke from damp firewood in one of the places he stayed. And the food was not always ideal: the local bread was sometimes inedible and the water, if drinkable at all, could upset the stomach. Once a heavy rain washed the road completely away, and the hills of Apulia, scorched by the sirocco, Horace found debilitating.29 Given the company he was keeping—his patron, the rich and powerful C. Maecenas was with him part of the way—Horace was about as safe and secure as anyone could be in antiquity while making his journey. But how would a small group of fugitive slaves, intending to run away, say, from Rome to Brundisium to board ship for an eastern destination, have fared along the same route? How would they have managed to evade detection, to find their way, to feed themselves, to find shelter? Like soldiers on the march, they will have had to make sure that they steered clear of pestilential stopping-places and contaminated water that might poison them, and they will have had to protect themselves from fatigue and sickness caused by the sun, if travelling in the summer, or from frosts and snows if in winter. But worse than this, they will have known that efforts would be made to track them down as soon as they were missed: their owner might pursue them himself, he might send professional slavecatchers after them, he might contact friends or relatives and give instructions to use any means possible to apprehend them. He would certainly circulate their description, and perhaps even resort to magic to prevent them leaving the city limits in the first place,30 Runaways, therefore, will always have had to move swiftly and clandestinely, perhaps travelling by night under the cover of darkness, avoiding all figures of authority, perhaps using a disguise and pretending to be free, taking refuge for sleep in isolated spots away from populous centres. They may have taken some provisions with them at the moment of departure, but after a few days the risks of procuring new rations will have had to be run, exposing them further to the possibility of discovery if they chanced to steal from the farms and 29  Horace’s account: see Fraenkel 1957: 105–12; Rudd 1966: 54–64. 30  Soldiers: Veget. 3.2. Slavecatchers: see Daube 1952. Magic: Plin. Nat. 28.13.

412

Bradley

villages they passed through. If they put their trust in harbourers, there was always the danger of betrayal once rewards for information were announced. As the fugitives coped with the exigencies of weather, terrain and sheer survival, therefore, every moment must have been filled with tension and uncertainty, in utter contrast to the ease with which Horace made his journey. And if in the end the port was safely reached, there was still the obstacle of boarding a ship with which to contend.31 All of this after the fugitives had first resolved to try to escape, had devised some ruse to conceal their departure (in the way the cuttle fish disappeared in a discharge of thick, inklike fluid, Artemidorus said (Onir. 2.14)), after they had made some sort of plan, had chosen a time to put it into effect, and had perhaps made decisions about leaving behind family members and close friends. All of this, too, in the certain knowledge that failure would be followed by punishment and might result in the disgrace of being chained and having to wear iron collars on which their names and addresses were inscribed in case they should ever try to escape again. The following story, though only a fable (Phaedrus, Appendix 20), reveals dramatically and depressingly the deterrent effect upon the slave of the prospect of punishment, and captures the dilemma in which many slaves must have found themselves as they hesitated between remaining in slavery and making a bid for freedom: A slave running away from a master of cruel disposition met Aesop, to whom he was known as a neighbour. ‘What are you excited about?’ asked Aesop. ‘I will tell you frankly, father—and you deserve to be called by that name—since my complaint can be safely entrusted to your keeping. I get a surplus of blows and a shortage of rations. Every now and then I am sent out to my master’s farm without any provisions for the journey. Whenever he dines at home I stand by in attendance all night long; if he is invited out I lie in the street until daybreak. I have earned my liberty, but I am still a slave though grey-headed. If I were aware of any fault on my part, I should bear this with patience. I have never yet had my belly full, and besides that I have the bad luck to suffer tyranny exercised by a cruel master. For these reasons, and others which it would be too long to recount, I have decided to go away wherever my feet shall take me’. ‘Now then, listen’, said Aesop, ‘these are the hardships that you suffer, according to your account, when you have done no wrong; what if you commit 31  The practical hazards faced by runaways are best appreciated from the accounts of escape given by slaves fleeing from the United States to Canada; see especially Drew 1856: 69, 75, 260–1, with Silverman 1985; Ripley 1986: 3–46.

Resisting Slavery

413

an offence? What do you think you will suffer then?’ By such advice the man was deterred from running away.32 Running away, then, was an enterprise requiring enormous courage and resourcefulness, an experience overladen with danger and emotional strain, a response to slavery demanding great resources of inner strength. In the ancient record it is often trivialised, and it is thus tempting for the historian, even now, to brand flight a form of irresponsibility on the part of the slave. Assessing what running away actually involved, however, reveals how almost insuperable odds had to be confronted by those who undertook it, which in itself is a straightforward measure of the simple horror of slavery. Yet those odds were repeatedly faced by great numbers of Roman slaves, and to understand why, the motivating power of the desire to be free must of course be brought to the fore; for in the last analysis, flight, as Ulpian said (Dig. 21.1.17.10), was ‘a form of liberty’ for the slave, that at once brought relief from the power of the master. The risks slaves ran and the costs they paid in resisting slavery must obviously have varied according to the mode of resistance undertaken. An owner might insist that the slave maintain silence while at work, but such a dictate could not be permanently enforced and the level of personal danger involved in wilful disobedience was thus presumably minimal. Yet every time a slave stole an item of clothing or food, or failed to tell the whole truth or sabotaged property, the degree of risk increased considerably. Moreover, plans to assault an owner, if they backfired, could bring severe reprisals for innocent as well as guilty slaves, a factor that dissidents must always have had to consider in their preparations. Plans, in any event, could easily spread within the familia and, failing through betrayal, bring destruction upon their sponsors. In antiquity, the reality of these risks was not a subject that commanded the interest of elitist commentators, but it should not for that reason be glossed over: the fact is that behind almost every record or anecdote of theft, deceit, flight, suicide and so on, there lies a human story, even if now only dimly perceptible, of enormous struggle.33 32  Punishment: Polyb. 1.69.4: already in the middle of the third century BC the slave Spendius, a Campanian who ran away and joined the Carthaginians, was afraid of being recovered by his owner, beaten or put to death. Collars: e.g. ILS 8726. On the emotional elements of flight, see the modern evidence in Drew 1856: 71–2, 211, 282. Unlike American slaves who fled from the South to the North and to Canada, Roman slaves could never know the experience of living in a slaveless society. 33  Plans: Dig. 29.5.1.26; 29.5.1.30. In the modern evidence, the least dangerous mode of resistance was the transmission from one generation to another of subversive trickster stories,

414

Bradley

Incontrovertible generalisations about which forms of resistance were the more prevalent cannot be made. Revolt seems to have been very uncommon, which can be explained by the argument that slaves could successfully express their protests in many other ways that were far less threatening to themselves. At the same time, however, a comprehensive record of slave behaviour is not available and it may be that there were many revolts about which nothing is now known. Similarly, sensationalistic episodes of suicide and attacks on owners found their way into the historical record precisely because they were sensationalistic, and how many other episodes went unrecorded it is impossible to say. No one can prove, therefore, that suicide was more common than assault, assault than theft, and so on. At the impressionistic level, however, running away and petty sabotage seem to have been the most widely deployed means of resistance: the interest in fugitives and ‘troublesome’ slaves displayed by those who created the historical record borders, after all, on the obsessive. Allowing slaves to challenge the existing order, to cause disruption, to prevent the owner from exploiting his property to the full, to reduce the owner’s standing before his peers when an absence of discipline within the household was achieved, flight and day to day resistance were effective and relatively safe strategies for slaves to pursue. Over time the general pattern may well have been affected by changing historical conditions, because New World studies indicate that rates of running away, for instance, increased at times of political weakness or upheaval in the dominant regime. Possibly, therefore, at times of crisis in Roman politics—say in the revolutionary era of the first century bc or in the era of disintegration in the third century ad—there were surges in the numbers of runaway slaves. But there seems to have been no time in the central era at large when Roman society was utterly free from the impact of slave rebelliousness.34 Rebelliousness, however, must not be confused with notions of class solidarity among slaves, and there is no indication that resistance was fuelled by ideological programmes rooted in the desire to secure radical alteration to the structure of society. There was certainly never anything at Rome comparable to the movement that led in St Domingue to the creation of the state of Haiti. Rather, those who made up the Roman slave population, in all its diversity, were concerned with improving their lives as individuals or as members of those for instance of Anansi the spider, Brer Rabbit, and John and Old Marster; see Levine 1977; Roberts 1989. In antiquity, Aesopic fables may have served a similar subversive purpose among slaves: see Bradley 1987a: 150–3. On Aesop himself as a trickster figure, see Hopkins 1993, and cf. Winkler 1985: 279–86. 34  Obsessive: Finley 1980: 111, on fugitives. New World studies: e.g. Lovejoy 1986: 74.

Resisting Slavery

415

small groups through whatever means of self-help they could find, so that it was personal, not collective, independence that was their object.35 It has been said that ‘resistance not acquiescence is the core of history’, a view that those who regard history as a ceaseless process of struggle between the forces of exploitation and the forces of opposition to exploitation would readily endorse. This is a view, however, that is based neither on a series of propositions that can be empirically verified or refuted nor on a set of conclusions drawn from a dispassionate analysis of evidence; and its implication that all the historically downtrodden have always consciously devoted themselves to resistance fails to allow for variations of individual strength and capability. Yet it is undeniably true that throughout history there have been many slaves who have withstood the oppression inherent in their condition. The proof lies in the combination of their words and actions, which leave the issue beyond doubt. Harriet Jacobs wrote that she was just fourteen when she first felt the conviction of having to fight slavery: ‘The war of my life had begun’, she said, ‘and though one of God’s most powerless creatures, I resolved never to be conquered’. William Troy, an American slave who took refuge in Canada, similarly told Benjamin Drew: ‘I intend to be a terror to the system while I live.’ For Roman slavery, statements of that kind do not exist. In all fundamentals, however, in terms of authority, control and the manipulation of power, the social relationship between slaveowner and slave was precisely the same at Rome as in the slave societies of the New World; and an objective assessment of the deeds and actions of Roman slaves shows that they fall into precisely the same categories of resistance behaviour associated with slaves in those same societies. It follows that resistance had a structural, and elemental, place in the history of slavery at Rome.36

List of Works Cited

Alföldy, G. (1969) Fasti Hispanienses. Wiesbaden. Alföldy, G. (1988) The Social History of Rome. Baltimore. Astin, A. E. (1978) Cato the Censor. Oxford. 35  See the interesting parallel provided by the Islamic slave society of the Sokoto Caliphate in nineteenth-century West Africa, as described in Lovejoy 1986, and cf. Karasch 1987: 325–6, explaining the absence of slave revolts in Rio de Janeiro by the absence of slave class cohesion. 36  Quotation: Herbert Aptheker in Craton 1986: 96 (cf. 113 n.1). Harriet Jacobs: Gates 1987: 353. William Troy: Drew: 1856: 355.

416

Bradley

Bagnall, R. S. (1993) ‘Slavery and society in late Roman Egypt’, in Baruch Halpern and Deborah W. Hobson, eds., Law, Politics and Society in the Ancient Mediterranean World, 220–38. Sheffield. Baldwin, B. (1978) ‘Trimalchio’s domestic staff’, A Class 21: 87–97. Barnes, T. D. (1981) Constantine and Eusebius. Cambridge, MA and London. Basler, R. P., ed. (1953) The Collected Works of Abraham Lincoln. New Brunswick, NJ. Bauer, R. A. and A. H. Bauer (1942) ‘Day to day resistance to slavery’, Journal of Negro History 27: 388–419. Bauman, R. A. (1980) ‘The “Leges iudiciorum publicorum” and their interpretation in the Republic, Principate and Later Empire’, ANRW 11 13: 103–233. Bellen, H. (1971) Studien zur Sklavenflucht in römischen Kaiserreich. Wiesbaden. Biëzuńska-Małowist 1. (1977) L’Esclavage dans l’Egypto gréco-romaine, Seconde partie: Période romaine. Warsaw. Birley, E. (1953) Roman Britain and the Roman Army. Kendal. Birley, E. (1980) ‘Law in Roman Britain’, ANRW II 13: 609–25. Blackburn, R. (1988) ‘Slavery—its special features and social role’, in Léonie J. Archer, ed., Slavery and Other Forms of Unfree Labour, 262–79. London and New York. Blassingame, J. W. (1975) ‘Using the testimony of ex-slaves; approaches and problems’, Journal of Southern History 41: 473–92. Blassingame, J. W. (1979) The Slave Community: Plantation Life in the Antebellum South. New York and Oxford. Bömer, F. (1957) Untersuchungen über die Religion der Sklaven in Griechenland und Rom I: Die wichtigsten Kulte und Religionen in Rom und in lateinischen Westen (2nd edn 1981). Wiesbaden. Boswell, J. (1988) The Kindness of Strangers: The Abandonment of Children in Western Europe from Late Antiquity to the Renaissance. New York. Bowie, E. L. (1977) ‘The novels and the real world’, in B. P. Rcardon, ed., Erotica Antiqua (Acta of the International Conference on the Ancient Novel, Bangor 1976), 91–96. Bangor. Bowie, E. L. (1985) ‘Theocritus’ seventh Idyll, Philetas and Longus’, CQ 35: 67–91. Bradley, K. R. (1986a) ‘Social aspects of the Roman slave trade’, MBAH 5:49–58. Bradley, K. R. (1986b) ‘Seneca and Slavery’, C&M 37: 161–72. Bradley, K. R. (1987a) Slaves and Masters in the Roman Empire: A Study in Social Control. New York and Oxford. Bradley, K. R. (1987b) ‘On the Roman slave supply and slavebreeding’, in M. I. Finley, ed., Classical Slavery, 42–64. London (= Slavery & Abolition 8: 42–64). Bradley, K. R. (1988) ‘Roman slavery and Roman law’, Historical Reflections/Réflexions Historiques 15: 477–95. Bradley, K. R. (1989) Slavery and Rebellion in the Roman World, 140 B.C.–70 B.C. Bloomington, Indianapolis and London.

Resisting Slavery

417

Bradley, K. R. (1990) ‘Servus onerosus: Roman law and the troublesome slave’, Slavery & Abolition 11: 135–57. Bradley, K. R. (1991) Discovering the Roman Family: Studies in Roman Social History. New York and Oxford. Braund, D. C. and G. R. Tsetskhladze (1989) ‘The export of slaves from Colchis’, CQ 39: 114–25. Brown, P. R. L. (1988) The Body and Society: Men, Women and Sexual Renunciation in Early Christianity. New York. Brunt, P. A. (1971) Italian Manpower 225 B. C.–A. D. 14. Oxford. Brunt, P. A. (1973) ‘Aspects of the social thought of Dio Chrysostom and of the Stoics’, PCPhS 19: 9–34. Brunt, P. A. (1977) ‘From Epictetus to Arrian’, Athenaeum 55: 19–48. Brunt, P. A. (1980a) ‘Free labour and public works at Rome’, JRS 70: 81–100. Brunt, P. A. (1980b) ‘Evidence given under torture in the Principate’, ZRG 97: 256–65. Brunt, P. A. (1987) ‘Labour’, in J. Wacher, ed., The Roman World 11 701–16. London and New York. Brunt, P. A. (1988) The Fall of the Roman Republic and Related Essays. Oxford. Brunt, P. A. (1990) Roman Imperial Themes. Oxford. Brunt, P. A. (1993) ‘Aristotle and Slavery’, in Studies in Greek History and Thought, 343–88. Oxford. Bugkland, W. W. (1908) The Roman Law of Slavery. Cambridge. Cadoux, C. J. (1925) The Early Church and the World: A History of the Christian Attitude to Pagan Society and the State Down to the Time of Constantinus. Edinburgh. Calvert, B. (1987) ‘Slavery in Plato’s Republic’, CQ 37: 367–72. Cambiano, G. (1987) ‘Aristotle and the anonymous opponents of slavery’, in M. I. Finley, ed., Classical Slavery, 21–41. London (= Slavery & Abolition 8: 21–41). Campbell, J. B. (1984) The Emperor and the Roman Army 31 BC–AD 235. Oxford. Carandini, A. (1981) ‘Sviluppo e crisi delle manifatture rurali e urbane’, in A. Giardina and A. Schiavone, eds., Società romana e produzione schiavistica, v. 2, 249–60. Rome and Bari. Cartledge, P. A. (1985) ‘Rebels & Sambos in Classical Greece: a comparative view’, in P. A. Cartledge & F. D. Harvey, eds., Crux: Essays presented to G. E. M. de Ste. Croix on his 75th birthday, 16–46. London. Casson, L. (1984) Ancient Trade and Society, Detroit. Chadwick, H. (1983) ‘New Letters of St. Augustine’, JThS 34: 425–52. Clarke, J. R. (1991) The Houses of Roman Italy, 100 B.C.–A.D. 250: Ritual, Space, and Decoration. Berkeley, Los Angeles and Oxford. Coarelli, F. (1988) Il foro boario: dalle origini alla fine della repubblica. Rome. Conrad, R. E. (1983) Children of God’s Fire: A Documentary History of Black Slavery in Brazil. Princeton.

418

Bradley

Corbier, M. (1989) ‘The ambiguous status of meat’, Food and Foodways 3: 223–64. Cotton, H. M. (1981) Documentary Letters of Recommendation in Latin from the Roman Empire. Königstein/Ts. Cotton, H. M. (1986) ‘The role of Cicero’s letters of recommendation: iustitia versus gratia?’, Hermes 1 14: 443–60. Craton, M. (1982) Testing the Chains: Resistance to Slavery in the British West Indies. Ithaca and London. Craton, M. (1986) ‘From Caribs to Black Caribs: the Amerindian roots of servile resistance in the Caribbean’, in Gary Y. Okihiro, ed., In Resistance: Studies in African, Caribbean, and Afro-American History, 96–116. Amherst, MA. Crawford, M. H. (1977) ‘Republican denarii in Romania: the suppression of piracy and the slave-trade’, JRS 67: 117–24. Crawford, M. H. (1985) Coinage and Money under the Roman Republic: Italy and the Mediterranean Economy. London. Crook, J. A. (1955) Consilium Principis: Imperial Councils and Counsellors from Augustus to Diocletian. Cambridge. Curtin, P. D. (1969) The Atlantic Slave Trade: A Census. Madison, Milwaukee and London. D’Arms, J. H. (1991) ‘Slaves at Roman convivia’, in W. J. Slater, ed., Dining in a Classical Context, 171–83. Ann Arbor, MI. Daube, D. (1946) ‘Two early patterns of manumission’, JRS 36: 57–75. Daube, D. (1952) ‘Slave-catching’, Juridical Review 64: 12–28. Davies, R. W. (1989) Service in the Roman Army. Edinburgh. Davis, D. B. (1966) The Problem of Slavery in Western Culture. Ithaca (augmented reprint, New York 1988). Davis, D. B. (1975) The Problem of Slavery in the Age of Revolution 1770–1823. Ithaca and London. Degler, C. N. (1971) Neither Black Nor White: Slavery and Race Relations in Brazil and the United States. New York and London. Desanges, J. (1976) ‘The iconography of the black in ancient North Africa’, in Ladislas Bugner, ed., The Image of the Black in Western Art I: From the Pharaohs to the Fall of the Roman Empire, 246–68. New York Dionisotti, A. C. (1982) ‘From Ausonius’ schooldays? A Schoolbook and its relatives’, JRS 72: 83–125. Drescher, S. (1986) Capitalism and Antislavery: British Mobilization in Comparative Perspective. New York. Drew, B. (1856) The Refugee: Or the Narratives of Fugitive Slaves in Canada. Boston. Duchêne, H. (1986) ‘Sur la stèle d’Aulus Caprilius Timotheus, sômatemporos’, BCH 110: 513–30. Dumont, J. C. (1987) Servus: Rome et l’esclavage sous la république. Rome. Dunbabin, K. M. D. (1978) The Mosaics of Roman North Africa: Studies in Iconography and Patronage. Oxford.

Resisting Slavery

419

Dunbabin, K. M. D. (1989) ‘Baiarum grata voluptas: Pleasures and dangers of the baths’, PBSR 57: 6–46. Duncan-Jones, R. P. (1974) The Economy of the Roman Empire: Quantitative Studies. Cambridge (2nd edn 1982). Dyson, S. L. (1992) Community and Society in Roman Italy. Baltimore and London. Edmondson, J. C. (1987) Two Industries in Roman Lusitania: Mining and Garum Production. Oxford. Edmondson, J. C. (1989) ‘Mining in the later Roman empire and beyond: continuity or disruption?’, JRS 79: 84–102. Eltis, D. (1987) Economic Growth and the Ending of the Transatlantic Slave Trade. New York and Oxford. Étienne, R. (1981) ‘Les rations alimentaires des esclaves de la « familia rustica » d’après Caton’, Index 10: 66–77. Evans, J. K. (1978) ‘The role of suffragium in imperial political decisionmaking: a Flavian example’, Historia 27: 102–28. Evans, J. K. (1980) ‘Plebs rustica. The peasantry of classical Italy II’, AJAH 5: 134–73. Finley, M. I. (1973) The Ancient Economy. Berkeley and Los Angeles (2nd edn 1985). Finley, M. I. (1980) Ancient Slavery and Modern Ideology. New York. Finley, M. I. (1981) Economy and Society in Ancient Greece. London. Flory, M. B. (1978) ‘Family in “familia”: kinship and community in slavery’, AJAH 3: 78–95. Fogel, R. W. (1989) Without Consent or Contract: The Rise and Fall of American Slavery. New York and London. Fox-Genovese, E. (1988) Within the Plantation Household: Black and White Women of the Old South. Chapel Hill and London. Foxhall, L. (1990) ‘The dependent tenant: land leasing and labour in Italy and Greece’, JRS 80: 97–114. Foxhall, L. and H. A. Forbes (1982) ‘Sitometreia: the role of grain as a staple food in classical antiquity’, Chiron 12: 41–90. Fraenkel, E. (1957) Horace. Oxford. Frayn, J. M. (1979) Subsistence Farming in Roman Italy. London. Frayn, J. M. (1984) Sheep-Rearing and the Wool Trade in Italy during the Roman Period. Liverpool. Freyburger, G. (1992) ‘Le droit d’asile à Rome’, LEC 60: 139–51. Gardner, J. F. (1991) ‘The Purpose of the Lex Fufia Caninia’, EMC 10: 21–39. Gardner, J. F. and T. E. J. Wiedemann (1991) The Roman Household: A Sourcebook. London and New York. Garnsey, P. D. A. (1968) ‘Why penalties become harsher: the Roman case, late Republic to fourth-century Empire’, Natural Law Forum 13: 141–62. Garnsey, P. D. A. (1970) Social Status and Legal Privilege in the Roman Empire. Oxford.

420

Bradley

Garnsey, P. D. A. (1988) Famine and Food Supply in the Graeco-Roman World: Responses to Risk and Crisis. Cambridge. Garnsey, P. D. A. (1991) ‘Mass diet and nutrition in the city of Rome’, in A. Giovannini, ed., Nourrir la plèbe, 67–101. Zurich. Garnsey, P. D. A. and R. P. Saller (1987) The Roman Empire: Economy, Society and Culture. Berkeley and Los Angeles. Garrido-Hory, M. (1981) Martial et l’esclavage. Paris. Gaspar, D. B. (1985) Bondmen and Rebels: A Study of Master-Slave Relations in Antigua. Baltimore. Gates, H. L., Jr., ed. (1987) The Classic Slave Narratives. New York. Genovese, E. D. (1972) Roll, Jordan, Roll: The World the Slaves Made. New York. Goodman, M. (1983) State and Society in Roman Galilee, AD 132–212. Totowa, New Jersey. Gow, A. S. F. (1953) The Greek Bucolic Poets. Cambridge. Gow, A. S. F. and D. L. Page, eds. (1965) The Greek Anthology: Hellenistic Epigrams. Cambridge. Graf, F. (1992) ‘Gestures and conventions: the gestures of Roman actors and orators’, in Jan Bremmer and Herman Roodenburg, eds., A Cultural History of Gesture, 36–58. Ithaca. Green, R. P. H., ed. (1991), The Works of Ausonius. Oxford. Griffin, M. T. (1976) Seneca: A Philosopher in Politics. Oxford. Grisé, Y. (1982) Le Suicide dans la Rome antique. Montreal and Paris. Gruen, E. S. (1984) The Hellenistic World and the Coming of Rome. Berkeley, Los Angeles and London. Guthrie, W. K. C. (1969) A History of Greek Philosophy: Volume III, The Fifth-Century Enlightenment. Cambridge. Gwyn, J. (1992) ‘The economics of the transatlantic slave trade: a review’, Histoire Sociale/Social History 25: 151–62. Hägg, T. (1983) The Novel in Antiquity. Berkeley and Los Angeles. Harris, W. V. (1979) War and Imperialism in Republican Rome 327–70 B.C. Oxford (repr. with corrections 1985). Harris, W. V. (1980) ‘Towards a study of the Roman slave trade’, MAAR 36: 117–40. Harris, W. V. (1989) Ancient Literacy. Cambridge, MA and London. Harris, W. V. (1990) ‘Roman warfare in the economic and social context of the fourthcentury B.C.’, in Walter Eder, ed., Staat und Staatlichkeit in der frühen römischen Republik, 494–510. Stuttgart. Hollis, A. S., ed. (1970) Ovid, Metamorphoses Book VIII. Oxford. Hopkins, K. (1978) Conquerors and Slaves: Sociological Studies in Roman History, Volume 1. Cambridge. Hopkins, K. (1993) ‘Novel evidence for Roman slavery’, P & P no. 138: 3–27. Hübner, R. (1978) ‘Four Oxyrhynchos papyri’, ZPE 30: 195–207.

Resisting Slavery

421

Jashemski, W. F. (1979) The Gardens of Pompeii. New Rochelle, NY. Jones, A. H. M. (1956) ‘Slavery in the ancient world’, Economic History Review 9: 185–99 (= M. I. Finley, ed., Slavery in Classical Antiquity (Cambridge and New York 1960, repr. with bibliographical additions 1968), 1–15). Joshel, S. R. (1992) Work, Identity and Legal Status at Rome: A Study of the Occupational Inscriptions. Norman and London. Karasch, M. C. (1987) Slave Life in Rio de Janeiro 1808–1850. Princeton. Kenney, E. J., ed. (1984) The Ploughman’s Lunch: Moretum: A Poem Ascribed to Virgil. Bristol. Kirschenbaum, A. (1987) Sons, Slaves and Freedmen in Roman Commerce. Washington, DC. Klein, R. (1988) Die Sklaverei in der Sicht der Bischöfe Ambrosius und Augustinus. Stuttgart. Kolchin, P. (1987) Unfree Labor: American Slavery and Russian Serfdom. Cambridge, M.A. and London. Kolendo, J. (1981) ‘L’esclavage et la vie sexuelle des hommes libres à Rome’, Index 10: 288–97. Lane Fox, R. (1986) Pagans and Christians. New York and London. Legrand, PH.-E., ed. (1967) Bucoliques Grecs II. Paris. Lepelley, CL. (1981) ‘La crise de l’Afrique romaine au début du Ve siècle d’après les lettres nouvellement découvertes de Saint Augustin’, CRAI 1981:445–63. Lepper, F. A. and S. S. Frere (1988) Trajan’s Column: A New Edition of the Cichorius Plates. Gloucester. Levick, B. M. (1990) Claudius. New Haven and London. Levine, L. W. (1977) Black Culture and Black Consciousness: Afro-American Folk Thought from Slavery to Freedom. New York and Oxford. Lévy-Bruhl, H. (1934) ‘Théorie de l’esclavage’, in Quelques problèmes du très ancien droit romain, 15–33. Paris (= M. I. Finley, ed., Slavery in Classical Antiquity (Cambridge and New York 1960, repr. with bibliographical additions 1968), 151–69). Lewis, B. (1990) Slavery in the Middle East: An Historical Enquiry. New York and Oxford. Lewis, N. (1983) Life in Egypt under Roman Rule. Oxford. Lovejoy, P. E. (1986) ‘Fugitive slaves: Resistance to slavery in the Sokoto Caliphate’, in Gary Y. Okihiro, ed., In Resistance: Studies in African, Caribbean, and Afro-American History, 71–95. Amherst, M.A. Lutz, Cora (1947) ‘Musonius Rufus: the Roman Socrates’, Yale Classical Studies 10: 3–147. Macmullen, R. (1986) ‘Judicial savagery in the Roman Empire’, Chiron 16: 147–66 (= MacMullen 1990: 204–17). Macmullen, R. (1987) ‘Late Roman slavery’, Historia 36: 359–83 (= MacMullen 1990: 236–49). Macmullen, R. (1990) Changes in the Roman Empire: Essays in the Ordinary. Princeton.

422

Bradley

Macqueron, J. (1945) ‘Le testament d’Antonius Silvanus’, RD 23: 123–70. Manning, C. E. (1989) ‘Stoicism and slavery in the Roman Empire’, ANRW 11 36.3: 1518–43. Martin, D. B. (1990) Slavery as Salvation. New Haven and London. Matthews, J. (1989) The Roman Empire of Ammianus. London and Baltimore. Mgfeely, W. S. (1991) Frederick Douglass. New York and London. Mcginn, T. A. J. (1990) ‘Ne serva prostituatur. Restrictive covenants in the sale of slaves’, ZRG 107: 315–53. Mclaurin, M. A. (1991) Celia, A Slave. Athens, GA. Mellon, J., ed. (1988) Bullwhip Days: The Slaves Remember: An Oral History. New York. Millar, F. G. B. (1965) ‘Epictetus and the imperial court’, JRS 55: 141–8. Millar, F. G. B. (1977) The Emperor in the Roman World. London (2nd edn 1992). Millar, F. G. B. (1981) ‘The World of the Golden Ass’, JRS 71: 63–75. Miller, R. M., ed. (1990) Dear Master: Letters of a Slave Family. Athens, GA and London. Montevecchi, O. (1973) La papirologia. Turin. Nicholson, O. (1989) ‘Flight from persecution as imitation of Christ: Lactantius’ Divine Institutes IV.18, 1–2’, JThS 40: 48–65. Nicolet, CL. (1991) Space, Geography, and Politics in the Early Roman Empire. Ann Arbor, MI. Nock, A. D. (1928) ‘Early gentile Christianity and its Hellenistic background’, in A. E. J. Rawlinson, ed., Essays on the Trinity and the Incarnation, 55–156. London (= Essays on Religion and the Ancient World (Oxford 1972), 49–133). Oxé, A. and H. Comfort, eds. (1968) Corpus Vasorum Arretinorum. Bonn. Packer, J. E. (1975) ‘Middle and lower class housing in Pompeii and Herculaneum’, in Bernard Andreae and Helmut Kyrieleis, eds., Neue Forschungen in Pompeji und den anderen vom Vesuvausbruch 79 n. Chr. verschütteten Städten, 133–42. Recklinghausen. Pagels, E. (1988) Adam, Eve, and the Serpent. New York. Patterson, O. (1982) Slavery and Social Death: A Comparative Study. Cambridge, MA and London. Peacock, D. P. S. (1982) Pottery in the Roman World: An Ethnoarchaeological Approach. London and New York. Pelteret, D. (1981) ‘Slave raiding and slave trading in early England’, Anglo-Saxon England 9: 99–114. Pitts, L. F. (1989) ‘Relations between Rome and the German “Kings” on the middle Danube in the first to the fourth centuries A.D.’, JRS 79: 45–58. Price, R., ed. (1979) Maroon Societies: Rebel Slave Communities in the Americas, 2nd edn. Baltimore. Price, S. F. R. (1986) ‘From Freud to Artcmidorus’, P & P 113: 3–37. Pucci, G. (1973) ‘La produzione della ceramica aretina. Note sull’ “industria” nella prima età imperiale romana’, D Arch 7: 255–93.

Resisting Slavery

423

Pucci, G. (1981) ‘La ceramica italica (terra sigillata)’, in A. Giardina and A. Schiavone, eds., Società romana e produzione schiavistica, v. 2, 99–121. Rome and Bari. Queirös Mattoso, K. M. De (1986) To Be A Slave in Brazil 1550–1888. New Brunswick, NJ. Ramage, N. H. and A. Ramage (1991) Roman Art. Englewood Cliffs, NJ. Rawick, G. P. (1972) The American Slave: A Composite Autobiography, Volume I, From Sundown to Sunup: The Making of the Black Community. Westport, CT. Rawson, B. M. (1986) ‘Children in the Roman familia’, in Beryl Rawson, ed., The Family in Ancient Rome: New Perspectives, 170–200. London and Sydney. Rawson, E. (1985) Intellectual Life in the Late Roman Republic. London and Baltimore. Reardon, B. P. (1991) The Form of Greek Romance. Princeton. Richardson, J. S. (1976) ‘The Spanish mines and the development of provincial taxation in the second century B. C.’, JRS 66: 139–52. Richmond, I. A. (1953) ‘Three Roman writing-tablets from London’, AntJ 33: 206–8. Rickman, G. E. (1980) The Corn Supply of Ancient Rome. Oxford. Ripley, C. P., ed. (1986) The Black Abolitionist Papers, Volume II, Canada, 1830–1865. Chapel Hill and London. Roberts, J. W. (1989) From Trickster to Badman: The Black Folk Hero in Slavery and Freedom. Philadelphia. Robertson, D. S. and P. Vallette, eds. (1965) Apulée: Les Métamorphoses II Paris. Robertson, N. (1987) ‘The Nones of July and Roman weather magic’, MH 44: 8–41. Robinson, O. F. (1981) ‘Slaves and the criminal law’, ZRG 98: 213–54. Robleda, O. (1976) Il diritto degli schiavi nell’ antica Roma. Rome. Rose, W. L., ed. (1976) A Documentary History of Slavery in North America. New York and London. Rossi, L. (1971) Trajan’s Column and the Dacian Wars. London. Rudd, N. (1966) The Satires of Horace. Cambridge. Rutherford, R. B. (1989) The Meditations of Marcus Aurelius: A Study. Oxford. Sacks, K. S. (1990) Diodorus Siculus and the First Century. Princeton. Ste Croix, G. E. M. de (1975) ‘Early Christian attitudes towards property and slavery’, Studies in Church History 12: 1–38. Ste Croix, G. E. M. de (1981) The Class Struggle in the Ancient Greek World from the Archaic Age to the Arab Conquests. London and Ithaca. Saller, R. P. (1982) Personal Patronage under the early Empire. Cambridge (1987) ‘Slavery and the Roman Family’, in M. I. Finley, ed., Classical Slavery, 65–87. London (= Slavery & Abolition 8: 65–87). Saller, R. P. (1991) ‘Corporal punishment, authority, and obedience in the Roman household’, in Beryl Rawson, ed., Marriage, Divorce, and Children in Ancient Rome, 144–65. Canberra and Oxford. Samson, R. (1989) ‘Rural slavery, inscriptions, archaeology, and Marx: a response to Ramsay MacMullen’s “Late Roman Slavery” ’, Historia 38: 99–110.

424

Bradley

Schlaifer, R. (1936) ‘Greek theories of slavery from Homer to Aristotle’, HSPh 47: 165–204 (= M. I. Finley, ed., Slavery in Classical Antiquity (Cambridge and New York 1960, repr. with bibliographical additions 1968), 93–132). Shaw, B. D. (1984) ‘Bandits in the Roman Empire’, P & P 105: 3–52. Shaw, B. D. (1990) ‘Bandit highlands and lowland peace: the mountain of IsauriaCilicia’, JESHO 33: 199–233, 237–70. Shelton, K. J. (1981) The Esquiline Treasure. London. Sherwin-White, A. N. (1966) The Letters of Pliny: A Historical and Social Commentary. Oxford. Silverman, J. H. (1985) Unwelcome Guests: Canada West’s Response to American Fugitive Slaves, 1800–1865. Millwood, NY, New York and London. Snowden, F. M., Jr. (1976) ‘Iconographical evidence on the black populations in GrecoRoman antiquity’, in Ladislas Bugner, ed., The Image of the Black in Western Art 1: From the Pharaohs to the Fall of the Roman Empire, 133–245. New York. Sotgiu, G. (1973/1974) ‘Un collare di schiavo rinvenuto in Sardegna’, ArchClass 25–26: 688–97. Speidel, M. D. and S. Panciera (1989) ‘From the North and Black Sea shores: Two new gravestones for boys of the “equites singulares Augusti” ’, Chiron 19: 119–26. Spurr, M. S. (1986) Arable Cultivation in Roman Italy c. 200 B.C.–c. A.D. 100. London. Spurr, M. S. (1990) ‘Agriculture and the Georgics’, in Ian McAuslan and Peter Walcot, eds., Virgil: Greece and Rome Studies, 69–93. Oxford. Stadter, P. A. (1980) Arrian of Nicomedia. Chapel Hill. Stampp, K. M. (1956) The Peculiar Institution: Slavery in the Ante-Bellum South. New York. Starr, C. G. (1949) ‘Epictetus and the tyrant’, CPh 44: 20–9. Stefanelli, L. P. B., ed. (1990) llbronzodei Romani: arredo e suppellettile. Rome. Strasburger, H. (1965) ‘Poseidonios on problems of the Roman Empire’, JRS 55: 40–53. Syme, R. (1979) Roman Papers I & II. Oxford. Syme, R. (1988) Roman Papers IV & V Oxford. Syme, R. (1991) Roman Papers VI & VII. Oxford. Tchernia, A. (1983) ‘Italian wine in Gaul at the end of the Republic’, in Peter Garnsey, Keith Hopkins, and C. R. Whittaker, eds., Trade in the Ancient Economy, 87–104, London. Tchernia, A. (1986) Le Vin de l’Italie romaine. Paris and Rome. Temperley, H. (1972) British Antislavery: 1833–1870. Columbia, SC. Thomas, J. A. C. (1975) The Institutes of Justinian. Amsterdam and Oxford. Thompson, E. A. (1957) ‘Slavery in early Germany’, Hermathena 89: 17–29 (= M. I. Finley, ed., Slavery in Classical Antiquity (Cambridge and New York 1960, repr. with bibliographical additions 1968), 191–203). Thompson, E. A. (1985) Who Was St. Patrick? Bury St Edmunds, Suffolk.

Resisting Slavery

425

Thompson, L. A. (1989) Romans and Blacks. London and New York. Toynbee, J. M. C. (1971) Death and Burial in the Roman World. London. Treggiari, S. M. (1969) Roman Freedmen during the Late Republic. Oxford. Treggiari, S. M. (1973) ‘Domestic staff at Rome during the Julio-Claudian period’, Histoire Sociale/Social History 6: 241–55. Treggiari, S. M. (1975) ‘Jobs in the household of Livia’, PBSR 43: 48–77. Van Hooff, A. J. L. (1990) From Autothanasia to Suicide: Self-Killing in Classical Antiquity. London and New York. Versnel, H. S. (1970) Triumphus: An Enquiry into the Origin, Development and Meaning of the Roman Triumph. Leiden. Veyne, P. (1987) ‘The Roman Empire’, in Philippe Ariès and Georges Duby, eds., A History of Private Life I: From Pagan Rome to Byzantium, 5–234. Cambridge, MA and London. Vlastos, G. (1941) ‘Slavery in Plato’s Thought’, The Philosophical Review 50: 289–304 (= M. I. Finley, ed., Slavery in Classical Antiquity (Cambridge and New York 1960, repr. with bibliographical additions 1968), 133–49). Vogt, J. (1975) Ancient Slavery and the Ideal of Man. Cambridge, MA. Wade, R. C. (1964) Slavery in the Cities: The South 1820–1860. New York. Walbank, F. W. (1979) A Historical Commentary on Polybius, Volume III. Oxford. Wallace-Hadrill, A. (1982) ‘The golden age and sin in Augustan ideology’, P & P 95: 19–36. Wallace-Hadrill, A. (1988) ‘The social structure of the Roman house’, PBSR 56: 43–97. Wallace-Hadrill, A. (1991) ‘Houses and households: sampling Pompeii and Herculaneum’, in Beryl Rawson, ed., Marriage, Divorce, and Children in Ancient Rome, 191–227. Canberra and Oxford. Watson, A. (1967) The Law of Persons in the Later Roman Republic. Oxford. Watson, A. (1968) ‘Morality, slavery and the jurists in the later Roman Republic’, Tulane Law Review 1968: 289–303. Watson, A. (1975) Rome of the XII Tables: Persons and Property. Princeton. Watson, A. (1983) ‘Roman slave law and Romanist ideology’, Phoenix 37: 53–65. Weaver, P. R. C. (1972) Familia Caesaris: A Social Study of the Emperor’s Freedmen and Slaves. Cambridge. Weaver, P. R. C. (1990) ‘Where have all the Junian Latins gone? Nomenclature and status in the early Empire’, Chiron 20: 275–304. Welwei, K.-W. (1988) Unfreie im antiken Kriegsdienst, III: Rom. Stuttgart. Whittaker, C. R. (1987) ‘Circe’s pigs: from slavery to serfdom in the later Roman world’, in M. I. Finley, ed., Classical Slavery, 88–122. London (= Slavery & Abolition 8: 88–122). Wiedemann, T. E. J. (1985) ‘The regularity of manumission at Rome’, CQ 35: 162–75. Wiedemann, T. E. J. (1987) Slavery: Greece and Rome, New Surveys in the Classics, no. 19. Oxford. Wightman, E. M. (1971) Roman Trier and the Treveri. New York and Washington, DC.

426

Bradley

Wilson, L. M. (1938) The Clothing of the Ancient Romans. Baltimore. Winkler, J. J. (1985) Auctor & Actor: A Narratological Reading of Apuleius’s The Golden Ass. Berkeley, Los Angeles and Oxford. Wiseman, T. P. (1985) Catullus and His World: A Reappraisal. Cambridge. Yavetz, Z. (1988) Slaves and Slavery in Ancient Rome, New Brunswick, NJ and London. Yellin, J. F., ed. (1987) Incidents in the Life of a Slave Girl. Cambridge, MA and London.

Body Work: Slavery and the Pauline Churches Jennifer A. Glancy Travelers to the towns and cities of the Roman Empire customarily sought shelter with those they knew, or they carried letters of introduction from families and friends. When Lucius, the hero of Apuleius’s Golden Ass, arrives in the town of Hypata in Thessaly, he inquires at a public inn for the home of Milo. Lucius is previously unacquainted with Milo, but a friend has provided him with a letter of introduction. The old woman who directs Lucius to Milo’s house notes as a sign of Milo’s meanness that, although he is a wealthy man, the only inhabitants of his house are his wife and a single female slave. Lucius knocks at Milo’s gate and his knocking brings the slave, Photis, to the gate. The rudeness of the house is apparent in her interrogation of the visitor. When Lucius presents her with the letter of introduction, she again bars the gate as she brings the letter inside to Milo.1 Acts of the Apostles presents a scene with similar details, set in another urban environment in the Eastern Empire, Jerusalem. After King Herod orders the killing of James, the brother of John, he orders the arrest of Peter. The night before Peter is scheduled to appear before Herod, an angel appears in his cell and releases his chains. The angel leads Peter past the guards. The iron gate opens for them, and they go into the city. Realizing where he is, Peter proceeds to the house of his fellow believer Mary. A prayer meeting is in progress at her house. When Peter pounds on the gate, Rhoda, a slave, arrives at the gate to see who is causing the disturbance. Like Photis, she leaves the visitor at the gate while she runs into the house to report the appearance of the visitor, although the text attributes her behavior not to rudeness but to flustered joy (12:2–14).2 The appearance of these two slaves when visitors knock at the gates of the houses in which they dwell ushers us into the world of urban slaves in the Eastern Empire. Both Photis and Rhoda live in modest slaveholding establishments. Households with a hundred or more slaves existed but were far less Source: Glancy, Jennifer A., “Body Work: Slavery and the Pauline Churches,” in Jennifer A. Glancy, Slavery in Early Christianity, New York: Oxford University Press, 2002, 39–70. By permission of Oxford University Press, © 2002 by Jennifer A. Glancy. 1  Apul., Met. 1.21–22. 2  Harrill argues that Rhoda exemplifies the servus currens, or running slave, a stock character in New Comedy (“Dramatic Function of the Running Slave Rhoda”).

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_018

428

Glancy

numerous than smaller slaveholding establishments, families that held one or two slaves or a half dozen slaves. In contrast to the miserly Milo with his single slave, the traveler Lucius is eventually joined even on his travels by two of his own slaves, who follow him on foot. Milo’s hospitality to Lucius extends to these two slaves. This is only a minor act of generosity; hosting a few extra slaves causes Milo little trouble. Apuleius mentions the crude mattress shared by Lucius’s slaves when its relocation to a place in the courtyard farther from Lucius’s door signals that Photis has been preparing for an evening of erotic activity with the visitor. One of these slaves later arises from a heap of straw in the stable to beat a troublesome ass, whom he fails to recognize as his master, Lucius, transformed. Acts of the Apostles does not specify the extent of Mary’s slaveholdings, but we may infer that she does not have an opulent household. Rhoda does not seem to be exclusively a gatekeeper: she must come to the gate when she hears the knocking. By contrast, a grander household would feature a slave for whom gatekeeping was an exclusive duty. For example, in the Gospel of John a woman who seems only to serve as a gatekeeper guards the entrance to the high priest’s complex.3 The Acts of Paul includes a reference to the enslaved doorkeeper of Thecla’s household because he serves as a witness to her nocturnal comings and goings.4 Seneca remarks on a surly doorkeeper who expects visitors to drop a small coin in his hand as he lets them cross the threshold. The wise visitor placates the doorkeeper, says Seneca, “as one quiets a dog by tossing him food.”5 Gatekeepers, cup bearers, hairdressers, paedagōgi accompanying young masters through the streets to school—roles played by domestic slaves in antiquity may seem at odds with images of slaves’ work influenced by the paradigm of heavy slave labor on the cotton-producing plantations of the American South. Domestic slaves did not always contribute to the wealth of the household, while the costs of maintaining the household increased with every slave. (“The only servant he [the miserly Milo] feeds is one young girl,” says the innkeeper as she supplies the traveler Lucius with information about his prospective host.6) Reliance on slaves as a source of wealth or income, however, is only one dimension of the anthropology of slavery.7 Paul Bohannan, 3  John 18:16. 4  Acts of Paul 18–19. 5  Sen., Con. 14.12. 6  Apul., Met. 1.21. 7  For an overview of comparative anthropological perspectives on slavery, see Lovejoy, Transformations in Slavery, pp. 1–18.

Body Work: Slavery And The Pauline Churches

429

who defines slavery in terms of a servile antikinship relationship in which the slave is subject to sale, notes: “The content of the master-slave relationship may vary greatly. One or the other aspect may be emphasized: economic, domestic, religious, sexual, or whatever. Any attempt to classify systems of servility in terms of the economic obligations and positions of the slave is to assume that this one point provides an index for the rest, when in fact such a situation must be shown empirically to exist or not to exist.”8 Pauline Christianity was an urban phenomenon. The relationships of slavery with which Paul was acquainted would have been principally the relationships of urban slavery. Slaves in Corinth or Philippi would not have been miners or agricultural laborers but, for example, craftspeople, prostitutes, managerial agents, and domestic slaves, including those whose domestic duties included sexual obligations. How would a new identity as a Christian affect an urban slave, and how did the presence of slaves and slaveholders in the population affect the growth and practices of the churches? Ancient understandings of slaves as bodies will again inform my analysis. I situate the slaves and slaveholding households of the Pauline orbit in the context of the practices of slavery in the early Empire before I narrow my focus to a consideration of the (in) compatibility between enslaved and thus sexually vulnerable bodies and the strictures of purity demanded within the Christian body.

A Walk through the Streets of a Provincial City

Who were the slaves that Paul encountered in the course of his travels? Where and in what contexts did he encounter them? Acts of the Apostles includes one account of a slave whom Paul met in the streets of Philippi. The unnamed female slave (paidiskē) was possessed by a “Pythian” spirit, who spoke through her. The slave’s oracular powers were a source of income to her owners and a source of annoyance to Paul and his companions. The woman followed after them, calling out that they were “slaves of the most High God.” A first-century traveler like Paul would not have been surprised to run across a fortune-teller plying her trade in the street. Fortune-telling was a common phenomenon in antiquity. Scraps of papyri record the questions that unknown persons from antiquity posed to fortune-tellers: questions about love and marriage, trade, gambling, and childbirth.9 Residents of Mediterranean cities would routinely have 8  Bohannan, Social Anthropology, p. 181. 9  When Paul exorcises the Pythian spirit from the slave, her owners lose a ready source of income. Luke’s treatment of the slave’s oracular powers is ultimately more respectful than that

430

Glancy

chanced on fortune-tellers. In The Golden Ass, for example, Lucius tells Milo of the predictions he received from an itinerant fortune-teller, and Milo replies by sharing the story of his own encounter with the same seer.10 According to Acts of the Apostles, Paul responded to the enslaved fortune-teller by performing an exorcism, which rid her of the possessing spirit and thus deprived her owners of a sure source of revenue.11 The slave practiced her lucrative trade publicly, in the streets. Evidence that female slaves and other women of humble status moved freely in urban streets and squares modifies the scholarly generalization that public spaces were “male” whereas private spaces were “female.” In the Hellenistic Jewish narrative of Judith, for example, Judith’s confinement to her home at the beginning and end of the story establishes her status as a respectable free woman of considerable means. It seems natural, however, for her female slave to travel through the streets, with no apparent companion, to invite the elders to a meeting in Judith’s house. In an article on Jesus’ conversation with the Samaritan woman at the well (John 4), Jerome Neyrey argues that the woman’s presence in a public space, at noon, and engaged in conversation with a stranger, marked her as “deviant.” Neyrey’s assessment of the Samaritan woman as deviant because of her presence in a public place at midday can only be sustained if we label as deviant all women of lower statuses (slaves, freedwomen, poor freeborn laborers), a considerable percentage of the female population. To support his characterization of a cultural division between public/male and private/female spaces, Neyrey quotes Philo: “Market-places and council-halls and law-courts and gatherings and meetings where a large number of people are assembled, and open-air life with full scope for discussion and action—all these are suitable to men…. The women are best suited to the indoor life which never strays from the house, within which the middle door is taken by the maidens as their of some other ancient authors, since he implies that the slave really had been possessed by a spirit of divination, albeit a spirit less powerful than the name of Jesus. In contrast, for example, Apuleius expresses cynicism toward the activities of oracles. When Lucius expresses confidence in the predictions of a Chaldean astrologer, who he encountered in Corinth, his host, Milo, debunks the tale by revealing how the same man had been revealed a fraud in Hypata. A mark of Apuleius’s genius is that he leaves his ultimate attitude toward the oracle an open question; the astrologer’s prediction that Apuleius will write a book about his journey ultimately proves true, as does his prediction that the book will not be taken as a serious work. Regarding the ambivalence of The Golden Ass toward all things supernatural, from fortune-telling and magic to the worship of Isis, see Winkler, Auctor and Actor. 10  Apul., Met. 2.12–14. 11  Acts of the Apostles 16:16–18.

Body Work: Slavery And The Pauline Churches

431

boundary, and the outer door by those who have reached full womanhood.”12 Neyrey fails to note that Philo’s articulation of the division between male and female worlds was expressly marked by status considerations. According to Philo, the woman who confined herself to her home except when she went to temple and then chose the hours when the market would be quiet deserved the honorable name of “freeborn lady” (Greek, eleuthera). Other literary references that associate women with interior spaces reinforce the impression that gender segregation was a phenomenon based on status and class. In the Acts of Thomas, for example, Charisius asks his wife sadly, “Why did you not have regard to your position as a free woman [emphasis added] and remain in your house, but go out and listen to vain words?”13 Indeed, Philo’s commentary on the symbolic division between male and female spheres devolves into a condemnation of women who behave in unseemly ways in public spaces, specifically, women who argue and fight in the marketplace.14 Philo’s censure hinges on the everyday presence of women in markets and other public places—and many of the women buying, selling, bargaining, and fighting in the marketplace would have been slaves. As Paul traveled from city to city, then, he would have found it impossible to avoid contact with slaves. When he went to the marketplace to find other craftspeople or to purchase food for dinner he would have mingled with both male and female slaves. A wide variety of evidence attests to the ubiquitous presence of slaves in marketplaces. In the Life of Aesop, for example, the slaveholder Xanthus orders his slave Aesop to cook dinner for a gathering of his pupils. Not only does Aesop serve as cook, he also goes to the marketplace to buy the provisions for the meal. The other shoppers he encounters in the marketplace would have included male and female slaves, freedmen and freedwomen, as well as freeborn folk of the lower economic strata. The merchants would have been of equally modest status. Funeral epitaphs of slaves as well as freedmen and freedwomen list such occupations as fishmonger, salt vendor, 12  Philo, Leg. Spec. 3.169. Quoted in Neyrey, “What’s Wrong with This Picture?” 13  Acts of Thomas 89. 14  Corley also notes that the relegation of women to the private sphere is a phenomenon of the elite. Whether lower-status women considered their “freer” access to public spaces an advantage, as Corley implies, or whether they construed it as a lack of protection is a question that our sources do not permit us to answer (Private Women, Public Meals, pp. 15–16). Sailer views the question from both perspectives: “The slave woman’s lack of honor removed moral inhibitions to give her space for independent action, at the same time it denied her certain protections” (“Symbols of Gender and Status Hierarchies in the Roman Household,” p. 89). For more sustained treatment of women in private and public spheres see Torjesen, When Women Were Priests.

432

Glancy

and grain merchant.15 Kathleen E. Corley has argued that the saying of Jesus comparing his generation to children in the marketplace may refer more plausibly to slaves sent by their owners to the marketplace to look for work as entertainers at a banquet.16 “They are like children [paidiois, which Corley proposes translating as “slaves”] sitting in the marketplace and calling to one another, ‘We played the flute for you, and you did not dance.’”17 Paul would have encountered slaves in the homes of the men and women who offered him hospitality. Even before he entered those homes, however, he would have interacted with slaves as he made his way around urban streets throughout the Empire.

Working Bodies: Occupations of Slaves

Slaves could be found in every occupation in Greco-Roman cities.18 The men who maintained the furnaces in the baths were often slaves. The women who served beer in beer shops were often slaves and, for that matter, often prostitutes. Slaves worked in pottery factories and on farms, in mines and as shepherds. In smaller establishments a slave might have multiple jobs. Slaveholders provided official notifications of the deaths of slaves in order to avoid capitation taxes. These death notices typically referred to the slave’s trade. They sometimes stated the deceased slave was not skilled in any particular trade.19 The staffs of larger households included slaves who managed the accounts and oversaw other slaves as well as slaves who carried household waste to public dump sites. Evidence is extensive for the involvement of slaves in the production of commodities, where they typically worked alongside free laborers. For example, during the early Empire, Arezzo was a center for the manufacture of red-glazed pottery. In workshops ranging in size from a dozen to sixty workers, slaves crafted pottery, which they marked with their own names.20 Slaves were ubiquitous in all ranks of garment workers, from weavers to dyers to seamstresses. Papyri that listed the occupations of slaves often cited weaving as a trade. An early third-century guardian’s account of the financial status of his 15  CIL 6.9801, 9683; IG 11244. 16  Corley, Private Women, Public Meals, p. 128. 17  Luke 7:31–32a. 18  Joshel discusses the relation of work to identity among slaves, freedmen and freedwomen, and the freeborn poor (Work, Identity, and Legal Status at Rome). 19  E.g., P.Oxy. 7.1030, 41.2957; 49.3510. 20  Dyson, Community and Society in Roman Italy, pp. 115–116.

Body Work: Slavery And The Pauline Churches

433

wards, two minor boys, listed the wages earned by a female slave employed as a weaver. The text’s modern editor notes that the slave’s wages were sufficient to lift the account into the profitable range.21 A sheaf of apprenticeship contracts suggests that slaveholders often sent female slaves, less often male slaves, to learn the trade of weaving.22 Such contracts delineated responsibilities for feeding and clothing the enslaved apprentice and specified the holidays the slave would enjoy while working under the weaver. Benefits to the slaveholder were multiple. The slaveholder received remuneration for the slave’s labor during the time of the apprenticeship. At the same time, the slave became more valuable to the slaveholder as she became more skillful in the designated craft. Along with labor in workshops, fields, and markets, slaves advanced their owners’ financial ends through serving as financial agents and managers of all kinds. Ostraca document the activities of slaves who served as financial agents, often with some autonomy. At least in Egypt, Roman families living abroad, including members of the imperial family, were most likely to rely on slaves as agents.23 Jesus alluded in the parables to slaves who managed other slaves, some serving their owners more faithfully than others.24 As overseers, slaves could exercise considerable power over other slaves within the household, but even outside the structure of the household a slave could manage a slaveholder’s fortune. A well-documented example of a slave heavily involved in financial management comes from a villa in the vicinity of Pompeii. Wax tablets excavated there record the activities of a slave named Hesychus, who acted as his owner’s agent in loaning 10,000 sesterces to an importer of foodstuffs named C. Novius Eunus. Hesychus also coordinated the rental of extensive food storage facilities. Moreover, this trove of tablets reveals some of the material benefits that could accrue to a slave entrusted with financial affairs. Within months of expediting his master’s loan in 37 ce, Hesychus became a creditor in his own right through a loan of 3,000 sesterces to C. Novius Eunus.25 Only a minority of slaves had such lucrative opportunities. The overseer parables, for example, call attention to the (praiseworthy or culpable) activities of the slave entrusted with affairs of the household. The parables allude to a greater number of other household slaves, under the supervision of the overseer, who held less-responsible positions. The ascendancy of a Hesychus was thus neither a 21   P.Oxy. 58.3921. 22  A bibliography accompanies a translation of an apprenticeship contract (P.Mich. inv. 5191a) in Pearl, “Apprentice Contract.” 23  Biezunska-Malowist, L’Esclavage dans l’Égypte Gréco-Romaine Seconde Partie, pp. 99–101. 24  Matt. 24:45–51; Luke 12:41–48. 25  S. Casson, Ancient Trade and Society, pp. 104–110.

434

Glancy

norm for slaves nor an anomaly. (Chapter four, which analyzes the parabolic figure of the slave, examines at greater length the evidence regarding managerial slaves.) It is difficult to assign a single job title to capture the work obligations of a slave in a small slaveholding establishment. A second-century contract from Oxyrhynchus gives an idea of the range of duties a slave who belonged to a more humble master or mistress might have to perform. Glaukos leased his slave Tapontos, a weaver, to Achillas for a period of a year. The lease specified that Achillas would be responsible for Tapontos day and night. However, Glaukos retained the right to send for Tapontos during the night to make bread.26 Clearly, Glaukos’s establishment was not large enough to include a full-time baker, as one would expect to find in a wealthy household. Our sources typically distinguish between urban and rural slaves. The opening act of Plautus’s Mostellaria, for example, features banter between a country-bumpkin slave and a scheming city slave with a superior attitude. The distinction between urban slaves, belonging to the familia urbana, and rural slaves, belonging to the familia rustica, was predicated as much on the slaves’ duties as on their place of residence.27 Smaller slaveholding households would not have included sufficient staff to maintain this demarcation. The Gospel of Luke includes a parable of Jesus in which an agricultural slave doubles as a cook and domestic attendant: “Who among you would say to your slave who has just come in from plowing or tending sheep in the field, ‘Come here at once and take your place at the table’? Would you not rather say to him, ‘Prepare supper for me, put on your apron and serve me while I eat and drink.’”28 Jesus’ parable does not evoke a vision of life on a vast estate, such as the extensive landholdings that dominated the landscape in Sicily and the Italian peninsula, but a smaller slaveholding household in which a slave’s multiple and varied duties ranged from agricultural and pastoral tasks to food preparation and service. In larger slaveholding establishments, slave labor could be highly specialized. Clement of Alexandria, for example, disapproved of bloated household staffs, with some slaves “to prepare and make the pastries, others to make the honey cakes, and still others to prepare the porridges.”29 Josephus mentioned that Herod the Great was especially fond of three of his eunuchs, each of whom had extremely limited duties. One eunuch did nothing but pour

26   P.Wisc. 5 (186 ce). 27  Bradley, Slavery and Society at Rome, p. 58. 28  Luke 17:7–8b. 29  Clement of Alexandria, Paedagogues 4.26.

Body Work: Slavery And The Pauline Churches

435

Herod’s wine. Another served his food. A third eunuch, who shared Herod’s bedchamber, helped Herod as he prepared for bed.30 Domestic slaves probably outnumbered slaves engaged in productive or managerial work.31 The horrors of life in the mines or on the vast agricultural estates of Sicily in many ways dwarfed the indignities of life as a domestic slave. Nonetheless, Keith Hopkins has argued, “Roman literature abounds with stories of incidental cruelty to individual domestic slaves…. Domestic slaves stood in the front line. They were more privileged and pampered than the tens of thousands of slaves who labored without hope in the fields and mines…. But domestic slaves also had more contact with their owners, and were more often subjected to their despotic whimsy.”32 In smaller households the very slaves who contributed to the production of commodities would also have been involved in household tasks, including food preparation, cleaning, removing waste, and caring for children. Along with producing clothes for sale, a slave skilled in wool working could also produce clothing for members of the household. In larger households the tasks performed by domestic slaves would be highly specialized. Thus, on a continuum, the slave in the Lukan parable moved from work in the fields to waiting on the master at the table. Aesop not only shopped for food but also prepared and served it to his master and his master’s guests. A wealthier household would have multiple slaves dedicated to particular cooking specialties. In The Golden Ass, for example, Lucius the ass eventually falls into the hands of two kitchen slaves, one specializing in sweets and the other in sauces.33 Descriptions of banquets suggest that elaborate household staffs would enhance the self-image of some slaveholders. The more frivolous a slave’s task, the clearer the evidence of the owner’s wealth. In the Satyricon, Encolpus describes the scene as he enters Trimalchio’s banquet, where Trimalchio is playing a ball game with some boys: If the ball hit the ground, he [Trimalchio] didn’t chase it, but had a slave with a bag full of balls give the players a new one. We noticed some other novelties: there were two eunuchs stationed at different points in a circle; one was holding a silver chamber pot, the other was counting the balls…. 30  Joseph., BJ 1.488–489. 31  One basis for this assertion is the census returns preserved in Egypt: “Slaves in the returns are evidently for the most part domestic servants (only one slave has a declared occupation …)” (Bagnall and Frier, Demography of Roman Egypt, p. 49). 32  Hopkins, “Novel Evidence for Roman Slavery,” p. 7. 33  Apul., Met. 10.13.

436

Glancy

While we wondered at the extravagance of this, Menelaus ran up and said, “This is the guy who’s throwing the party! What you see is only the prelude to dinner.” As Menelaus spoke, Trimalchio snapped his finger as a signal to the eunuch to hold out the chamber pot for him as he continued to play. After emptying his bladder, he called for water for his hands, sprinkled it lightly on his fingers and then wiped them dry on the head of a young slave.34 Trimalchio has set up this scene as a tableau for his guests so they can witness the extent of his wealth. As Menelaus pointedly observes, “This is the guy who’s throwing the party!” Jesus and his followers encountered the slaves of opulent households on those occasions when their activities came to the attention of civil or religious authorities. For example, during the events surrounding the arrest of Jesus, he and his followers had dealings with slaves belonging to the grand household of the high priest. One of the high priest’s male slaves had an ear severed during the arrest of Jesus; several of the high priest’s female and male slaves accosted Peter to accuse him of accompanying Jesus.35 The Pauline letters and Acts of the Apostles include references to a number of households that included slaves.36 However, on the basis of these texts, we cannot reach firm conclusions about the size of those households nor the degree to which their slaves specialized in productive, managerial, domestic, or sexual duties. We have no evidence to suggest that Paul interacted with slaves or slaveholders in households as lavish as Trimalchio’s. Still, when he accepted hospitality from a slaveholder, domestic slaves would have tended to his needs, from washing his feet upon entering the household to preparing the food for communal meals.

Domestic Bodies: Family Life among Slaves

Our sources help us detect the shadowy ghosts of slaves proceeding through the streets of Mediterranean cities in the course of their varied duties or laboring in workshops and kitchens. The sources are less helpful when we try to reconstruct the domestic arrangements of slaves. What kind of quarters did slaves occupy? How did the layout of ancient houses facilitate or impede family ties among slaves? Moreover, how did residential layouts shape the quality of 34  Petr., Sat. 27. 35  Matt. 26:51, 69, 71; Mark 14:47, 66, 69; Luke 22:50, 56; John 18:10, 16–18. 36  E.g., 1 Cor. 1:11, 16:15; Philem.; Acts of the Apostles 10, 16:13–14, 16:27–34, 18:8.

Body Work: Slavery And The Pauline Churches

437

relationships among slaves who succeeded in establishing, at least for a time, stable family ties? Literary sources afford scant cues. Archaeological remains are in need of interpretation. Sometimes slaves lived outside the households of their owners. We have a glimpse into such living arrangements in a legal petition submitted in 60 ce, in which a slaveholder named Theon complained that thieves had twice broken into the living quarters of his slave Epicharis, who lived in an apartment of some kind in a house in another district.37 More often, slaves lived with their owners. Larger houses had separate slave quarters—dark, cramped, far less ornamented than the rooms in which the head of the household received guests or conducted business. Smaller households did not have separate accommodations for slaves. Regardless of whether the slaves had their own quarters, however, no corner of the household was off limits to slaves. Attending children, preparing food, working on accounts—the slaves, who were integral to every aspect of the household’s function, would have been found in every nook and cranny of a house during the hours of work.38 At night, a slave might have slept in a closet that was also used for food storage or curled up in a corner of an owner’s bedroom. The information we have about ancient slave quarters is, at best, minimal. Evidence so fragile cannot sustain generalizations about the quality of domestic life among those slaves fortunate enough to establish (at least temporarily) family bonds. Michele George notes, “The development of slave families is also problematic. In large households separate suites for slaves with considerable responsibilities may have been usual. However, in houses of average size this would have been impossible, and here slave families must have evolved despite proximity to the free household, and in a cultural context with notions of intimacy and personal space which were different from our own.”39 (It seems worth noting that if slaves lived with less privacy than families in modern industrialized societies, so too did slaveholders.) Envisioning the domestic configurations of a family of slaves owned by a modest slaveholding establishment is difficult. Did they enjoy common leisure time? Where did they congregate? Were they able to spend time with one another without intrusions from their owners or, for that matter, from other slaves in the household? The questions become still more complex when we try to imagine the domestic relations among a family of slaves with multiple owners or the domestic relations among a family that included freedpersons as well as slaves. 37   P.Oxy. 58.3916. 38  George, “Repopulating the Roman House.” 39  George, “Repopulating the Roman House,” p. 317.

438

Glancy

Our inability to envisage how slaves arranged their family time and space should not be construed as evidence of a failure on their part to do so. Dale Martin has reviewed funeral epitaphs from various regions in the Empire. The dedications of these epitaphs most commonly indicate that an immediate family member (spouse or child) of the deceased was responsible for the memorial. This held true for slaves and freedpersons as well as for freeborn persons. Martin encourages caution, however, in drawing conclusions based on the epitaphs. We cannot easily infer from epitaphs, which employ stereotyped formulas of affection and respect, the actual feelings that ancient family members had for one another. Most slaves from antiquity were not remembered in formal epitaphs, which suggests that the slaves so memorialized represented an atypical cross-section of ancient slaves. Nonetheless, the picture sketched by the epitaphs is remarkably consistent. At least in death and most likely in life, the slaves who dedicated epitaphs to one another valued their life partners and their offspring.40 We do not know when slave families made time for one another or how they managed to find a place for that time together, but epigraphic evidence suggests that many did so. Martin concludes that “these [epigraphic] studies illustrate the existence and importance of the immediate family for a significant minority of slaves.”41 Many more slaves, however, lived and labored under conditions that did not enable them to sustain stable family connections. For example, the sexual demands that slaveholders made on female and young male slaves may have strained family relationships among slaves themselves. Or perhaps not—slaves who cared about other slaves in demand as their owners’ sexual playthings did not leave diaries recording their emotional reactions to the intimate violation of their loved ones. Studying architectural remains of private dwellings in Italy and the Eastern Empire tells us only that slaves and slaveholders often shared remarkably close quarters. It does not tell us how the resultant lack of privacy impinged on the ability of slaves to construct their own lives and worlds of meaning.

The Head of the Body: The First Christian Slaveholders

Paul’s encounter with the slave possessed by an oracular spirit led to conflict with her owners rather than to an opportunity to evangelize her household, although Acts of the Apostles does include four accounts of household conversions. In each case, Peter or Paul gained entry through contact with the head of 40  Martin, Slavery as Salvation, pp. 2–7. 41  Martin, Slavery as Salvation, p. 7.

Body Work: Slavery And The Pauline Churches

439

the household. A dream of Cornelius led him to send two slaves and a soldier to find Peter (10:7). Peter decided to baptize the household when the holy spirit descended on the assembled members, including, presumably, those enslaved (10:44–48). When Paul encountered Lydia, the Lord opened her heart, which led not only to her own baptism but also to the baptism of her entire household (16:14–15). A jailer’s experience of God’s power in opening the doors of the prison precipitated his invitation to Paul and Silas to preach to his household and then to the household’s baptism (16:27–34). Brief mention of the conversion of the synagogue official Crispus notes that his entire household followed his lead (18:8). The historical accuracy of these summaries is difficult to assess. Luke’s theological emphasis is on the power of the spirit at work building the church. Luke supposed that his readers would find nothing amiss when a slaveholder determined the religious practices of the household. Indeed, even contemporary scholars evince little concern about the legitimacy of conversion and baptism of slaves in such circumstances.42 Scholars debate whether the household baptisms represented in Acts of the Apostles involved children, but they do not debate whether household baptisms included slaves. They assume and assert that this was the case. Moreover, they do not seem troubled by this assumption. For example, as James Dunn reviews the household conversions in Acts, he equivocates on the inclusion of children in household baptisms but not of adult slaves. Dunn writes of Lydia’s situation, “Household here need not include children since the term was commonly used to include household slaves and retainers.” Of the conversion of the jailer’s household, Dunn writes, “It is equally unclear whether household slaves and other adults alone [emphasis added] are in view or also children.” Finally, regarding the conversion of Crispus’s household, Dunn notes, it is “not clear whether a family is in view or simply the household slaves and retainers.”43 Like Dunn and other commentators on Acts, ancient readers would have understood these households to include slaves. Unlike these commentators, however, I think that household baptisms masterminded by slaveholders raise uncomfortable questions about the social dynamics within the Pauline churches, as well as questions about the reception of the gospel by those who participated in the ritual of baptism.

42  For example: “Incorporation into the Christian community as a consequence of social relationships rather than personal conviction is known to have been widespread in the early church. Household baptisms after the conversion of the oikodespotēs are reported in Acts … and alluded to by Paul” (Taylor, “Social Nature of Conversion in the Early Christian World,” p. 132). 43  Dunn, Acts of the Apostles, pp. 220, 223, and 243.

440

Glancy

Paul’s initial contact in Philippi, according to Acts of the Apostles, was a dealer in purple goods named Lydia. She was a godfearer, that is, a Gentile who worshiped the God of Israel but had not converted to Judaism. Her openness to Paul’s message, we have noted, led not only to her own baptism but also to the baptism of her household (Acts 16:13–15). Paul and his companions accepted her invitation to stay with her, so her house was large enough to accommodate visitors. Whether she was a wealthy woman or a relatively small business agent is not clear, further muddied by the difficulty of identifying social strata of ancient society in terms familiar to modern readers. She could have been a freedwoman who first entered the garment trade when she was a slave.44 Perhaps some of those who lived in her house were slave apprentices. Perhaps she owned her workers, or perhaps they were freedmen and freedwomen. Ivoni Richter Reimer has advanced the hypothesis that Lydia’s house was a “contrast society” in a Roman colony. As a female head of household, Lydia was not a pater familias. The citizen pater familias (Latin, “father of the family,” or more generally “head of household”) possessed, at least according to law, near total powers over members of the household, especially his offspring (including adult offspring) and slaves. Despite an earlier mention in Acts of the Apostles of a female Christian slaveholder, Mary (12:12–14), Reimer minimizes the likelihood that Lydia was a slaveholder, seeing her instead as an “independent woman … who provides shelter for other people in her house.” Even conceding that Lydia’s home could have included slaves, Reimer claims that Lydia’s home was an egalitarian retreat: “Because, for example, there is no longer a pater familias, there is no more patriarchal subordination, and all can be equal sisters and brothers.”45 A later papyrus from Karanis offers a glimpse into a household headed by a female garment worker, which hardly conforms to Reimer’s hypothesis that a female-headed household would constitute a contrast society. An apprenticeship contract for a female slave to learn the trade of weaving identified Aurelia Libouke as a weaver who had the right to act without a guardian because she had raised three children. Aurelia Libouke agreed to teach the slave the craft of weaving and promised that the slave would attain proficiency suitable for a girl her age. Unlike most apprenticeship contracts, which specified regular holidays throughout the apprenticeship period, this contract stipulated that the slave would be available for labor with no days off

44  The possibility is mentioned by Reimer, Women in the Acts of the Apostles, p. 108, and Witherington, Acts of the Apostles, p. 491. 45  Reimer, Women in the Acts of the Apostles, pp. 125–126. For discussion of the slaveholder Mary and the enslaved Rhoda, see Martin, “Acts of the Apostles,” pp. 783–784.

Body Work: Slavery And The Pauline Churches

441

for leisure or even illness.46 The mere presence of a woman such as Lydia as head of household did not, therefore, transform a household from a hierarchical to an egalitarian structure.47 Richard Sailer notes, “Though the head of the household was stereotyped as male by use of pater familias, in reality Roman women owned property and must often, in the absence of husbands, have wielded power over households with dependents. This gendered language causes historians to lose sight of female heads of households, even when they know better.”48 The baptism of her household was Lydia’s initiative. Luke does not suggest that the motivations or reactions of other household members concerned either Lydia or Paul. Acts of the Apostles does not present the Christian message as a challenge to slaveholding authority. Through the representations of household baptisms in Acts of the Apostles, Luke reinscribes the power of the head of the household over the lives of those in the house, including resident adult slaves. Chris Frilingos has argued that, in his letter to Philemon, Paul employed the vocabulary of family in such a way as to establish himself as the affectionate and commanding pater familias with authority over both slave and slaveholder. In this picture of Christian origins, the new Christian community unsettled existing hierarchical modes of relating among members of Christian households, even between slaveholders and slaves.49 In contrast, by suggesting that the spirit responded to the invitations of slaveholders, household by patriarchal household, Acts of the Apostles treats enslaved members of households as dependent bodies subject to the intellectual and spiritual authority of slaveholders.50 An authorial propensity to identify with the viewpoint of slaveholders rather than slaves may also influence two summaries of the exodus event in Acts of the Apostles that, curiously, lack any reference to the enslavement of the 46   P.Mich. inv. 5191a (271 ce), translated in Pearl, “Apprentice Contract.” 47  I have argued elsewhere for the necessity of a hermeneutics of suspicion in interpreting ancient representations of relations between slaveholding and enslaved women. See Glancy, “The Mistress-Slave Dialectic” and “Family Plots.” 48  Sailer notes that the “institution of slavery may have increased the independence of elite women” (“Pater Familias, Mater Familias, and the Gendered Semantics of the Roman Household,” p. 197). A female head of household would be known in Latin as a domina rather than as a mater familias, which connoted a chaste and respectable wife (Sailer, “Pater Familias, Mater Familias, and the Gendered Semantics of the Roman Household,” esp. p. 196). 49  Frilingos, “‘For My Child, Onesimus.’” 50  Drawing on postcolonial theory, Jill Gorman has begun an investigation of such “household coercions” (private correspondence).

442

Glancy

Israelite people. When Paul preaches at the synagogue in Antioch, he encapsulates the events surrounding the liberation of the Israelites without even a cryptic allusion to the coercive nature of their continuing sojourn in Egypt: “The God of this people Israel chose our ancestors and made the people great during their stay in the land of Egypt, and with uplifted arm he led them out of it” (13:17). In Stephen’s speech before the council, he represents the kings of Egypt as evil not because they enslaved the Israelites who had been their guests but because they promoted infanticide: A king who had not known Joseph came to power, declares Stephen, and he “dealt craftily with our race and forced our ancestors to abandon their infants so that they would die” (7:19). Although in the nineteenth century African-American abolitionists would celebrate the exodus story as a narrative of God’s historical intervention against slavery, these first-century synopses of the exodus expunge any memory of the enslavement of the Israelites. Paul identifies the Christian community as the body of Christ. Descriptions of household conversions in Acts of the Apostles suggest that slaveholders played a disproportionate role in the baptisms of their households and therefore a role in the Christian body that derived not from a gift of the spirit but from their secular status. In 1 Corinthians Paul refers to a message he received from “Chloe’s people,” members of Chloe’s household, probably her slaves.51 Since Paul expected his readers to be familiar with Chloe, he had no reason to specify whether she was herself a Christian. She may well have been. If she was a believer, Paul’s reference to “Chloe’s people” hints that Christian slaveholders had a higher profile within the church than enslaved members of their households who were also baptized, an impression entirely consistent with Luke’s depiction of the movement of the spirit in Acts of the Apostles. According to Acts, Peter and Paul gained access to households not through humble members of the householdchildren or slaves—but through their heads. Slaveholders chose baptism not only for themselves but also for others who belonged to them, including, presumably, enslaved adult members of the household. What was the position of these enslaved bodies within the Christian body? Given the circumstances of a household baptism, could slaves’ experiences as members of the Christian community obliterate the disadvantages of their slave status, even within the cult?52

51  1 Cor. 1:11. 52  Barclay, “Paul, Philemon, and the Dilemma of Christian Slave-Ownership.”

Body Work: Slavery And The Pauline Churches



443

Obstacles to Slaves’ Participation in Pauline Churches

Paul claimed that membership in the body of Christ dissolved barriers between slave and free.53 Largely on this basis, scholars have inferred that slave status did not entail any special liabilities for participation in the Christian community. However, because slaves were their masters’ sexual property, their obligations to their masters would at times have included actions defined as polluting or aberrant in the Christian body. Slaves whose owners were not members of the church would have been especially vulnerable, since their owners would not have been subject to the community’s censure. The argument of the remainder of this chapter has two stages. The first stage analyzes evidence concerning slaves and their lack of control over sexual activity in the ancient Mediterranean world. Ancient texts record acceptance of masters’ control over their slaves’ sexuality and establish a strong link between slavery and prostitution. The second stage of the argument analyzes Paul’s discourse on porneia, or sexual immorality, in 1 Thessalonians 4:3–8 and 1 Corinthians 5–7 in light of evidence concerning the sexual vulnerability of slaves. No New Testament text details the sexual obligations of slaves to their masters.54 Nonetheless, consideration of the sexual demands of first-century slavery requires that we revise, or at least qualify, our claims about the composition and practices of Pauline communities. Wayne Meeks’s assessment that “a Pauline congregation generally reflected a fair cross-section of urban society” has dominated recent reconstructions of Pauline circles.55 To assume that these congregations never encountered potential converts among the many urban slaves who numbered sexual activities among their duties would modify our picture of Pauline communities in an erratic manner. How Pauline communities responded to these slaves is another question. Either the community excluded slaves whose sexual behavior could not conform to the norms mandated within the Christian body, or the community tolerated the membership of some who did not confine their sexual activities to marriage. The first possibility challenges the assumption that slavery did not jeopardize the standing of individuals in the Christian community; the second possibility suggests that Pauline communities viewed some sexual activities as morally neutral. In 1 Thessalonians Paul advises as an antidote to porneia the acquisition and control of a vessel, skeuos. What he means by this counsel remains notably 53  1 Cor. 12:13; Gal. 3:28; cf. Col. 3:11. 54  Bartchy notes that “neither the sexual risks for slaves nor the related temptations for their owners are mentioned specifically in NT documents” (“Slavery, Greco-Roman,” p. 69). 55  Meeks, First Urban Christians, p. 73.

444

Glancy

obscure. Some commentators believe that he is urging (male) members of the Thessalonian congregation to acquire wives. Other commentators believe that he is urging the Thessalonian Christians to control their own bodies or, perhaps more specifically, their sexual organs. The Christian men of Thessalonika who first received this letter were accustomed to thinking of slaves as morally neutral sexual outlets. I suggest that we revisit Paul’s advice to the Christians in Thessalonika to ask how the first male recipients of this letter would have understood Paul’s instruction that each man should acquire a vessel for himself. We do not know the extent to which the Corinthian community adapted Paul’s teachings in 1 Corinthians. I argue, however, that his teachings on sexuality in 1 Corinthians 5–7 would have complicated and even barred the participation of many slaves in Christian life. The passage 1 Corinthians 5 establishes parameters for the discourse, raising the question of what constitutes porneia. Paul situates porneia as a problem not only for individuals engaged in such behavior but for the entire community. In 1 Corinthians 6:12–20, Paul explains why a Christian community cannot include a man who frequents prostitutes. Paul’s logic is predicated on the incompatibility of prostitution with the Christian life. Enslaved prostitutes who had no control over their owners’ decisions to profit from their bodies would thus have been excluded from membership in the Christian body. Finally, Paul’s insistence that sexual activity should be confined to marriage (1 Cor. 7) posed difficulties for slaves whose masters insisted on sexual relations with them. In Paul’s understanding, could the Christian body accommodate the sexually available bodies of slaves? That is, would the inclusion of Christians who were legally and culturally unable to protect the boundaries of their own bodies have threatened the maintenance of clear boundaries for the Christian body? Or does the presence of urban, domestic slaves and slaveholders among the congregations of the Pauline orbit require that we revise our estimations of the Pauline definition of porneia, sexual impurity?

Body Work: Sexual Availability of Slaves

Concerning the sexual availability of slaves, Moses I. Finley writes, “This is treated as a commonplace in Graeco-Roman literature … only modern writers have managed largely to ignore it, to the extent that the fundamental research remains to be done.”56 Modern scholars usually overlook the problem of whether slaves’ sexual availability affected their membership in the Christian community. The ancient church, however, explicitly considered this question 56  Finley, Ancient Slavery and Modern Ideology, pp. 95–96.

Body Work: Slavery And The Pauline Churches

445

by (at the latest) the third century. The Apostolic Tradition, usually attributed to Hippolytus, included in its discussion of baptism a list of those whose participation in the Christian community was proscribed or otherwise regulated.57 The situation of enslaved concubines received particular attention. Men who kept concubines could only join the community if they formalized those relationships through legitimate marriages. Enslaved concubines were welcome in the community if they had been faithful to their masters and if they had raised their children. (A complicating factor unrecognized by Hippolytus is that the owner rather than the enslaved concubine would have decided whether to rear or expose any child born to her.58) Contemporary scholars should exercise caution in asserting the relevance of a third-century document to the first-century church. What is striking is that an ancient Christian source articulated separate sexual standards for female slaves (who did not own nor control their own sexual activity) and free men (who did control their own sexual activity), thus calling attention to a moral conundrum intrinsic to the situation of enslaved Christians, which modern biblical scholarship has neglected. In Finley’s formulation, a chief characteristic of ancient slavery was that slaves were answerable with their bodies.59 Such answerability manifested itself in three major ways. First, corporal punishment was largely restricted to slaves, and slaves were in constant danger of corporal punishment. Although others, including sons and daughters of elite Roman families, were legally subject to beatings by the pater familias, corporal punishment was symbolically linked to servile status in Roman thought.60 Indeed, the nexus between slaves and the whip is a leitmotiv of the comedies of Plautus, as in this exchange between two slaves in Persa:

Sagaristo: Down at the mill I’ve been promoted To first sergeant in charge of flogging. Toxilus: Oh, you’re an old hand at that; you’ve earned your stripes.61

57  For attempts at reconstruction of The Apostolic Tradition, see Chadwick, Treatise on the Apostolic Tradition of St. Hippolytus of Rome; Botte, La Tradition Apostolique de Saint Hippolyte; and Hanssens, La Liturgie D’Hippolyte. Controversy surrounds the attribution of this text to Hippolytus in Rome in the early third century, especially since the document seems to have been more influential in the East than in the West. However, since I am only concerned to show that early Christians were more attuned to the complications of slavery than are modern scholars, the document’s exact provenance is irrelevant to the present discussion. 58  Rousselle, Porneia, p. 104. 59  Finley, Ancient Slavery and Modern Ideology, pp. 94–96. 60  Sailer, “Corporal Punishment, Authority, and Obedience in the Roman Household.” 61  Plaut., Persa 1.1.42–43.

446

Glancy

A second sense in which slaves were answerable with their bodies was the practice of requiring all testimony by slaves to be given under torture. Finally, Finley observed, “The third, qualitatively different and ubiquitous, manifestation of the answerability of slaves with their bodies, [was] their unrestricted availability in sexual relations…. Prostitution is only one aspect. More interesting in the present context is the direct sexual exploitation of slaves by their masters and the latter’s family and friends.”62 Finley adduced as evidence an aphorism of the elder Seneca, which referred to the moral position of the receptive partner in male homosexual encounters: “Unchastity [impudicitia] is a crime in the freeborn, a necessity for the slave, a duty [officium] for the freedman.”63 While the (im)moral evaluation implicit in the saying is satirical, its presupposition of the availability of slaves for sexual service is germane. Enslaved girls, women, boys, and young men were frequently sexual targets for their masters. Moreover, as Susan Treggiari writes, “The assumption seems implicit in Roman society that intercourse with a slave, who had no moral responsibility and no choice, was morally neutral for the free initiator.”64 Freeborn Roman men typically married in their late twenties or early thirties.65 Since Roman society placed little value on men’s sexual abstinence, young men often satisfied their sexual desires with household slaves or prostitutes (who were typically slaves). From the man’s perspective, one important benefit of such sexual encounters would be that any children conceived in those circumstances would not compete with legitimate children, conceived later, for claims to the estate. For the same reason, many Roman men chose sexual partners who were slaves or freedwomen after their wives’ deaths.66 Enslaved women were additionally vulnerable to retaliation from wives who knew or suspected their husbands’ sexual involvements with household slaves. While some women may have appreciated their husbands’ circumspection in seeking other sexual outlets, a variety of evidence suggests that wives often reacted negatively to slaves who attracted such attention. Sexual jealousy might prompt a wife’s resentful notice. A wife might also be angry if she thought her husband had squandered household resources to purchase a sexual partner or to pay for the partner’s upkeep. Annalisa Rei notes of the comedies of Plautus, “Characteristically, a wife takes action to avenge not a husband’s infidelity, but his violation of the property which the couple shares or which even belongs to 62  Finley, Ancient Slavery and Modern Ideology, pp. 95–96. 63  Sen., Controv. 4 praef. 10. 64  Treggiari, Roman Marriage, p. 301. 65  Sailer, “Men’s Age at Marriage and Its Consequences in the Roman Family.” 66  Sailer, “Slavery and the Roman Family,” pp. 68–72.

Body Work: Slavery And The Pauline Churches

447

the wife entirely. The pursuit of courtesans, both by married and by unmarried men, was not considered immoral in Roman society.”67 The mistress of a household had a variety of means at her disposal for harassing slaves, including physical abuse. In Genesis Rabbah, several rabbis commented on the scriptural observation that “Sarai dealt harshly” with her slave Hagar (Gen. 16:6). Rabbi Berekiah speculated that Sarai (Sarah) attempted to humiliate Hagar by requiring her to carry towels and water buckets for bathing or that Sarai slapped Hagar’s face with a shoe.68 Galen claimed that his mother used to bite female slaves when she was angry with them; Augustine recorded an incident in which his paternal grandmother ordered the household slaves to be beaten because she perceived them to be disloyal to her daughter-in-law, Augustine’s mother, Monica.69 Given the ancient stereotype that women are less able than men to control their passions, writers may have exaggerated the violence of slaveholding women against their slaves.70 Still, we should guard against the equally distorting assumption that a slaveholding woman and a woman she owned would have naturally presented a joint front because of their shared gender identity. Potentially faced with retaliation by their mistresses, many slaves might have preferred to avoid sex with their masters. The choice, however, was not theirs to make. In the world constructed by the romances, noble young women who were enslaved in unfortunate circumstances were typically able to resist the sexual overtures of their owners. Still, the threat of sexual compromise was constantly present for them. In Chariton’s Chaereas and Callirhoe,71 for example, the title characters consummate their marriage before being torn apart. Callirhoe finds herself the slave of Dionysius, a good man who is desperate to sleep with her. She manages to resist until she realizes that she is pregnant, whereupon she sends a message to Dionysius that she is willing to be his wife but not his concubine. Dionysius joyously accepts. Callirhoe convinces her new husband that the child she is carrying is his. What rings true about the episode is the master’s desire for an attractive young slave. However, Dionysius’s deference to Callirhoe’s sexual reluctance is implausible, as is his acceptance of her 67  Rei, “Villains, Wives, and Slaves in the Comedies of Plautus,” p. 104. 68  Neusner, Genesis Rabbah, vol. 2, p. 152. 69  Galen cited in Hopkins, “Novel Evidence for Roman Slavery,” p. 9; August., Confessions 9.9.20. See my discussion of Augustine’s grandmother and her slaves (Glancy, “The Mistress-Slave Dialectic,” pp. 1–2). 70  Clark, “Women, Slaves, and the Hierarchies of Domestic Violence.” 71  Hock describes Chaereas and Callirhoe as “the most contemporary document we have for Paul’s world in the Greek East of the first century” (“Support for His Old Age,” p. 68).

448

Glancy

marriage proposal. Livy’s version of the story of Verginia assumes the inability of a woman with the legal status of a slave to protect the integrity of her own body. In Livy’s version, Appius Claudius has Verginia legally declared a slave so that her father cannot defend her chastity. Elaine Fantham writes, “It is telling that Verginius should protest in terms of his daughter’s status, ‘his daughter would have been dearer to him than life itself if she had been allowed to live free and chaste; but … he saw her dragged off to be raped like a slave girl.’”72 While the romances drew attention to the sexual vulnerability of their enslaved heroines and heroes, the genre’s insistence on happy endings necessitated that those heroines and heroes should preserve their chastity through utterly implausible stratagems. Slaves did not have the legal right nor cultural power to say “no” to their owners’ sexual demands. In her dissertation on women in Greek novels, Brigitte M. Egger notes that although the preservation of the sexual integrity of enslaved protagonists is a recurring theme, the romances do not evince concern for the sexual integrity of other enslaved characters.73 Masters in the romances take for granted that their slaves are their sexual property. The excitement of the stories builds on the possibility that noble young women separated from their protective environments will be treated in a way considered appropriate for common household slaves. So too on the stage. In Plautus’s Persa, a young woman who is caught in a scheme of her father’s that involves selling her into slavery in order to reveal that the slave dealer is trafficking in freeborn “merchandise” worries that even a temporary enslavement will deter any decent man from later marrying her. No matter how brief her period of (pseudo)enslavement, for a short time she will not be in a legal or cultural position to guard her chastity.74 In Plautus’s Curculio the slave Palinurus responds with alarm to his master’s announcement of plans to seduce a young woman until Phaedromus, his master, specifies that the object of his lust is a slave rather than a proper young woman. Palinurus proclaims, “No one bans or prohibits you from buying what’s for sale from this place, if you have the cash. No one prohibits a man from using the public street, so long as you don’t cut a path through an enclosed property: so long as you keep off wives, widows and virgins, young men and boys of free birth, make love to anyone you choose.”75

72  Fantham, “Stuprum,” p. 276. 73  Egger, “Women in the Greek Novel,” p. 167. 74  See discussion in Sailer, “Social Dynamics of Consent to Marriage and Sexual Relations.” 75  Plaut., Curc. 32–37. Translation by Fantham in “Stuprum,” p. 274.

Body Work: Slavery And The Pauline Churches

449

The frequency of manumission inscriptions that imply that the freed slave was the master’s sexual partner indicates the prevalence of such liaisons in Greco-Roman societies.76 One manumission inscription from Delphi in the late first century bce illustrates the asymmetry in power that persisted even after a master freed his concubine. Kleomantis manumitted Eisias, the mother of his sons. However, he stipulated that she had to remain with him his entire life, “doing everything that is ordered like a slave…. If Eisias does not remain or does not do what is ordered, let Kleomantis have the power to punish her in whatever way he wishes, by beating her and selling her.”77 Far from the respect accorded the heroines of the romances, many slaves (and even former slaves) would have encountered physical abuse and the threat of being sold should they resist their masters’ advances.78 Resistance of male or female slaves to sexual overtures sanctioned by their owners was not acceptable servile behavior. A scrap of papyrus from Oxyrhynchus records a crude proposition or, more accurately, a threat from two males to a third. It reads: “Apion and Epimas proclaim to their best-loved Epaphroditus that if you allow us to bugger you it will go well for you, and we will not thrash you any longer.” Epaphroditus typically appears in the papyri as a name associated with a slave. Here, a young slave, or perhaps a freedman, seems to have two “options”: to submit to unwanted sexual activities or to allow two other men to (continue to) beat him.79 The bluntness of the message highlights the implausibility that a slaveholder such as Chariton’s Dionysius would respect a slave’s privacy and sexual integrity. On the other hand, for every Epaphroditus who resisted such sexual advances and perhaps even saw them as rape, there may have been another slave who accepted sexual overtures with equanimity. For some slaves, cohabitation with a master or, very occasionally, a mistress was certainly a route to what we would recognize as upward social mobility.80 Our sources are silent regarding the actual reactions of slaves to their sexual obligations. The silence is revealing. Ancient societies were not concerned with the reactions of slaves to sexual coercion, since the slave had no legal or cultural right to say “no.”81 76  Tucker, “Women in the Manumission Inscriptions at Delphi,” p. 231. 77  Tucker, “Women in the Manumission Inscriptions at Delphi,” p. 230. 78  Degradation and alienation typified the ancient Mediterranean slave trade. See Bradley, “‘The Regular, Daily Traffic in Slaves.’” 79   P.Oxy. 42.3070, discussed in Montserrat, Sex and Society in Greco-Roman Egypt, pp. 136–138. 80  Grubbs, Law and Family in Late Antiquity, p. 28 and chap. 6. 81  Bradley, Slaves and Masters in the Roman Empire, p. 118.

450

Glancy

Bodies for Rent: Slavery and Prostitution

Just as the typical slave had little opportunity to resist a master’s sexual advances, prostitutes who were slaves had little if anything to say about the work their owners required of them. A suggestion by Margaret Y. MacDonald that Christianity represented an economic threat to slaveholding pimps would require enslaved prostitutes to be in a position to refuse their assigned labor.82 This is highly improbable. By the late second century, Tertullian even complained that a persecution of Christians had extended to forcing a Christian woman into prostitution, the first known instance of a practice that lasted until the fourth century.83 Aline Rousselle asks, “Can we say that slaves and prostitutes enjoyed sexual freedom in this sense when we know that the leno, or procurer, took them to the ports when he heard that a ship had arrived? We should not imagine that all the young girls who were abandoned and then brought up to be slaves or prostitutes or both, and later freed by a procurer, were demi-mondaines.”84 Roman law and custom suggested that female prostitutes enjoyed a sexual freedom parallel to that of men, but such freedom was more apparent than real. Prostitutes did not choose their sexual partners; they were chosen as sexual partners. In his work on prostitution in ancient Rome, Thomas A. J. McGinn concludes, “This is a reflection of the unequal distribution of power implicit in the exchange of sex for money, an inequality especially notable when the woman [or the male prostitute] is a slave.”85 Roman law recognized the validity of certain restrictive covenants in the sale of slaves. One of those covenants, ne serva prostituator, forbade the buyer (including subsequent buyers) from forcing the slave into prostitution. A buyer who violated a ne serva prostituator covenant faced several potential penalties. In some instances the seller would reclaim ownership of the slave. In other instances the buyer would be forced to pay a financial penalty to the seller, who presumably had taken a financial loss in selling the slave under a restrictive covenant. In still other instances the slave would be manumitted as a Junian Latin, without benefit of citizenship. Although the ne serva prostituator covenant benefited the slave, the law was essentially concerned with the rights of the original vendor.86 Why would a slaveholder impose the ne serva prostituator covenant? Slaveholders may have had widely divergent motivations. 82  MacDonald, Early Christian Women and Pagan Opinion, p. 55. 83  McGinn, Prostitution, Sexuality, and the Law in Ancient Rome, p. 310. 84  Rousselle, Porneia, p. 94. 85  McGinn, Prostitution, Sexuality, and the Law in Ancient Rome, p. 210. 86  McGinn, Prostitution, Sexuality, and the Law in Ancient Rome, chap. 8.

Body Work: Slavery And The Pauline Churches

451

A promise that a slave would never be prostituted could function as either an incentive or a reward for hard work. A slaveholder who sold his lover could be reluctant for that person to become a prostitute. A slaveholder who sold a son or daughter might have some qualms about the prostitution of his progeny. A slaveholder who sold a slave who had worked as an intimate attendant for his wife might feel that the slave’s later work as a prostitute could taint the household. Some slaveholders may have been genuinely concerned about the well-being of slaves with whom they had interacted on a daily basis. All of this, however, is speculation, since slaveholders did not record their reasons for imposing the covenant. The very existence of such covenants highlights Roman awareness both that many slaveholders compelled their slaves to work as prostitutes and that slaves were not themselves in a position to resist such work. Dio Chrysostom described newly enslaved women forced to work as prostitutes who “feel shame and revulsion.”87 Slaves would have been aware that the decision whether or not to prostitute them belonged to the master. McGinn points to the practice of forcing Christian women into brothels as evidence that compulsion to work as a prostitute loomed as a potential punishment for slaves: “If some slave women might be rewarded for their loyal service through exemption from forced prostitution through the imposition of the covenant, others might be punished for their failings through installation in a brothel. We have a glimmer of what might be described as the [ne serva prostituator] covenant’s ‘evil twin’: a practice perhaps more widespread as a private form of punishment than the sources allow us to see.”88 In the Hellenistic world and later during the Roman Empire, exposed infants constituted one supply of enslaved prostitutes.89 Clement of Alexandria and Justin Martyr warned against frequenting prostitutes for fear of committing incest with a daughter one had abandoned in infancy.90 Slave dealers had a stake in promoting trade in prostitutes. Prostitutes commanded a higher sale price than other female slaves.91 Indeed, slave dealers moved easily between the role of leno (pimp or procurer) and auctioneer of female flesh.92 In the comedies of Plautus, for example, the figure of the leno schemes alternately to 87  Dio Chrys., Or. 7.134. 88  McGinn, Prostitution, Sexuality, and the Law in Ancient Rome, p. 310. 89  Pomeroy, Goddesses, Whores, Wives, and Slaves, pp. 140, 192. 90  Justin Martyr, 1 Apology 27; Clement of Alexandria, Paedagogues 3.3.21. 91  Based on records from Egypt in the third century bce. See Pomeroy, Women in Hellenistic Egypt from Alexander to Cleopatra, p. 130. 92  Pomeroy, Women in Hellenistic Egypt from Alexander to Cleopatra, p. 130.

452

Glancy

make money as a pimp and to strike it rich by selling a few prized slave prostitutes to men infatuated with them. Although Ulpian extended the definition of pimp to include someone who maintained free prostitutes, he primarily defined a pimp as a man who dealt in the services of enslaved prostitutes. McGinn writes, “The pimp who keeps slave prostitutes is emphasized either because this type was more significant as a socio-economic phenomenon or because a special legal reason (not evident to us) required this.” However, he regards the former reason—the high proportion of slaves in the population of prostitutes—as more likely.93 Although contemporary scholars do not have the data necessary to calculate what percentage of prostitutes were slaves, all available evidence suggests a strong correlation between prostitution and slavery.94 Whether most prostitutes were slaves is a more difficult question. McGinn expresses agnosticism on this point but adds, “It is safe to say that most prostitutes in the provinces were peregrines or slaves before A.D. 212.”95 Dominic Montserrat suggests that the documentary evidence is likely to downplay the percentage of slave prostitutes: “Since their ‘bodies’ were entirely at the disposal of their masters, a slave forced to work as a prostitute would need only to be recorded by social rather than professional status—the potential status as prostitute (or dancer or wet-nurse or wool-worker) was implicit in the designation as slave.”96 In her study of female prostitution in the Roman Empire, Rebecca Fleming emphasizes the substantial intersection between slaves and prostitutes. She characterizes the institution of prostitution in the Roman world as “part of the slave economy, so though the labor is hers the profits are not.”97 Artemidorus’s dream logic provides a thoroughly nonscientific confirmation that ancient writers expected prostitutes to be slaves: “A prostitute dreamt that she had entered the holy place of Artemis. She was freed and gave up her prostitution. For she would not enter the temple unless she were to abandon her profession.”98 And she could not abandon her profession as prostitute while she still belonged to another. 93  McGinn, Prostitution, Sexuality, and the Law in Ancient Rome, p. 254, n. 284. 94  Furthermore, “while the question of the extent to which Roman prostitutes found themselves in a condition of slavery or subservience cannot be addressed here, it is certainly true that many prostitutes, whether slave or free, would not have been able to decide how much to charge, how hard to work, or whether to remain a prostitute” (McGinn, “Taxation of Roman Prostitutes.” p. 91). 95  McGinn, Prostitution, Sexuality, and the Law in Ancient Rome, p. 16, n. 80. 96  Montserrat, Sex and Society in Greco-Roman Egypt, p. 107. 97  Fleming, “Quae Corpore Quaestum Facit,” p. 50. 98  Artemidorus, Oneirocritica 4.4.

Body Work: Slavery And The Pauline Churches

453

Although I am stressing the evidence that links prostitution and slavery, I am not thereby implying that free persons who became prostitutes willingly chose the occupation with any regularity. The documentary evidence from Greco-Roman Egypt suggests that free persons who engaged in prostitution often entered the trade under duress, through dire economic necessity. An indigent woman would not typically establish herself as a prostitute. Rather, another family member, often a father, would set her up as a prostitute because of the family’s destitution. Fleming writes, “Female bodies clearly counted among the economic resources, not only of slave-dealers and owners, but also of any family network.”99 Not until the rise of Coptic Christian literature is there any intimation that women enter the ranks of prostitutes to satiate their own lusts.100 This literature, we may safely surmise, reflects the preoccupations of its authors rather than the erotic appetites of prostitutes. In 40 ce, Caligula instituted a tax on prostitutes. This was not a “vice tax,” designed with the ulterior motive of curbing the widespread practice of prostitution. The goal of the tax was to raise revenue. The tax was collected throughout the Empire. In Egypt, the levy on prostitutes was ten times higher than other occupational taxes, suggesting that prostitutes (or their owners or brothel keepers) had sufficient income to pay a very high tax.101 Prostitutes and prostitution were thus common and visible throughout Greco-Roman cities in the first century. Through collection of the tax on prostitutes, the emperor directly profited from the trade.102 Furthermore, even the tax code recognized that many prostitutes were slaves: in some locales, records specify how to collect the tax from enslaved prostitutes and their owners. Popular literature of the ancient Mediterranean often features the character of the enslaved prostitute. Fleming observes that “with one partial exception, slaves fill all literary brothels where the status of their denizens is clear.”103 In An Ephesian Tale, the heroine, Anthia, is sold as a prostitute and forced to display herself outside the brothel. She manages, improbably, to retain her virginity by pretending to have epileptic seizures.104 The enslaved prostitute is a familiar figure in the Plautine corpus. The pseudo-meretrix, a free woman of good reputation whose virtue prevails despite her temporary enslavement, is a 99  Fleming, “Quae Corpore Quaestum Facit,” p. 42. 100  Montserrat, Sex and Society in Greco-Roman Egypt, pp. 107–109. 101  McGinn, “Taxation of Roman Prostitutes.” For qualifications to McGinn’s argument, see Bagnall, “A Trick a Day to Keep the Tax Man at Bay?” 102  Fleming, “Quae Corpore Quaestum Facit,” p. 56. 103  Fleming, “Quae Corpore Quaestum Facit,” p. 43, n. 19. 104  Xenophon of Ephesus, An Ephesian Tale 5.7.

454

Glancy

stock character.105 In several plays (Rudens, Poenulus), a freeborn female is rescued by her long-lost father just before her sale by a leno. In another plot line, a clever slave arranges the sale of a woman already enslaved as a prostitute to a man who cares about her (Pseudolus). These plots converge in Persa, in which a conspiracy entraps Dordalus, a leno, into buying a young freeborn woman. When her identity is revealed, she is freed and Dordalus ruined. Through the fraudulent sale, a slave named Toxilus acquires the money he needs to buy from Dordalus the prostitute Lemniselenis, who he desires. Although the formulas of the romances and comedies dictate happy resolutions to these stories, the vast majority of enslaved prostitutes in the Greco-Roman world could not reasonably expect to retain nor regain control over their own sexual practices. In general, Greco-Roman societies expected that men would frequent prostitutes and did not morally condemn them for doing so. Susan Treggiari writes, “Objections to intercourse with a common prostitute were aesthetic rather than moral or prudential.”106 Nor was moral condemnation of prostitutes themselves common, although prostitutes did occupy a position in society distinct from that of respectable matrons. In a culture whose definition of honor emphasized the preservation of a woman’s sexual integrity and a man’s sexual inviolability, prostitutes had no honor to preserve. Nonetheless, many in the ancient world perceived prostitution as pivotal to the smooth operation of the honor-shame system. By diverting the sexual energies of men away from those whose honor mattered, prostitution protected the honor of respectable families.107 The Greek Anthology includes an epitaph attributed to Marcus Argentarius: “Psyllus, who used to take to the pleasant banquets of the young men the venal ladies that they desired … who earned a disgraceful wage by dealing in human flesh, lies here. But cast not those stones…. Spare him, not because he was content to gain his living so, but because as keeper of common women he dissuaded young men from adultery.”108 Dio Chrysostom, the orator and philosopher who synthesized Stoic and Cynic ideas, offered a rare condemnation of prostitution. In doing so, he claimed that many of his contemporaries espoused the view that prostitution was a “wonderful elixir to produce chastity in our cities, your motive to keep those open and unbarred brothels from contaminating your barred

105  Rei, “Villains, Wives, and Slaves in the Comedies of Plautus,” p. 95. 106  Treggiari, Roman Marriage, p. 224. 107  McGinn, Prostitution, Sexuality, and the Law in Ancient Rome, p. 17. 108   Anth. Graec. 7.403.

Body Work: Slavery And The Pauline Churches

455

homes and inner chambers, and keep men who practice their excesses abroad and openly at little cost from turning to your free-born [eleuthera] and respected wives!”109 Acceptance of prostitution as an inevitable support for the honorshame rubric was not universal, but many were willing for enslaved women and men to pay with their bodies the price of the chastity of freeborn wives, daughters, and sons. Traditional Roman religion ritualized the division between prostitutes and other women. Prostitutes claimed a number of Roman holidays as their own.110 April 23, the Vinalia, was the first of two wine festivals; prostitutes celebrated by sacrificing to Venus at her temple outside the Colline Gate. The Floralia (or Ludi Florae) on April 27, dedicated to Flora, the goddess of flowers and fertility, was another festival for prostitutes. More striking perhaps were the celebrations of April 1, the Veneralia, when prostitutes and respectable matrons separately worshiped the forces of love. Prostitutes (and possibly other base-born women) went to the men’s baths to worship Fortuna Virilis, while respectable matrons worshiped Venus Verticordia. As Sarah Pomeroy summarizes these Roman ritual practices, “Thus the dichotomy between respectable women and whores was dramatized: the former worshipping an apotheosis of conjugal ideals, the latter worshipping sexual relationships having nothing to do with wedlock.”111 These holidays were associated with the city of Rome and were not celebrated elsewhere in the Empire. They are therefore of limited relevance for understanding the structure of religious life in the cities of the Greek East where Paul evangelized. I mention these holidays because they dramatize social (not moral) distinctions between prostitutes and respectable matrons, which pertain throughout the Empire. Roman cultic practice offers a narrow window into the complexity of social relations in ancient Mediterranean societies. During the Matralia, from which enslaved women were generally barred, freeborn matrons ritually beat the single slave woman ushered into the assembly. Scholars of early Christianity often assume the gospel message broke down walls between slave and free. In what follows we consider the challenge that the sexual availability of enslaved bodies poses to that assumption.

109  Dio Chrys., Or. 7.140. 110  For a survey of traditional Roman festivals, see Scullard, Festivals and Ceremonies of the Roman Republic. 111  Pomeroy, Goddesses, Whores, Wives, and Slaves, pp. 208–209.

456

Glancy

Enslaved Bodies and the Body of Christ

Paul was active in the mid-first century in cities of the Eastern Empire, although he also had contacts in the Western Empire, even in Rome itself, where he spent his final days. Those attracted to his preaching included male and female slaves, freedmen and freedwomen, and freeborn men and women, some of whom were slaveholders. Paul insisted that authority within the congregation was a gift of the spirit, given for the building up of community. Nonetheless, as we have seen, Acts of the Apostles records several stories in which Paul’s entry into a household occurred when he had contact with the head of the household. Corresponding stories in which Paul’s entry into a household occurred through contact with a subordinate member of the familia (Latin, “family,” implying not only biological kin but also slaves and freedpersons attached to the family unit) do not appear in Acts. How accurately Acts of the Apostles represents the history of the early Pauline communities is difficult to ascertain. As these communities struggled to define themselves as one body in Christ, Paul specified the limits of behavior, sexual and otherwise, compatible with membership in that body. Given the ubiquity of the sexual use of slaves, Paul would inevitably have encountered slaves whose obligations included sexual relations with their owners and those to whom their owners permitted sexual access, including enslaved prostitutes. In this section I examine Paul’s advice on sexual matters in light of these wider contexts. I argue that recognition of the somatic obligations of ancient slaves leads us to revise or modify commonly held positions in Pauline studies. We do not have sufficient evidence to determine whether the sexual obligations of slaves were an obstacle to their participation in the Christian community, or whether, like others in the first century, Paul and the churches regarded some sexual activity as morally neutral. Two interrelated questions structure the inquiry. First, how does a recognition that slaves were treated as available bodies affect our interpretation of Paul’s instructions on sexual ethics? Second, in light of what we have learned about the sexual use of slaves, how does Paul’s discourse on porneia, or sexual immorality, affect our reconstruction of primitive Christianity as a social movement, especially among slaves? In the late third century, Lactantius proclaimed that Christian men and women should adhere to the same sexual standards. Christian men should not have sexual partners other than their wives, not even slaves. Such an idea was not utterly unprecedented. The Stoic Musonius Rufus, for example, had argued that by engaging in sexual activities with slaves, freeborn men exhibited less self-control than women, who avoided such entanglements. He therefore

Body Work: Slavery And The Pauline Churches

457

thought that men should refrain from sexual relations with slaves.112 However, no evidence suggests that this Stoic advice had an appreciable impact on behavior in the Roman world. In contrast, later Christian writers repeated versions of Lactantius’s mandate, although they often acknowledged that the reality of people’s behavior did not always conform to these scrupulous standards. Ambrose urged men to avoid sexual relations with slaves in order to avoid giving their wives grounds for divorce. Jerome contrasted Christian with pagan morals by alluding to the tolerance that the latter exhibited for liaisons between free men and their female slaves.113 Jerome wrote, “Among the Romans men’s chastity goes unchecked; seduction and adultery are condemned, but free permission is given to lust to range the brothels and to have slave girls, as though it were a person’s rank and not the sexual pleasure that constituted the offense. With us what is unlawful for women is equally unlawful for men.”114 Does this view of sexual ethics arise over the course of several centuries’ development of Christian thought? Or is this view of sexual ethics implicit in the earliest Christian moralizing on sexual matters as preserved in the letters of Paul? Modern commentators have assumed that Paul found the sexual use of slaves immoral. However, unlike Lactantius, Ambrose, and Jerome, Paul does not refer explicitly to sexual relations between slaveholders and their human property. Given the ubiquity of the assumption that slaves were sexually available to their owners with no moral consequences for the slaveholder, we cannot simply assume that Paul’s audience implicitly understood that participation in the body of Christ precluded such sexual activity. We must return to the Pauline texts themselves. We must see what they say and, perhaps more important, what they do not say. In several places, Paul exhorts newly formed Christians to abstain from porneia. To assume that this instruction includes a tacit prohibition on (consensual or nonconsensual) sexual contact between slaveholder and slave begs the question. The question precisely at hand is whether Paul understands such relations to constitute porneia. Is Paul silent on the question because the sexual use of slaves is abhorrent to him and he expects other Christians to share his perspective? Or is Paul silent on this question because he does not challenge cultural norms regarding the sexual use of slaves? 112  Musonius Rufus, frag. 12. See discussion in Geytenbeck, Musonius Rufus and Greek Diatribe, pp. 71–77. 113  Grubbs, Law and Family in Late Antiquity, pp. 90–91, 233–249. 114  Jerome, Ep. 77.3.

458

Glancy

Musonius Rufus, Lactantius, Ambrose and Jerome all address the moral position of the slaveholder. What of the moral position of the slave? Sensitivity to the moral position of persons who are not in a position to say “no,” who are unable to maintain their bodies as sexually inviolate, invites us to reconsider the consequences of Pauline sexual ethics for slaves who sought membership and full participation in the Christian body. If indeed Christian communities understood Paul’s exhortations on sexual matters to preclude carnal contact between slaveholders and slaves, consequences would have differed for those in a position to control their own bodies and those not in a position to protect the integrity of their own bodies. Paul insisted that porneia was not primarily a matter of individual conscience. He argued that in order to avoid pollution the Christian community should expel any person who engaged in porneia. If we conclude that Paul included carnal relations between slaveholders and slaves under the umbrella of porneia, slaves whose owners insisted on using them sexually could be seen as a source of potential contamination in the pure Christian body. Under these circumstances, a slave owned by a non-Christian would be at greater risk than a slave owned by a Christian. A non-Christian slaveholder would have found a request that he abstain from using a slave sexually to be laughable. Still more tenuous would be the position of an enslaved prostitute who heard the gospel preached and sought baptism and communion with a Christian community. The business of a prostitute (Greek, pornē) was porneia. In 1 Corinthians, Paul reacts with horror to the presence in the Christian community of a man who had frequented a prostitute. It seems unlikely that he would have responded more warmly to the presence of a prostitute. For at least some denizens of the Roman Empire, the conditions of their enslavement would have created barriers to participation in the life of the Christian body. 1 Thessalonians 4:3–8 The earliest Christian document that we possess, 1 Thessalonians, offers guidance on sexual matters. In giving that advice, Paul refers explicitly to counsel on sexual ethics that he had previously delivered to the Thessalonian church in person. Introducing his advice in 4:3–8, Paul reminds his readers that “you learned from us how you ought to live…. You know what instructions we gave you” (4:1–2). Although twenty-first-century readers find aspects of this advice to be obscure, the Thessalonian Christians had the benefit of hearing these remarks in the context of Paul’s prior exhortations. The NRSV translation renders the passage with artificial clarity, lending cogency to Paul’s advice that is not warranted by his own phrasing: “For this is the will of God, your sanctification: that you abstain from fornication [=porneia];

Body Work: Slavery And The Pauline Churches

459

that each of you know how to control your own body in holiness and honor, not with lustful passion, like the Gentiles who do not know God; that no one wrong or exploit a brother or sister in this matter, because the Lord is an avenger in all these things, just as we have already told you beforehand and solemnly warned you” (4:3–6; emphasis added). This translation glosses over several difficulties. First, Paul instructed the (male) Thessalonian Christians to abstain from porneia, or sexual immorality. Whether Paul understood porneia to encompass precisely the field of activities connoted by the modern concept of “fornication” is unclear and even unlikely. Second, by supplying gender-neutral language in place of Paul’s gender-specific language, the NRSV translation artificially abstracts Paul’s advice from the patriarchal culture in which he wrote. Paul expressed concern that some Christians might wrong their brothers in sexual matters, an uneasiness that derived its particular charge from its patriarchal context. The NRSV neutralizes and obscures this context by implying that Paul was concerned with injuries to men or women in the community. Third, and most problematically, in the NRSV translation Paul tells the Thessalonian Christians that it is important that “each of you know how to control your own body.” What Paul actually wrote is that each (male) Thessalonian Christian should know how to “obtain his own vessel.” While some translators and commentators understand this as an idiomatic expression for controlling one’s own body or, possibly, sexual organs, other translators take “vessel” (skeuos) as a euphemism for wife, so that the passage would encourage the (male) Thessalonians to avoid sexual immorality by obtaining wives.115 I propose a third possibility. Paul’s advice could be construed as instructions to the male Thessalonian Christians to find morally neutral outlets for their sexual urges. And in the first century, domestic slaves were considered to be morally neutral outlets for sexual urges—vessels, we might say. Ernest Best notes, “The attitude of the Christian is now contrasted with that of the pagan, which had been his own only a few months before and which he could not be expected to forget completely on becoming a Christian.”116 Reflecting on the immediacy of that pagan context for the Thessalonian Christians, I find it surprising that Paul did not explicitly reinforce a prohibition on the sexual use of slaves, if he believed that Christianity demanded such a discipline. Commentators and critics typically insist that 1 Thessalonians 4:3–8 restricts sexual activity to marriage. David Williams writes, for example, that porneia “regularly means to have dealings with a prostitute (pornē), but 115  For a review of scholarship, coupled with an argument for reading “vessel” as “sexual organ.” see Elgvin, “‘To Master His Own Vessel.’ ” 116  Best, Commentary on the First and Second Epistles to the Thessalonians, p. 165.

460

Glancy

it was also used of any form of illicit sex—illicit from the Jewish/Christian point of view—that is, of any sexual relationship outside of marriage.”117 This reading, however, is a construction of what Williams believes Paul must have meant, rather than a straightforward exposition of what Paul actually said. Paul warned against porneia, but he did not define the parameters of porneia. As Williams acknowledges, a “restriction of sex to marriage struck the pagan world of that day as odd, for it tolerated and even encouraged, at least in the case of men, various forms of extramarital sexual activity.”118 Indeed, Paul urged the (male) Thessalonian Christians to adhere to more careful sexual standards than those tolerated by the Gentiles. Unfortunately, he did not clarify in what ways the newly formed Christians were to differentiate their behavior from that of their Gentile neighbors. The insistence of contemporary scholars that this advice required Christians to restrict their sexual activity to marriage exceeds what Paul wrote to the Thessalonians. O. Larry Yarbrough assimilates Paul’s advice in 1 Thessalonians 4:3–8 to the proscription against adultery, “which was of course fundamental to Jewish paraenesis, being one of the Ten Commandments.”119 G. P. Carras argues more broadly that Paul’s instruction to the Thessalonian Christians to distinguish their sexual behavior from that of their Gentile neighbors was an admonition to follow Jewish sexual norms. Thus, Carras suggests, Paul would expect the newly formed Christians to avoid incest, homosexual practices, prostitution (although it is unclear whether first-century Judaism univocally condemned prostitution), and adultery.120 In the ancient world, however, adultery referred specifically to sexual activity between an honorable married woman (therefore, not a slave) and a man who was not her husband. Carras’s argument addresses the behaviors that Paul’s counsel would exclude but does not address those behaviors still implicitly permitted by a call to observe the Jewish sexual code, which did not call into question a man’s casual sexual access to his own slaves. The crux of this difficult passage is Paul’s advice to “obtain a vessel.” Since antiquity, commentators have debated whether Paul advised the (male) Thessalonians to control their bodies/sexual organs or to acquire wives. Both possibilities present difficulties to the interpreter. What does it mean to obtain or acquire one’s own body or sexual organs, for example? While the perfect tense of the verb ktasthai could be extended to denote “to possess,” the present 117  Williams, 1 and 2 Thessalonians, p. 71. 118  Williams, 1 and 2 Thessalonians, p. 71. 119  Yarbrough, Not Like the Gentiles, p. 77. 120  Carras, “Jewish Ethics and Gentile Converts.”

Body Work: Slavery And The Pauline Churches

461

tense (which Paul uses) does not have this denotation. On the other hand, referring to a wife as a skeuos, a vessel, does not accord her a particularly dignified position. Since neither translation is the obvious choice, Beverly Gaventa advocates translating skeuos neither as body/sexual organs nor as wife but literally, as vessel, to preserve the inherent ambiguity of Paul’s phrasing.121 Moreover, translating skeuos as vessel leaves open the possibility that the Thessalonian Christians would have understood Paul’s counsel to be consistent with the pursuit of other sexual options, which did not infringe on the honor of other free men in the community. (I am not insinuating that Gaventa espouses this final possibility. She writes, “Whatever Paul means in this particular text, it is certain … that Paul was opposed to all sexual expression outside of monogamous marriage.”122) Despite the desire of some critics to see in 1 Thessalonians a representation of “marriage in the sense of full partnership,” Paul’s words do not evoke such an elevated view of sexual relations between men and women.123 An appreciation for mutuality and reciprocity in marriage informs 1 Corinthians 7. However, critics should demonstrate and not simply assume continuity between the earlier letter and the later letter. Skeuos is a term used to refer to inanimate objects, household goods. Ernest Best rightly suggests that the “rendering wife [which he prefers] appears repulsive to modern thought, but this should not deter us if on other grounds it is the better meaning [as opposed to body], for it would not have been so abhorrent to the ancient world with its very different view of women.”124 However, Best then assimilates the passage to the later text in 1 Corinthians: a Christian man is “not to look on her [his wife] as a tool to satisfy his lusts but to appreciate her as a being who with himself is part of God’s holy people (1 Cor. 7.2–6).”125 Best insists that Paul did not want men to consider their wives as tools. Nonetheless, Paul’s advice that, in order to control his sexual urges, a man should “obtain his own vessel” does promote an instrumental attitude toward women. Some commentators cite the concern that viewing Christian wives as vessels would diminish the significance of marriage to support the argument that the vessel to which Paul refers is the body/sexual organs.126 My suggestion, 121  Gaventa, First and Second Thessalonians, p. 53. 122  Gaventa, First and Second Thessalonians, p. 53. 123  Schrage, Ethics of the New Testament, p. 227. 124  Best, Commentary on the First and Second Epistles to the Thessalonians, p. 161. 125  Best, Commentary on the First and Second Epistles to the Thessalonians, p. 164. 126  Morris, First and Second Epistles to the Thessalonians, p. 121; Richard, First and Second Thessalonians, p. 198.

462

Glancy

that Paul’s advice to men “to take a vessel” could encompass sexual access to slaves in one’s own household, acknowledges that referring to sexual partners as vessels treats them not as ends but as means. Given the pervasive assumption in the ancient world that slaveholders had free sexual access to their human chattel, Paul’s failure to iterate, or reiterate, a prohibition on such behavior is peculiar, but he may have delivered a stern warning to the Thessalonians on this matter when he was among them: “you know what instructions we gave you” (4:2a). Modern readers of 1 Thessalonians, however, did not hear those instructions, and we should be wary of assuming that we have easy access to their content. Perhaps Paul instructed the Thessalonians that participation in the body of Christ precluded sexual exploitation of slaves. Perhaps he did not deliver such a charge. However, an admonition to avoid sexual contact with one’s slaves would have been sufficiently countercultural that Paul would have done well to return explicitly to the matter when he urged the Thessalonian Christians to contain their sexual urges. Against this argument, Paul does admonish the Thessalonians that their actions should be “in holiness and honor, not with lustful passions like the Gentiles” (4:4b–5a). Would a man’s sexual use of his slaves violate the holiness of the Christian body? If so, how would a newly formed Christian man come to understand that sexual use of slaves would pollute his body? Assuming for a moment that Paul had delivered such instructions in person, would later generations of Christians, who passed this letter along, have preserved and respected that mandate? I emphasize the question of pollution and holiness because a man’s sexual use of his own slaves would emphatically not have violated first-century notions of honor. Male slaves were considered to be without honor, and female slaves without shame. Furthermore, Paul supplies as a rationale for his sexual advice that the new Christians are to avoid infringing on each other’s sexual rights: “that no one wrong or exploit a brother in this matter” (4:6a). Scholars sensitive to feminist concerns have resisted the tendency to neutralize the gender-specific language of the passage.127 Gaventa writes, “Engaging in sexual intercourse outside marriage involves violating the property rights of another male. On this reading, verse 6 refers to the injustice done a male when another male engages in sexual intercourse with a woman whose sexual activity “belongs” to him.”128 However, in the first century, a slave’s body, and thus her sexual activity, ‘belonged’ to

127  Fatum, “1 Thessalonians,” p. 259; Perkins, “1 Thessalonians,” p. 441; Gaventa, First and Second Thessalonians, p. 52. 128  Gaventa, First and Second Thessalonians, p. 52.

Body Work: Slavery And The Pauline Churches

463

the slaveholder. A slaveholder was not infringing on any other male’s property rights when he had sexual relations with his own slave. New Testament critics uniformly assert that in 1 Thessalonians Paul urges male members of the community to avoid sexual immorality by restricting sexual activities to marriage. This is not, however, what Paul says. Whether his words in 1 Thessalonians built on explicit earlier instructions to the community, which specified these behavioral parameters, is a matter of speculation rather than evidence. What Paul actually says is that male members of the Christian community should avoid immorality by “obtaining a vessel.” In the first century, many who heard such counsel would understand it as consistent with reliance on slaves as morally neutral sexual outlets. Moreover, the motivations that Paul offers for confining one’s sexual urges are, to a significant degree, consistent with a slaveholder having sexual relations with his own human chattel: the preservation of honor, for example, and respect for the rights of Christian brothers. Whether a slaveholder’s sexual relations with his own slaves would be consistent with the preservation of the holiness of the Christian body is less clear. Paul expresses no concern for the feelings, experience, or moral struggles of the vessel. Does his silence betoken an indifference to the subjectivity of (free) women or of slaves? How we answer this question may depend on whether we assume that there was continuity or development between Paul’s sexual ethics in 1 Thessalonians and in 1 Corinthians. In 1 Corinthians Paul speaks sympathetically of the wife’s role in marriage, but even there, he does not address a practice that was universally tolerated in the first century: the sexual exploitation of slaves as bodies. 1 Corinthians 5:1–13 Chapters 5–7 of 1 Corinthians constitute Paul’s most extended discussion of sexuality and Christian living. Paul cautions against porneia, sexual immorality, in 1 Corinthians 5:1–13 and 6:12–20.129 In chapter 7 Paul responds to questions the Corinthians had written him about marriage and life in the Christian body. Many commentators suggest that chapter 7 begins a new section in the overall structure of the letter.130 Other commentators note that, as Paul begins his response in chapter 7 to the Corinthians’ questions, he takes his cue from 129  Weiss, Der Erste Korintherbrief; Barrett, Commentary on the First Epistle to the Corinthians; Fee, First Epistle to the Corinthians. 130  Allo, Première Epître aux Corinthiens; Conzelmann, 1 Corinthians. Deming argues on other grounds for the structural unity of 1 Cor. 5–6, which would also suggest a break between chaps. 5 and 6 (“Unity of 1 Corinthians 5–6”).

464

Glancy

the previous discussions of porneia. A preoccupation with sexual immorality inaugurates and governs Paul’s discussion of marriage. Antoinette Clark Wire argues that Paul rehearses the arguments of chapters 5–6 so that his readers will understand the urgency of his concern in chapter 7. For Paul, Wire contends, marriage is the antidote to porneia.131 What relevance did any of this discussion have for enslaved members of the congregation, who could neither resist their masters’ sexual demands nor contract legitimate marriages? Paul begins his discussion of the Corinthians’ sexual practices by condemning the community for tolerating a man who had a sexual relationship with a woman who had been (and perhaps still was) his father’s wife. Paul labels the relationship an instance of porneia. He furthermore claims that this kind of porneia is not even found among Gentiles. He expresses surprise that the church has been less scrupulous than the surrounding pagan world in delineating the boundaries of appropriate sexual conduct: “It is actually reported that there is sexual immorality among you, and of a kind that is not found even among pagans; for a man is living with his father’s wife.”132 Paul advises, “When you are assembled … you are to hand this man over to Satan for the destruction of the flesh, so that his spirit may be saved in the day of the Lord.”133 Drawing on the metaphor of yeast permeating an entire batch of dough, he encourages them to remove the incestuous man from their midst. In the future, he says, they should not tolerate the presence of immoral persons in their community: “I wrote to you in my letter not to associate with sexually immoral persons—not at all meaning the immoral of this world, or the greedy and robbers, or idolaters, since you would then need to go out of the world. But now I am writing to you not to associate with anyone who bears the name of brother who is sexually immoral or greedy…. For what have I to do with judging those outside?”134 The phrasing of Paul’s complaint implies that individuals and groups could vary in their perceptions of what constituted porneia. Even in the first century this was a category open to interpretation. Paul’s purpose is not to provide an exhaustive treatment of porneia. Not surprisingly, then, he does not address a question immediately relevant to the present discussion: whether sexual relationships between slaves and their masters constitute porneia.135 As I have 131  Wire, Corinthian Women Prophets, pp. 73–79. 132  1 Cor. 5:1. 133  1 Cor. 5:5. 134  1 Cor. 5:9–12. 135  In his discussion of 5:1–13, Fee notes the divergence between first-century sexual mores and later Christian teaching, although he does not focus on the sexual vulnerability of

Body Work: Slavery And The Pauline Churches

465

argued, a positive answer to this question would not have been self-evident to a Christian congregation whose members had grown up in a social context that accepted the normalcy of such behavior. The target of Paul’s accusation is the Corinthian church rather than the individual man involved in the transgression.136 In fact, Paul demonstrates less consternation about the incestuous man than about the community that accepted and perhaps applauded his actions.137 Hans Conzelmann suggests, “Paul does not explicitly state the ground of his judgment, because the ground is self-evident: the community is the temple of God (6:19).”138 The presence of the incestuous man in the community threatens the soundness of the whole, a threat Paul represents with the metaphor of leavening (5:6–8). Paul alludes to his instructions from an earlier letter to avoid pornoi, sexually immoral men, and, presumably, women and clarifies these earlier instructions. The Christian community cannot cut itself off from the world, but it should avoid the intrusion of the world into the community. The church is not a refuge for anyone who practices immorality, and he catalogs those whom he considers immoral, from the sexually immoral to greedy persons to thieves. (See also 6:9–11, in which Paul asserts that the Corinthian community includes some who had formerly numbered among such unsavory characters. Here, as well, Paul lists pornoi before other categories of wrongdoers.) Since porneia represents such a threat to the well-being of the community it would seem the entire community would have an interest in eliminating it. However, the boundaries of behavior Paul defines as porneia remain unclear. Would slaves who submitted sexually to their owners number among the pornoi? At stake, although not explicitly addressed, is the status of slaves in the community. Although Paul is primarily concerned with the health of the community, he is also concerned with the destiny of the transgressing man (5:4–5). However,

slaves. He states that “the Judeo-Christian moral restrictions on human sexuality were not easily absorbed by pagan converts” (First Epistle to the Corinthians, pp. 196–197). 136  Conzelmann, 1 Corinthians, p. 95; Senft, Première Epître de Saint Paul aux Corinthiens, p. 73. 137  This view is almost universally held in the commentaries and other secondary literature; see, for example, Robertson and Plummer, Critical and Exegetical Commentary on the First Epistle of Paul to the Corinthians, p. 95; Fee, First Epistle to the Corinthians, p. 195; Meeks, First Urban Christians, p. 130; Mitchell, Paul and the Rhetoric of Reconciliation, pp. 112–116; Campbell, “Flesh and Spirit in 1 Cor. 5:5”; and Hays, “Ecclesiology and Ethics in 1 Corinthians.” An exception is Barrett, who holds that Paul’s “first concern” is for the transgressor (Commentary on the First Epistle to the Corinthians, p. 127). 138  Conzelmann, 1 Corinthians, p. 96.

466

Glancy

he expresses no interest in the moral status or destiny of the woman involved.139 Commentators generally agree that Paul’s lack of interest in the issue indicates that the woman was not a member of the Christian community.140 Lending credence to this position is Paul’s insistence in the same passage that Christians have no business judging those outside the community (5:12–13). Extending this logic, one might ask whether Paul’s silence on the sexual position of slaves indicates that he believes that slaves who are sexually involved with their owners are therefore alienated from the Christian body. Wire proposes that Paul is silent about the stepmother’s complicity because she is not in a position to be responsible for her actions, either because of extreme youth or enslavement.141 Wire’s proposal recognizes the lack of clarity that surrounds the early Christian community’s treatment of slaves’ sexual activity, and she may be right. Paul may have believed that the forced sexual activity of slaves was beyond moral judgment. His surviving words do not help us decide the question. The passage 1 Corinthians 5:1–13 offers a variety of challenges for understanding the impact of Pauline Christianity on slaves. Recognition that GrecoRoman slaveholders regarded their slaves as bodies in turn complicates our reading of the chapter. The firstcentury category of porneia does not map neatly onto modern categories of sexual impropriety. The question of whether Paul thought that sexual relations between slaves and their owners fell in the category of porneia (for both slaveholder and slave or for either slaveholder or slave) is especially important. Paul insists that porneia represents a threat to the entire community and not only to the individuals immediately involved. A focus on the sexual dimensions of slavery opens the question of why Paul 139  Paul does not specify whether the man’s father is alive or dead, whether the father has divorced the wife, or whether the man is living with his stepmother as wife or concubine. Most commentators think it unlikely that the father, if living, is still married to the woman in question because Paul does not refer to an adulterous relationship, but even this is unclear. For a discussion of relevant issues see Allo, Première Epître aux Corinthiens, pp. 117–120. Deming has suggested the possibility that the son might have an ongoing relationship with his stepmother as prostitute (“Unity of 1 Corinthians 5–6,” p. 295). Paul had no reason to be more specific about these questions because the Corinthian congregation was already familiar with the details of the situation. 140  Robertson and Plummer, Critical and Exegetical Commentary, p. 96; Allo, Première Epître aux Corinthiens, p. 118; Conzelmann, 1 Corinthians, p. 96, n. 24; Barrett, Commentary on the First Epistle to the Corinthians, p. 121; and Fee, First Epistle to the Corinthians, p. 197. Weiss seems alone in his impression that Paul’s silence on the woman stems from a belief that the responsibility for sexual transgression rests with the man (Erste Korintherbrief, p. 125). 141  Wire, Corinthian Women Prophets, p. 74. Deming refers to Wire’s suggestion as “unlikely” without giving reasons for dismissing it (“Unity of 1 Corinthians 5–6” p. 298, n. 35).

Body Work: Slavery And The Pauline Churches

467

remains silent concerning the woman’s complicity in the case of incest he reports in 1 Corinthians 5. 1 Corinthians 6:12–20 From the earlier discussion of porneia, broadly defined as sexual irregularity, Paul moves to a discussion of porneia, more narrowly defined as prostitution. In the centuries before the Roman occupation of Corinth, the city’s temples legendarily enslaved a thousand prostitutes.142 The practice of sacred prostitution disappeared long before Paul arrived in the city, but in the first century Corinth was still associated with secular prostitution. Many Corinthian coins bore the image of Aphrodite, a kind of early advertising for a popular service industry in the city.143 As a port city Corinth attracted travelers, a group well known in antiquity for their patronage of prostitutes. Perhaps it was inevitable that a member of the Corinthian church would find his way to a brothel. In 1 Corinthians 6:12–20, Paul relies on the language of economics and the slave trade to condemn the buying and selling of bodies in prostitution. Given the high proportion of prostitutes who were slaves, the passage raises questions about the implications of Pauline Christianity for slaves. Paul quotes what seems to be a slogan among some Corinthian Christians: “All things are lawful for me.” He supplies a retort: “but not all things are beneficial.”144 Paul writes: The body is meant not for porneia [sexual immorality/prostitution] but for the Lord, and the Lord for the body…. Do you not know that your bodies are members of Christ? Should I therefore take the members of Christ and make them members of a prostitute? Never! Do you not know that whoever is united to a prostitute becomes one body with her? For it is said, “The two shall be one flesh.” But anyone united to the Lord becomes one spirit with him. Shun porneia! Every sin that a person commits is outside the body; but porneia sins against the body itself. Or do you not know that your body is a temple of the Holy Spirit among you, which you have from God, and that you are not your own? For you were bought with a price; therefore glorify God in your body.145 [All secondperson pronouns in the passage are plural.]

142  Strab., Geography 8.6.20. 143  Engels, Roman Corinth, p. 52. 144  1 Cor. 6:12. 145  1 Cor. 6:13–20.

468

Glancy

The focus of 6:12–20 is on a Christian male who visits a prostitute. Jouette Bassler writes, “The prostitute is merely a vehicle for sexual freedom; Paul shows no theological interest in her.”146 Paul reminds the Corinthians that the body is made for the Lord and will ultimately share in the Lord’s resurrection. His lack of interest in the prostitute suggests that, unlike the body of the believer, her body “is not destined for resurrection.”147 Although commentators tend to assume that Paul’s lack of interest in the stepmother in 5:1–13 indicates that she was not a member of the community, the thought that a prostitute might be a member of the Christian body seems almost absurd. Dale Martin writes that, for Paul, a prostitute “is not a person in her own right (as if such a thing were imaginable for Paul) but a representative of the cosmos that is estranged and opposed to God and Christ.”148 The possibility that the Corinthian church might embrace women or men still working as prostitutes is not considered in the secondary literature. Wire writes: “The fact that Paul does not censure the prostitutes themselves suggests that they are not to his knowledge participants in the community. Yet most cross-class urban groups would include some prostitutes. Paul does say that many Christians practiced immorality before their freedom was purchased by Christ (6:9–11), and among those could be prostitutes living single or married lives in the community.”149 Perhaps freeborn or freed prostitutes attracted to the Christian message were able to seek other means of supporting themselves, although this is doubtful.150 However, enslaved prostitutes would certainly lack such control over their occupation. The scholarly consensus that Paul understood membership in the Christian body to be incompatible with prostitution thus has unrecognized implications for understanding the difficulties that slaves would face in joining the church. The language that Paul uses to challenge the common assumption that a sexual encounter with a prostitute has no lasting effects on the customer has broad implications for reconstructing his understanding of the nature of sexual activity.151 Paul claims that whoever has sex with a prostitute becomes one body with her and supports his position by quoting Genesis 2:24: “The two 146  Bassler, “1 Corinthians,” p. 332; also Conzelmann, 1 Corinthians, p. 112; and Mitchell, Paul and the Rhetoric of Reconciliation, p. 120. 147  Fee, First Epistle, p. 260. 148  Martin, Corinthian Body, p. 176. 149  Wire, Corinthian Women Prophets, p. 76. 150  Fleming, “Quae Corpore Quaestum Facit.” 151  For a clear discussion of Paul’s position on the theological ramifications of patronizing a prostitute, see Fisk, “ΠOPNEΥEIN as Body Violation,” pp. 540–558.

Body Work: Slavery And The Pauline Churches

469

shall be one flesh” (1 Cor. 6:16). Such a union, Paul says, is incompatible with spiritual union with the Lord (verse 17). Fee clarifies, “It is not the sexual union that is incompatible with Christ, it is such a union with a prostitute.”152 The believer belongs to Christ as a member, so his sexual activities affect the Lord to whom he belongs (verse 15). Martin graphically writes, “The man who has sex with a prostitute is, in Paul’s construction, Christ’s ‘member’ entering the body of the prostitute. Since her body is only part of a larger whole, the cosmos, the simple act of copulation between a man and a woman becomes for Paul copulation between Christ and the cosmos.”153 It is hard to see how, on Paul’s view, any sexual relations involving a believer could be morally neutral. All sexual relations would affect the Lord and indeed the entire Christian body (6:19–20).154 Again, the moral status of slaves (including, but by no means exclusively, prostitutes) whose owners insisted on using them sexually remains an urgent question. Paul relies on the metaphor of buying and selling bodies, of slave trading, to exhort the Christian community to avoid prostitutes. Ironically, Paul configures the male Christian body and the body of the church as objects of sale. The participle that Paul uses to express the Christian’s relationship with a prostitute or with the Lord, kollōmenos (6:16–17), implies economic obligation or subordination.155 In Paul’s view a man who patronizes a prostitute abridges his own freedom by obligating himself to her. Paul urges members of the community to obligate themselves instead as slaves to the Lord. The image of the slave market becomes more explicit. Paul informs the Corinthian community, “You [pl.] are not your own…. You were bought with a 152  Fee, First Epistle to the Corinthians, pp. 259–260. 153  Martin, Corinthian Body, p. 176. 154  In verse 19 Paul refers to “your” (plural) body as a temple, and in verse 20 he urges the Corinthians to glorify God in “your” (plural) body. I take the reference to be primarily to the body of the community, the collective body; the entire community should be vigilant about its purity, as in the discussion of 5:1–13. Others who interpret “your body” in 6:19–20 as an allusion to the community include Mitchell, Paul and the Rhetoric of Reconciliation (see index for references to 6:19–20), and Wire, Corinthian Women Prophets, p. 77. Most commentators, however, interpret “your body” primarily as an allusion to the individual Christian’s body, despite the parallelism to 3:16. See, for example, Barrett, Commentary on the First Epistle to the Corinthians, p. 151; Fee, First Epistle to the Corinthians, p. 249, n. 5, and p. 260; and Senft, Première Epître de Saint Paul aux Corinthiens, p. 85. Conzelmann also takes this position in his discussion of verse 19, although elsewhere he seems to assume that verses 19–20 refer to the Christian community (1 Corinthians, p. 96, cf. p. 112). 155  Porter, “How Should κoλλωμϵνoϛ in 1 Cor. 6:16.17 Be Translated?” See also Harrill, “Indentured Labor of the Prodigal Son (Luke 15:15).”

470

Glancy

price” (6:19b–20a).156 The price paid is Christ’s death, yet the metaphor also positions Christ as a customer at a slave auction, purchasing the Christian body.157 This use of language seems ironic in light of the economic realities implicit in the exchanges of prostitution, in which sexual access to a woman’s (or man’s) body is bought and sold. The bodies of enslaved prostitutes were of course subject to an additional kind of sale, that of slave trading. Paul expresses concern for the vulnerabilities of the Christian body (the body of the individual male and the body of the community) without expressing concern for the vulnerabilities of the (frequently enslaved, usually female) prostitute’s body. S. Scott Bartchy seems accurate in his assessment that “Paul did not want any Christian who was in slavery to think … that his slave-status was any disadvantage in his relation to God.”158 Although scholars debate the extent to which “the liminal transcendence of societal oppositions that was declared in baptism” affected patterns of social interaction and hierarchy among early Christians,159 the assumption that the message of the gospel extended without prejudice to slave as well as free has largely gone unchallenged. However, Paul’s insistence on the fundamental incompatibility between the body of prostitution and the body of Christ seems to leave even those prostitutes who were enslaved beyond the boundaries of Christian community. Paul draws on the metaphor of slave trading to warn the Christian community that it imperils itself when members frequent prostitutes rather than to express concern for the involuntary sexual use of the bodies of slaves, even in prostitution. 1 Corinthians 7:1–24 In 1 Corinthians 7, Paul addresses, at some length, questions pertaining to Christians and marriage. Although Paul writes explicitly of the situation of slaves in 1 Corinthians 7:21–24, the relevance of chapter 7 as a whole to the lives of slaves is problematic, since slaves were unable to contract legal marriages. J. Albert Harrill observes that, on the standard view, “Paul’s entire response to the Corinthian congregation on marriage, therefore, has little relevance to slaves.”160 At the same time, by limiting the legitimate range of sexual expres156  Fee, First Epistle to the Corinthians, p. 264. Conzelmann suggests that the idea of redemption as “ransom” is traditional and says, “The metaphor is not developed. The point is merely that you belong to a new master. Beyond this the metaphor should not be pressed. There is, for example, no reflection as to who received the payment” (1 Corinthians, p. 113). 157  Martin, Corinthian Body, p. 178. 158  Bartchy, First-Century Slavery, p. 175. 159  Meeks, First Urban Christians, p. 161. See also Gnilka, Der Kolosserbrief, p. 190. 160  Harrill, Manumission of Slaves in Early Christianity, p. 123.

Body Work: Slavery And The Pauline Churches

471

sion to marriage, Paul implicitly suggests that slaves who oblige their masters sexually are engaged in porneia. This section considers how Paul’s advice on marriage shapes our understanding of the challenges faced by enslaved Christians. It also reconsiders Paul’s ambiguous directive to slaves in 7:21 in the context of the sexual dimensions of first-century slavery. Paul introduces his discussion of marriage by allowing that matrimony protects men and women against porneia (7:2). He reinforces this declaration by conceding that, although a single life is desirable, marriage is nonetheless preferable to sexual incontinence (7:9). Continence, then, encompasses both celibacy in the single life and fidelity in the married life; the alternative, it seems, is porneia. Although later Christian tradition has often assumed that this demarcation of sexual practice was normative, we have seen that the majority of those living in Greco-Roman cities viewed a wider range of sexual practices as acceptable and ordinary. As we have seen, the master’s right to have sexual relations with one of his slaves was not accompanied by moral judgment against the master or the slave. Christian slaveholders who heard Paul’s words could choose to redirect their sexual attention to their wives. The same is not true for enslaved Christians. According to Wire’s reconstruction, the church at Corinth included a number of powerful women ascetics. She writes, “The life-support arrangements of these women—where they live and what they do—are suggested only indirectly in the various domestic situations Paul describes. Slave women … probably retain traditional domestic responsibilities.”161 Wire does not consider the complicating factor that the “traditional domestic responsibilities” of many enslaved women included sexual obligations. Indeed, Paul seems to assume that those whom he addresses in 7:1–16 could freely choose to marry or not to marry.162 He offers specific advice to various categories of married and unmarried persons.163 After addressing those who are married to believing partners and those who are not married, he addresses “the 161  Wire, Corinthian Women Prophets, p. 91. 162  Deming, Paul on Marriage and Celibacy. The virgins of 7:25–28, 36–38 are a different case; Paul does not seem concerned about their willingness to marry. 163  Although commentators offer various divisions of the chapter, they agree that the advice to various groups provides structure to the discussion. For example, Conzelmann and Senft categorize verses 1–7 as general directions and verses 8–16 as advice to particular groups (Conzelmann, 1 Corinthians, p. 114; Senft, Première Epître de Saint Paul aux Corinthiens, pp. 87–90), while Fee’s taxonomy includes verses 1–7 in advice to the married, verses 8–9 in advice to the “unmarried” (who Fee posits are widowers) and widows, verses 10–11 in advice to believing couples, verses 12–16 in advice to couples with only one believing spouse, verses 25–38 in advice about virgins, and verses 39–40 in advice to married women and widows (Fee, First Epistle, pp. 268, 287–288).

472

Glancy

rest” (verse 12). The advice he then delivers pertains particularly to Christians whose spouses are not believers, and commentators accordingly delimit the range of “the rest” to those whose spouses do not belong to the Christian community.164 Most commentaries do not consider who else could be included in “the rest.” They assume that Paul’s list exhaustively covers persons in all possible situations. According to this view, Paul has nothing to say about the options open to enslaved Christians, who could not contract marriages and could be used sexually by their owners. Commentators interpret Paul’s silence about the moral status of the stepmother in a case of incest (5:1–13) and the prostitute (6:1–12) as evidence that these women are not participants in the Christian body. Paul’s silence about the plight of sexually exploited slaves could be construed as evidence that he perceives their (compelled) behavior to place them outside membership in the Christian body. Harrill has offered the intriguing suggestion that “the rest” in verse 12 includes not only those married to unbelievers but also “those not empowered to marry at all unless enfranchised (7:21–23),” that is, slaves.165 He makes this proposal as he discusses 7:21, which he translates: “You were called as a slave. Do not worry about it. But if you can indeed become free, use instead [freedom].”166 Because of the difficulties of this text, an overview of Harrill’s position is essential. He begins his discussion of 7:21 with a summary of the long-standing debate over the translation of the verse.167 As he says, scholars base their translations not only on the syntax of the passage and its immediate context in the letter but also on their understandings of ancient slavery. These understandings often minimize the brutality of Greco-Roman slavery. Fee, for example, who offers a careful discussion of the syntax of the passage, asserts that slavery “provided generally well for up to one-third of the population in a city like Corinth or Rome. The slave had considerable freedom and very often experienced mutual benefit along with the owner.”168 Along with an analysis of the syntax of

164   Robertson and Plummer, Critical and Exegetical Commentary, p. 141; Conzelmann, 1 Corinthians, p. 121; Barrett, Commentary on the First Epistle to the Corinthians, p. 163; Fee, First Epistle to the Corinthians, p. 298; and Senft, Première Epître de Saint Paul aux Corinthiens, p. 165. 165  Harrill, Manumission of Slaves in Early Christianity, p. 123. 166  Harrill, Manumission of Slaves in Early Christianity, p. 122. 167  Harrill, Manumission of Slaves in Early Christianity, pp. 77–108. 168  Fee, First Epistle to the Corinthians, p. 319.

Body Work: Slavery And The Pauline Churches

473

verse 21,169 Harrill provides a discussion of seventeen philological parallels that support his contention that Paul urges slaves to “use instead [freedom].”170 He concedes that his translation requires Paul first to establish (in verses 17 and 20) and then violate a principle: each should remain content with the state in which he or she was called. Throughout chapter 7, however, Paul offers exceptions to the rule he has laid down. Although he says that spouses should be content to remain married, for example, he establishes guidelines for spouses seeking separation (verse 11). He encourages those who are single to be content with that status but states that they do not sin if they choose to marry (verse 28).171 This rhetorical pattern helps clarify Paul’s logic in 7:21. He encourages slaves not to be disturbed by their bondage but then offers an exception to this principle: “If you can indeed become free, use instead [freedom].” I find Harrill’s argument that in 7:21 Paul urges slaves to take advantage of opportunities for freedom to be convincing. Moreover, I find it plausible that Paul gives such advice in the midst of his discussion of marriage because he is aware that slaves, who are unable to contract marriages, are the sexual property of their owners. Harrill observes that “seizing such opportunities for liberation would have enabled a slave to escape a master’s violent coercion and, as a freedman/woman, to secure more control over his or her own body and daily activities to pursue asceticism.172 (Harrill does not explicitly address the question of sexual vulnerability, although such awareness may be implicit in his formulation.) Some forms of manumission still gave a master control over the freedperson, but freedpersons were in general less vulnerable than slaves to the demands (including sexual demands) of their masters. (Note that this reading of 1 Corinthians 7:21 does not include any claim about how Paul would respond to slaves or freedpersons who were still being used sexually by their owners or former owners—whether he would believe that such activity excluded the slave or freedperson from the Christian body or whether he would perceive coerced sexual activity as morally neutral.)

169  See also the discussion of syntax in Fee, First Epistle to the Corinthians, pp. 316–317. 170  Harrill, Manumission of Slaves in Early Christianity, pp. 108–121. 171  See also 7:15, 36, 39, and Harrill, Manumission of Slaves in Early Christianity, pp. 123–126; Fee, First Epistle to the Corinthians, pp. 268, 318; and Deming, “Diatribe Pattern in 1 Cor. 7:21–22.” 172  Harrill, Manumission of Slaves in Early Christianity, p. 122. Robertson and Plummer write: “He [Paul] regarded marriage as a hindrance to the perfection of the Christian life (verses 32–35). Was not slavery, with its hideous temptations, a far greater hindrance?” (Critical and Exegetical Commentary, p. 148).

474

Glancy

However, I do not support Harrill’s argument that Paul intends to include slaves among “the rest” of verse 12: “To the rest I say—I and not the Lord—that if any brother has a wife who is an unbeliever, and she consents to live with him, he should not divorce her.” With the majority of commentators, I categorize verses 17–24 as Paul’s articulation of his governing principle with two examples: circumcision/uncircumcision and slavery.173 His governing principle, again, is that each should be content with the state in which he was called. To include slaves among “the rest” awkwardly leaves an articulation of a general principle with a single illustration (circumcision/uncircumcision) followed by advice to slaves, which only obliquely pertains to the overall topic of the chapter, marriage. I follow the generally held view that when Paul refers to “the rest” in verse 12, he is referring narrowly to the case of Christians whose spouses are not Christians. In articulating the governing principle of the chapter he provides two examples. As he sets forth the second example, however, he is aware of the problematic ramifications of slavery with respect to sexual expectations and marital status, so he also offers one of the many exceptions that he scatters throughout the chapter. We have seen that Paul rejects the idea that sexual couplings can be a matter of moral indifference (1 Cor. 6:16), and he furthermore implies that all sexual unions involving Christians have implications for the entire community (1 Cor. 6:15–20). The passage 1 Corinthians 5–7 is marked by a concern for the integrity of the Christian body, which encompasses a desire to confine sexual activity to marriage, a standard to which many slaves would not have been able to adhere. Despite Paul’s belief that servile status should not affect one’s life in Christ, many slaves may well have found that sexual demands made by their masters complicated their membership in the Christian body. We simply do not know how Paul responded to the situation of slaves who were used sexually by their owners. Perhaps, however, concern over this vulnerability prompted him to advise slaves to take advantage of opportunities for freedom as they arose. Body Politics Bartchy writes that, according to Paul’s understanding, God’s call in Christ “had made irrelevant every particular social and religious status.”174 The assumption that servile status was in no way a disadvantage with respect to one’s 173  Robertson and Plummer, Critical and Exegetical Commentary, p. 144; Barrett, Commentary on the First Epistle to the Corinthians, p. 167; Conzelmann, 1 Corinthians, p. 114; Fee, First Epistle to the Corinthians, p. 306; and Senft, Première Epître de Saint Paul aux Corinthiens, p. 95. 174  Bartchy, First-Century Slavery, p. 173. (Bartchy is referring specifically to 1 Cor. 7:17–24.)

Body Work: Slavery And The Pauline Churches

475

membership in the church has been pervasive in discussions of New Testament texts pertaining to slavery. James D. G. Dunn argues, “If ‘Christ is everything and in everything,’ then nothing can diminish or disparage the standing of any one human in relation to another or to God.”175 J. N. Aletti claims that Galatians 3:28 and Colossians 3:11 indicate that there was no disparity in the “capacité éthique des croyants.”176 F. F. Bruce asks, “What real difference could there be for a Christian between bond and free?”177 In making such statements, New Testament scholars ignore the ancient equation between slaves and bodies. Even if we assume—and it is a major assumption—that Christian masters did not force their slaves to provide them with sexual services, pagan masters would not have shared such scruples about the bodies they owned. Although New Testament texts are silent on the sexual availability of slaves, a vast array of Greco-Roman sources assume that masters had unlimited sexual access to their slaves. In 1 Thessalonians Paul urges male Christians to avoid porneia by obtaining their own vessels. Accustomed to relying on slaves as morally neutral sexual outlets, the newly converted Thessalonian Christians could easily have construed these words as advice to maintain the honor of respectable freeborn Christians by turning to slaves to satisfy their sexual inclinations. Indeed, it is strange that Paul does not explicitly condemn the sexual use of slaves if he believes this practice to be inconsistent with the Christian ethos. New Testament scholars have uncritically assumed that later Christian rejections of the sexual use of slaves are already implicit in the earliest Christian document we possess. A slaveholder had the right to profit from a slave’s body by forcing her or him into prostitution. Paul’s unequivocal separation of the body of prostitution from the body of Christ would seem to exclude all prostitutes, even enslaved prostitutes, from membership in the church. More subtle would have been the problems faced by slaves whose masters used them sexually. In 1 Corinthians Paul distinguishes sexual relationships within marriage from all other sexual relationships, and he encourages the Corinthian Christians to confine their sexual practices to marriage. Perhaps this is one reason he encourages slaves to seek a change in their status when possible, taking advantage of opportunities for freedom. We do not know precisely how Paul responded to the situation of slaves whose masters insisted on sexual relations with them. However, attention to this dimension of ancient slavery makes it impossible to maintain both that servile status was no impediment to full involvement in the Christian 175  Dunn, Epistles to the Colossians and to Philemon, p. 227. 176  Aletti, Saint Paul Epître aux Colossiens, p. 235. 177  Bruce, Epistles to the Colossians, p. 151.

476

Glancy

body and that within the Christian body the only sexual relations tolerated were those between husband and wife. Paul believes that baptism breaks down the barriers between slave and free (Gal. 3:28; 1 Cor. 12:13). (It may be useful to differentiate between the claim that for those in Christ there are no distinctions between slave and free and the claim that slavery presents no obstacles to those who want to join the Christian body. In fact, Paul never clearly makes the latter claim, which has been repeatedly made for him by modern scholars.) Bruce notes that it would have been easier for Paul than for an enslaved person to maintain the position that slave status was irrelevant for those in the church.178 John M. G. Barclay has argued that masters and slaves in the same church would have found it difficult in practice to transcend those roles. For example, a slave might have found it very hard to follow Paul’s advice to admonish a member of the community who had transgressed, if that transgressor were his or her owner.179 We have seen that Acts of the Apostles accords householders (who would have been slaveholders) a disproportionate role in fostering the growth of the church, a picture that raises even more questions about the dynamics between slave and free members of the Christian body. Although Paul himself minimizes the significance of servile status, we do not have the coeval words of slaves, describing their own attempts to live a new life in Christ. In a society that represented and treated slaves as bodies, ecclesial incorporation of slaves would have exposed the body of Christ to the somatic vulnerabilities of enslaved members.

178  Bruce, Epistles to the Colossians, p. 168. 179  Barclay, “Paul, Philemon, and the Dilemma of Christian Slave-Ownership.”

How and Why Ancient Slavery Came to an End* Marc Bloch In the Roman world of the first centuries ad, the slave was everywhere: in the fields, in shops, in workshops, in offices. The rich kept hundreds or thousands, and one had to be quite poor not to own at least one. This is definitely not to say that servile labor had a monopoly on any activity, however humble. Many artisans were in a free condition, and innumerable fields were cultivated by peasants, small land-owners or tenant-farmers who had never been the property of a master. Vespasian reserved for the free workers of Rome those hard tasks that he refused to give to machines. Nonetheless, neither the material life of Greco-Roman societies, nor even their civilization at its most exquisite, could be conceived of without the existence of this forced labor. The Germans also had their slaves, either as servants or field hands. On the other hand, the Europe of modern times, with a few rare exceptions, has not known slavery on its own soil. For the most part, this transformation, one of the most profound that mankind has known, took place very slowly in the course of the high Middle Ages. At the time of the barbarian invasions and in the early days of their kingdoms, there were still many slaves in all parts of Europe. There were more, it would appear, than during the early days of the Empire. The great source of slavery had always been war. The victorious expeditions of the legions during the Roman conquest populated the slave-pens of Italy. Similarly, beginning in the fourth century, the incessant struggles of Rome against her enemies, the battles that these enemies frequently waged between them and the brigandage of regular soldiers or of professional bandits (the distinction was not always easy to make, any more than it is in contemporary China), accumulated in the hands of one group or the other this booty of Source: Bloch, Marc, “How and Why Ancient Slavery Came to an End,” in Marc Bloch (William R. Beer transl.), Slavery and Serfdom in the Middle Ages: Selected Essays, Berkeley: University of California Press, 1975, 1–31. © 1975 by the Regents of the University of California. Published by the University of California Press. * This essay was published as it was found among Bloch’s papers by the editors of the Annales d’Histoire Economique et Sociale. It has a number of lacunae which the editors found it would be impossible to fill without an immense review of Bloch’s work. Therefore, the lacunae in the data on the tenants of Saint-Germain and Saint-Rémi are merely noted.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_019

478

Bloch

flesh and bone which was only rarely given back when the fortunes of war changed sides. “There is no house, however poor, where one does not find a Scythian slave”—that is, according to the usual vocabulary of the author, a Gothic slave—the African Synesius wrote, around the year 400. He was thinking of the eastern regions of the Empire, the only ones he knew from experience. But if we replace “Gothic” with a term more general, such as “barbarian,” there is no doubt that in this form the observation retains its truth for all of the world that was still Roman. As for the invaders themselves, we know that a large number of the inhabitants of Romania, of all classes, had been reduced to servitude by them. In the biography of Saint Severin, which presents a day to day account of a sojourn in small towns on the Danube which were ceaselessly threatened by the German tribes around them, these raids for captives appear to be a common occurrence. Here and there in the texts we come across some tragic fates, which must have resembled many others: think of the great noblewoman of Cologne who, as a prisoner of the barbarians, served them as a slave for a long time. Or consider the other Gallo-Roman noblewoman whom brigands carried off: they displayed her for sale in the market of Clermont. The fate of runaways was not always better. Among the wanderers that the misfortune of the times cast upon the roadways of Romania, more than one fell into slavery, a victim of the same peoples among whom he had sought refuge. The warrior who won his captives in great number with his sword did not keep them all among his followers. The principal profit that he expected was from selling them. Barbarians also came to offer slaves of Roman blood for sale in land that was still Roman. Such offers were so frequent that in 409 an imperial law could only recognize the validity of these sales, but on condition that the slave could always buy back his freedom, either by paying back to his new master the sum that he had paid, or by serving him for five years. The invasion of Illyria and Thrace, according to Saint Ambrose, had dispersed men for sale “throughout the world.” Later, Gregory the Great saw Lombards leading prisoners “with ropes around their necks, like dogs,” whom they had captured during an expedition to Rome, and for whom they thought they would find buyers in the Kingdom of the Franks. The great disturbances in Europe led to a recrudescence of the trade; the poverty of the people acted in the same way. In spite of Roman law, fathers sold their children; the fact was recorded in the sixth century in Corsica. Whereas during the first century, a time of peace and prosperity, Pliny the Younger complained that slave manpower was so rare, and while in the third century a slave was quite expensive, at the beginning of the Middle Ages, human merchandise had become abundant again at a reasonable price. The trade continued to be very active throughout the era of the barbarian kingdoms and up to Carolingian times. As great businessmen before God, the

How And Why Ancient Slavery Came To An End

479

Jews played an important part, but they were far from being the only ones to practice the trade. The biographies of saints, laws, and formulae mention it constantly. Great Britain furnished many slaves to the continent, as far away as Provence and Rome, torn as she was by frequent wars between the AngloSaxon kings or against the Celtic speaking peoples, who were themselves prey to internal strife. On the fields of the wealthy, slaves of every origin would rub shoulders, barbarians as well as Romans. As objects in ongoing exchanges, they served as a medium in transactions that at the time were very numerous, in which money did not appear, as a standard, and at times as small change. The texts show us a certain Gaul acquiring a field against payment of a sword, a horse, and a Saxon woman. In enumerating the principal “species” that merchants customarily sold, a capitulary cites gold, fabrics, and slaves. The traders’ caravans were not just seen traveling from one country to another within Europe; in their commercial balance, servile livestock counted among the principal export products. The trade sent large numbers to Muslim Spain, and a lesser number, perhaps by way of Venice and the plains of the East, to the Greek and Arab Orient. However, to look more closely, very clear signs indicate that after the ninth century slavery was far from holding a place in European society comparable to that which it previously had held. To understand and weigh these signs of decline, we must first trace the changes undergone by the economic implementation of servile manpower after the end of the Roman era. Two methods were open to the master who wanted to make use of the living force that the law placed at his total direction. The simplest consisted of supporting the person, as one would a domestic beast and, as with the animal, to use his labor in any way whatever. But the slave could also be set up on his own account. In this case the master exacted, in various forms, part of his time and the products of his toil, while leaving him the task of supporting himself. Now in the last centuries of the Empire, this second procedure was more and more widespread. Even in industry the two procedures had been in competition. The wealthy, who owned great troops of slaves, had always recruited domestic workers from their ranks, thereby saving themselves, on many tasks and manufactures, from recourse to salaried labor or trading. This custom persisted until the ninth century. However, once the household needs were satisfied, was there a duly qualified surplus of manpower on hand? At all times one was forced to find a remunerative outlet for it in production for the market. This could be done by setting up vast workshops where the slaves were made to live laboring under the orders of and for the sole profit of the master of the plant. Here and there, in the first centuries of our era, we find actual factories, such as the famous

480

Bloch

workshops of Graufesenque and Lezoux in Gaul. Besides free workers they probably comprised unfree workers who either belonged to the employer or whom he rented from other masters. After the third century these establishments declined. Many manorial workshops did remain, but they hardly provided for more than the manor itself, and the imperial factories delivered their products to the state alone. Market demand had always been met by small-scale crafts, which thereafter encountered no competition. For want of work for laborers he could not afford to leave idle, the owner was forced to cut back. The slave exercised his trade for the public and after clothing and feeding himself on his revenue, turned over the rest to his master in various forms, which were frequently determined in advance. This practice, as old as craft production itself, became sufficiently widespread that it appeared necessary to rule on the judicial problem it raised. In the barbarian world itself, Burgundian law ruled. It was in agriculture, however, that the transformation proved especially profound. Little farms had always occupied a large part of the soil of Romania— the greater part of it, probably, except in various regions of Italy. Their servile personnel were naturally very limited. Besides these, at the beginning of the Christian era, there were immense domains cultivated by veritable armies of slaves in bands, comparable to the Negroes of modern colonial plantations. Toward the end of the Empire, this system was generally abandoned. The large landowners, taking advantage of their possession of large expanses, parceled them into little farms whose occupants paid rent in a variety of forms. Among the beneficiaries of these distributions appeared a large number of slaves taken from the central work gangs, each one charged with the responsibility of his own fields. Some were enfranchised the moment they were settled. Many others, though they had become tenant farmers, remained legally in their previous condition. Of course, the slave-tenant was not entirely new. He had long been present, notably on small properties whose owners could scarcely run the risk of too-extensive enterprises. But his spread was a new thing. The phenomena of parceling out latifundia and the decline of slave manufacturing, while of primary interest to the history of slavery, obviously go considerably beyond it. They amount to the triumph of small-scale over largescale enterprise. They could not alone account for all the changes that thenceforth affected the employment of servile manpower, though. It would be very inexact to speak of a total disappearance of large-scale farming. The creation of little farms had considerably reduced the extent of the resources for direct exploitation, but it did not cause them to disappear. Around the end of the Empire, and until the ninth century, the majority of the great landlords still kept sizeable farms under their administration, though even then the methods of cultivation came to be modified.

How And Why Ancient Slavery Came To An End

481

Of course, the master did not cease to feed, house, and clothe the slaves under his control who helped him to cultivate his fields. They were, however, less and less suitable to the task. From then on, it was the tenant farmers, whose land had been placed under the tenure of the principal domain, who were called to do the greater part of the work required for its profitability. Doubtless some were peasants long used to living in dependency upon a powerful landowner or who gradually became so. Others had been settled only recently on their parcels of land. By giving up a part of his estate, the large landlord could thereby assure himself of the labor power demanded by the rest. As we have seen, there were many slaves among these newly established tenant farmers. They continued to toil for their master, but they were no longer supported by him, any more than a factory owner supports his workers today. The land that had been granted to them, aside from being subject to rents that do not concern us here, was like their salary, upon which they had to live. What considerations, therefore, had induced slave owners, who also possessed vast plantations, to henceforth prefer the system of sharecropping to the seemingly more practical, direct utilization of human livestock? In all societies that have used slave labor, from large-scale to the simplest form—those of Roman latifundia and those of the plantations of the East Indies—its use has been in response to conditions that are always the same, implacably required by its very nature. The slave is a bad worker; his output has always been recognized as fairly low. He represents, moreover, perishable capital. The present-day owner who loses a worker through death or sickness may have some trouble replacing him if the manpower market is unfavorable. But if he manages to replace him, he has suffered no loss since the wage remains the same no matter who the man. However, the master whose slave died or became ill or quite simply got old, had to purchase another. At one stroke he lost the amount of his initial investment. Of course, to fill certain gaps, one could rely on certain slaves born into the household. They could not fill all the gaps, however; of all livestock breeding, that of man is the most delicate. These inconveniences were not very serious as long as the slave inventory remained abundant and hence of relatively low cost. To accomplish a small task, one had to waste a lot of slaves; if one of them turned up missing, it was neither a strain nor an expense to find a substitute for him. This was the state of affairs at the beginning of the Christian era, which had been created by so many victorious wars waged by Rome: it explains the existence of the great slave-gangs. But soon the recruitment of slaves became more difficult. Their value increased. This was when people turned toward the land-tenure system.

482

Bloch

Let us imagine the slave established on a little farm on his own account. Living in better organized families, his kind perpetuate themselves more securely. On the fields that have been granted to him, his work is of better quality. Because the rent must willy-nilly be paid, the surplus of the produce on which his livelihood is based depends upon his own labor. There remain the obligatory services on the lands of the master. Doubtless their rendering was not the best, and this may have been one of the reasons that, much later, from the tenth century on, led in turn to their abandonment. At least one could expect that, in not wanting to see his holding taken away, a holding that was only granted to him in return for carrying out these tasks, the slave-tenant would acquit himself less badly than one who ate in the communal slave-stable. The renewal of the slave trade at the time of the invasions may have provoked a return to the old use of slave labor in vast rural work gangs. The documents are too imprecise to allow us positively to affirm or deny it. What is certain is that there was no revolution of any great scale. The change had been made. Moreover, the Germanic chieftains, into whose hands so many great domains were falling at this time, were prepared to adopt the tenant-farming system. It was part of their peoples’ traditions. In ancient Germany, general economic conditions were not favorable for any kind of large-scale enterprise. The noble and rich had many lands, many of which lay fallow, and had many slaves often captured in warfare. To make use of these vast reaches as well as they could, there was nothing for them to do but divide them up. To feed so many people, it was absolutely necessary to allot each one a plot of land, since it would not have been convenient to maintain them in the household of the master. At a time when the slave-tenant was still a rarity in Italy, Tacitus was noting its frequency on the other side of the Rhine. Now this slave-tenant doubtless remained a slave in his personal status. Even in the Carolingian period, legislative precedents were forced to mention the distinction between servus and other dependants of the manor, such as the coloni. On many estates, while the duties owed by free men were generally fixed, the lord reserved the right to demand the labor of unfree tenants any time he deemed fit, “whensoever it should be commanded of them.” It seems that their wives—and only their wives—were drafted into the manorial workshop to work there under the orders of the master, and they alone provided him with fabrics and linen. However, in practice, the destiny of the slave, established on a little plot of land whose cultivation was entrusted to him, was different from what the word “slavery” implies. He paid over to his master only a part of the fruits of his labor, and gave him only a part of his work-time, because even though his duties were theoretically limited, obviously the necessity that obliged the master to allow the villein enough free time to extract his

How And Why Ancient Slavery Came To An End

483

livelihood from the holding and to pay his rents, also prevented those tasks from taking up all his time. He did not live out every moment under the orders of another man. He had his home and hearth, and he managed the cultivation of his fields himself. If he was particularly hard-working or particularly shrewd, he ate better than his neighbor; or insofar as there was a market, he could sell his surplus produce. Juridical institutions were not at all slow to recognize the peculiarities of his condition. Since he was one of those tillers of the soil whose efforts were, above all, important to the prosperity of the Empire, the laws of the fourth century prohibited the master from depriving the slave-tenant of his land, as they also protected the free tenant. Doubtless this rule of “attachment to the glebe” was only observed for a little time and faded in the ruin of the imperial state that had proclaimed it. But between the slaves who were chasés—that is, each provided with a house (casa) and adjoining lands—and those who were not, Carolingian law made a very important distinction: the former were considered as part of the real estate, the latter as part of the furniture. The laws governing their disposal were therefore entirely different. Above all, after the second half of the ninth century, the custom of the manor, which in the absence of a written law had long served to regulate the relations between the lord and his free tenants, extended its protection to the slave-tenant. Instead of the arbitrary power of the master, there was substituted the rule of a local tradition that was frequently quite harsh, but in applying to high and low alike prevented or was supposed to prevent new oppressions from being imposed. Even regarding strict law, the condition of servus casatus differed considerably from pure slavery. From an economic standpoint, the use that was made of his work did not correspond at all to the ordinary definition of slave labor. Moreover, the way of life of many slaves soon changed from the classical mode; their very number rapidly diminished. To approach this phenomenon, let us put ourselves in the ninth century. Dappled with light, or to put it better, with half light, between two great darknesses, this century offers us, in its manorial rent rolls, the elements of statistics that are still very imperfect and quite fragmentary. But neither the preceding nor the following centuries can give us the slightest equivalent. We really have no enumeration of the number of slaves who were not chasés. Some texts—the accounts of the abbey of Corbie, or the abbey of Notre Dame de Soissons—enumerate the serfs who received the day workers’ wage. But as they are preoccupied above all with the order of distributions, they neglected to make note of the differences in status of the people who took part. On the other hand, as far as the slaves who were chasés is concerned, the information is as clear as we could wish. On the lands

484

Bloch

of Saint-Germain-des-Prés, near the end of the reign of Charlemagne, lived (here there is a lacuna in the text) tenants of every sex and age, only (here there is a lacuna in the text) were of servile status. On the lands of St. Remi de Reims near the middle of the century, the proportion was (here there is a lacuna in the text). Doubtless these data are only valid for Gaul and Italy. Definite indicators permit us, however, to assert that in Germany the situation was analogous. As for England, if we want definite figures, we must look later, at the time of the Domesday Book, that is to say, the year 1078. Since the evolution of English society seems to have been substantially behind that of the Continent—where in 1078, as we will see, there were practically no more slaves to be found— this disparity in the dates of the documents is not terribly inconvenient. The Domesday Book only enumerates a total of … (here there is a lacuna in the text). Reduced to these data, there is nothing that would allow us to affirm that as time passed the classes of slave-tenants were thinned out. This leaves the door open, in fact, to another interpretation—that these classes might always have been scanty. But let us take our observations further. On the manors of Frankish Gaul and Italy, the greater part of the land given over to small farms, depending on the central domain, was cut up into indivisible tenures that were generally called mansi. These were not at all of the same sort: there were different categories alongside one another, each subject to its particular obligations. The most widespread classification took as its point of departure the personal status of the tenant. According to whether he was a slave or a free man, the mansus—to confine ourselves to this unit—was called servile or free, and taxed accordingly. At least that was the original principle. After a time that, for reasons discussed below, we can say coincided with the fall of the Roman Empire, this exact parallel between the condition of the man and that of the land ceased to be maintained. Whatever became of the legal situation of the tenant, the mansus henceforth kept its original status, free or servile according to the case, and remained bound by the obligations that this term expressed, with the result that the distribution of mansi of different types remained as a geological testimony to a long-vanished distribution of people. Now, on the lands of Saint-Germain-des-Prés in the ninth century only (here there is a lacuna in the text) servile mansi were really in the possession of slaves; on those of Saint Remi de Reims the figures are (here there is a lacuna in the text) on the one hand and (here there is a lacuna in the text) on the other. Does this raise the hypothesis of a simple coming and going of holdings among groups of people, each one of which remained of equal importance? In fact one does come across free mansi that have passed to slaves, but there are far fewer of these: (here there is a lacuna in the text) out of (here there is a lacuna in the text) at Saint Germain (here there is a lacuna in the text), at Saint Remi. Obviously, it

How And Why Ancient Slavery Came To An End

485

was the number of slave-tenants, on the whole, which had diminished. At this time many portions of land that had been formerly assigned to them were now occupied by free men. What had happened? It would be absurd to think that among the tenants some mysterious psychological decay had set in among the slaves and the slaves alone. Certainly the free men who farmed the mansi originally created for the slaves were for the most part the direct inheritors of the original tenants. But at a given moment the family had received its freedom. Since no necessary relation existed any more between the status of the land and that of its possessor, doubtless among the tenants of mansi belonging to free men, there slipped in, besides slaves still attached to their servitude, the descendants of slaves who had been freed at some former time. Also, the texts themselves, in spite of terrible gaps, remind us that there were a great many manumissions that were extended to very large groups of people during the epoch of the barbarian kingdoms. Not only is there scarcely any type of act of which the collections of formulae used by notaries offer us more examples. We also know of enough examples taken from real life so that we can be sure that they were frequent and widespread. How did it come about that so many slaves had thus received their freedom? The necessity for answering this question leads us to introduce a factor whose influence on practices is always infinitely difficult to evaluate: representations of a religious order. For once, a favorable circumstance is actually going to simplify our task. On the threshold of the Middle Ages, we are fortunate to find no longer before us a religious doctrine in the process of forming, with all the contradictory movements that this stage never fails to involve. From this time onward, Western Christianity had determined its positions on slavery. Just as they were at the time of the great councils of the Peace of the Church, or when Gregory the Great was writing (and in spite of the formal modifications introduced by the rebirth of the hard social philosophy of Aristotle), we find them inspiring the thought of Thomas Aquinas, of Luther, or of Bossuet. The problem had two aspects depending on whether one considered the sources of slavery or the institution as it was already formed. One could not avoid asking under what conditions, if ever, it was legitimate to reduce a human creature to servitude. Once this first difficulty was resolved, the existence in society of many slaves who for a long time and frequently by inheritance had been attached to their condition, remained an undeniable reality. In the face of this fact, what line of conduct ought to be adopted? Let us return to the first point later. Towards slaves subjugated from this moment onward, the attitude of the best authorized religious opinion can be summed up in several concise precepts that follow.

486

Bloch

Nobody doubted that slavery in itself was against divine law. Were not all men equal in Christ? In this primordial thesis, pagans converted to Christianity could recognize an idea that their own philosophers and legalists had made familiar to them, and that, at any rate, had not been without influence on Christian thought itself—except that where the Church talked of “divine law,” paganism had said “natural law.” The parallel was so close that from the Carolingian era onwards theologians tended to identify the two notions with one another. We must avoid underestimating the practical value of the principle of equality thus proclaimed. But even if it could lead to the better treatment of individuals, even to their treatment in a fashion that contrasted with the classic use of slave labor, of course, it must not attack the institution itself at its foundations. Taken literally, the entire social edifice would have crumbled; all hierarchies and even private property were struck by the same theoretical condemnation. Doubtless before God, the slave was the equal of his master, just as in full conformity with the lessons of the Church, the Emperor Louis the Pious said in a capitulary that he was the equal of his subjects. However, no masters thought of abdicating their authority any more than the sovereign did and no one asked it of them. Natural law had always been conceived as subject to the correction of the particular laws of each state. As for the accommodations that divine law was obliged to bend itself to, theologians had learned to justify them by the myth of the Fall, from the first few centuries ad onward. Divine law had only reigned on earth before the great tragedy of the ancestral couple, and all the faults of society followed from the original sin. “It is not nature which has made slaves,” wrote Abbot Smaragdus of Saint Mihiel under Louis the Pious, thus mixing the two vocabularies, pagan and Christian, “rather it is the Fall.” And already in the sixth century, Isidore of Seville wrote, “Slavery is a chastisement inflicted on humanity by the sin of the first man.” The thought of Saint Augustine, which was penetrated by dualist elements until after his conversion, dominated the Middle Ages, whose religion, even though maintained in the careful ways of orthodoxy, never repudiated some Manichean strain, notably in the conception of the Devil. Only the City of the Devil is of this world: the City of God is of the beyond. And actually, all ideology aside, that was the deep feeling that reigned in people’s souls. Since this life is only a place of transition and by definition evil, and since the great task here below is to prepare for Eternal Life, attempting to reform the established social order from top to bottom in the hope of bringing about the triumph of a happiness that was in itself impossible could only be a vain undertaking. Even more, it would be a sacreligious waste of forces that ought to be reserved for a higher and more urgent task. Whether in thought or in action, one should not lose sight of this mystical background to whatever weighs on the medieval

How And Why Ancient Slavery Came To An End

487

mind. Not all consciences were at all equally sensitive to it, nor did all perceive its presence with an equal intensity in all moments of life. But it nonetheless constantly gave the fantastic and fleeting characteristic of a stage set about to fall away to realities that seem to us essentially the concrete matter of our endeavors, such as society and nature. Of course, this is not to say that in traditional instances the practice of Christian virtues was not strongly required. But each condition had its peculiarities and one’s duty was to accept the particular case. The word of Saint Paul remained the law of the Church. By the same token, the legitimacy of slavery was recognized. It seemed so evident to Saint Augustine that in encountering on his way the rule of Hebrew law according to which a slave of the Jewish religion had to be freed from his condition after six years of servitude, he had a great deal of difficulty in explaining how the new law prevented its application to the Christian slave. The councils of the Frankish era confined their ambitions to forbidding the export of slaves—especially their sale overseas, that is to say to Muslims or pagans—and to forbidding Jews to possess or trade in Christian slaves whose faith had to be protected against possible conversions. Also, individual members of the clergy, as well as the Church itself, which was a very large landlord as an institution, possessed a great many slaves. There is no doubt that in some isolated cases, the clergy drew the more difficult conclusions from the notion of original equality. Care was taken to condemn those who did so. In 324, in a canon that Western compilations would unfailingly reproduce, the Council of Granges proclaimed, “If anyone, under the pretext of pity, leads a slave to despise his master, to remove himself from slavery, to not serve with good will and respect, let him be anathematized.” Practical life presented priests with affairs of conscience, and ecclesiastical authorities gave them solutions conforming simultaneously to Christian charity and to the established order. For instance, Raban Maur was asked if it was allowed to say masses for a fugitive slave who dies during his escape and hence in a state of sin. Of course, replied Raban, but one had to remember that as long as the slave lived, the preachers of Christ had an obligation to exhort him to return to his master. Finally, in 916, the Council of Altheim, in referring (inexactly) to a text of Gregory the Great, did not shrink from making a parallel between the slave who fled his master and the churchman who abandoned the Church, striking them both with equal anathema. Moreover, the very existence of sizeable masses of slaves placed a delicate problem before the Church. Should one allow them into the priesthood? The question does not seem to have been raised before the fourth century. From the moment when it was, the response seems to have been what it was always

488

Bloch

to remain: uniformly negative. The principle of equality had bent before considerations of discipline, which the clergy could not defy without betraying its mission. How could a man whom the law placed under the aboslute domination of a master maintain the independence necessary to those who dispense the sacraments? The danger was all the more felt because despite repeated prescriptions of popes and councils, the ordination of slaves continued to take place in practice from time to time, and the troublesome consequences were thereby constantly obvious. This care for dignity, if not a horror of some original blemish attached to servitude, was so much the true motivation for the prohibition that we can see it equally applied in the Merovingian kingdom to coloni, who were juridically free men, but under the strict domination of a large landowner. Also, enfranchisement sufficed to remove the prohibition unless because of the very conditions of the act the slave remained in a state of too rigorous subjection to his former master. It was nonetheless true that in thus barring slaves from entering the orders, the Church once more strengthened slavery. However, it was no small thing to have said to the “tool with a voice” (instrumentum vocale), as the old Roman agronomists called him: “you are a man” and “you are a Christian.” This principle inspired the philanthropic legislation of the emperors during pagan times as well as after the triumph of the new faith. The Church did not forget this. After all, the maxim of Saint Paul was two-edged, and was directed at masters as well as at slaves. Of course, we do not know very well to what extent masters heeded the urging, and if we go by the texts of the councils and the penitentials, the churchmen’s efforts to remind the forgetful of it do not seem to have been very well upheld. In the ninth century, Regino of Prum bade the bishops to pay attention, on their pastoral rounds, to the conduct of those who possessed slaves, but it was only to exhort them to deprive of communion, for two years, those who had killed slaves without a trial. Ordinary maltreatment evidently seemed to him unworthy of attention. Slightly earlier, in Great Britain, the so-called Penintential of Theodorus, renewing in a way the Roman legislation concerning savings, forbade the master to take away from the slave the money he had earned in payment for his work. This was a significant symptom of social evolution which tended to assure a feeling of economic independence to servile labor. But all that did not go very far. A much more important fact was the recognized religious validity of marriage contracts made by slaves. By this, ecclesiastical legislation consolidated orderly households that multiplied the necessities of everyday life on the great estates. It thus gave its aid to the general movement that transformed slavery. Above all, manumission, which pagan morality of the previous few centuries

How And Why Ancient Slavery Came To An End

489

had always considered a gesture of pity, passed the status of being an act of piety. Since God had originally created all men equal and since, in addition, Christ had suffered for all men equally and at the price of his blood had manumitted them from the servitude of Original Sin, to bestow liberty was for the master not a stern duty but at least an infinitely commendable act whereby the faithful, raising himself so far as to imitate the perfect life of the Savior, worked for his own salvation. In fact, if we are to believe the preambles of the charters of “manumission” which the barbarian epoch has left us, no other motive had inspired their authors. Everyone knows that this phraseology must never be taken literally. The reasons that a man gives, publicly, for his acts are not always those he obeys in his innermost heart—far from it. Also, faced with a particular problem, the enfranchisement of its own slaves, the Church had to concern itself with putting an end to untimely acts of generosity. Its goods were in principle unalienable, and it was not the task of one of its temporal administrators to dispose of them even though this was to satisfy his concern, in the selfish sense, for his own salvation. Two conciliar canons, cited by the Regino of Prum and thereafter ceaselessly repeated, forbade a bishop, unless he first recompensed the church for the loss, from his own personal goods, to free its slaves, and forbade an abbot to grant freedom to those slaves who had been given to his monks. But it would be childish to deny that the idea of the world to come, of its penalties and rewards, had contributed to the inspiration of more than one manumission. Among the manumissions we have record of, many are parts of the terms of wills. The practice was already in favor during the Roman period, but there is no doubt that Christianity had greatly contributed to its spread. At the moment when anxiety about the world to come constrained with unusal force the soul about to depart, when a man also considered with more indifference than in the past those temporal goods that he would no longer enjoy, it was natural that even at the expense of his heirs, the rich slave owner thought about procuring for himself the benefit of a final act of charity. But would these considerations have been enough to overthrow the social order? This is the more difficult to believe because while to liberate slaves was indisputably a good act, to keep them under one’s domination was not, after all, an evil one. Manumissions were a good deed; this is indisputably the truth, but in itself this cannot explain their frequency. If the frequency of manumissions at this point was considerable, it is because, as well as being a good act about whose nature slave owners were far from indifferent, the freeing of slaves constituted an operation from which economic conditions of the moment had removed all danger, revealing nothing but its advantages.

490

Bloch

Apparently the right of manumission during the barbarian era was extremely complex. The forms of the act not only varied according to the countries, they presented a great diversity in their content. This is because the heritage of several juridical histories weighed upon the societies of these times. Sometimes people used procedures that came from the old customs of Germany. Sometimes, even though one might personally be from Germany and under German law, one had recourse to procedures of Roman law, which was always alive in the former Empire except in Great Britain, which unified the contributions of classical legislation and doctrine with new practices spread by Christianity. Each tradition offered its own procedures, which were more or less transformed over time. Moreover, the Latin tradition had a stereotyped phraseology that across the ages was to be applied, after the slaves of the high Middle Ages, to the serfs of the following centuries. But if we ignore the details of the juridical practices, we can attend to the concrete consequences of the act, and we see the different types grouped into two large categories. The collection of formulae distinguish them under the expressive names of manumissions with or without obedience, manumissio cum or sine obsequio, the only distinction which was really important from the point of view of the social structure. It could happen that the master, in making a free man of his slave, discharged him forever from any obligation to him. He opened to him, as certain acts put it, the four ways of the world. This case was rare. Neither the Roman nor the Germanic tradition was favorable to it. In Rome, not only did the offspring of the freed man have to wait until two generations had passed before having access to the rights of a citizen; the practice of the slave owners ordinarily kept them dependent on the author of the manumission and his successors, nearly indefinitely. Among a great many German nations-Franks, Saxons, Lombards, Bavarians—the slave, freed from servitude, could not thereby simply become a member of the tribe. He remained lite or alduin, from father to son, confined in an inferior legal status and, at the same time, attached to his former master for posterity. In order to tear him loose from this status, if one thought it fitting, a new manumission was required. Also, in a troubled society such as that of the barbarian kingdoms, within states where the protection, theoretically exercised by the central power, seemed in practice quite far away, almost negligible, absolute independence would most likely turn out to be other than an advantage. Hence, a man (mostly the common man) found it less difficult to accept a master than to live in fear of finding himself without a defender. And for whom would this isolation be more fearful than for the former slave, who was without any legal family?

How And Why Ancient Slavery Came To An End

491

A Lombard charter, giving testimony to the freed slaves themselves, puts these words into their mouths: “Vulpo, Mitilde, their sons, their daughters and their offspring have said that they did not want the four ways and were satisfied, for their future freedom, to receive it under the supervision and protection of the priests and deacons of Santa Maria Maiora of Cremona.” Formulae of this sort are rarely completely sincere, but we can believe that in this case the writer was not too unfaithful an interpreter of Vulpo’s mind and the minds of many like him. The interest of freed men, like that of the owners of slaves, contributed to the general application on manumission “with obedience,” an obedience that, of course, assumed a hereditary character. Sometimes the former master himself kept the benefit of these new powers; at other times, he gave his rights to a third party that, more often than not, was a church. Already an act of piety in itself, a manumission became, in this case, doubly so, since it accompanied a donation to the servants of God. The obsequium not only consisted of a general duty of subjection, with more or less vague outlines and a promise of support in return, ordinarily it comprised very precise obligations, at times specified by the act of manumission itself, at others prescribed by a group custom known to all. Traditionally, the lites, the freed men of Frankish law, and probably their cousins under other Germanic legal systems, paid to their lord, one by one, a sort of annual tax in kind or in money, fixed once and for all. In barbarian societies, this taxation slowly became very widespread. It became the custom to demand it of all kinds of dependents, above all of freed men, without distinction between the modes of their manumission. Often the lord kept a certain part of the inheritances for himself, and at times he levied a tax on the occasion of marriages. Most important, the freed slave had most often been a tenant farmer after the time of his slavery. Having left his servitude, he naturally kept his landholding, subject to the customary obligations. This is why freeing a slave was at times expressed in terms of making a colonus out of him, that is to say a tenant farmer, who was free but still very strictly subordinate to the master of the land. Now certainly, in the Roman world as in Germany, masters had always been able to obtain various benefits, on the order of material revenue as well as social influence, that were reserved for them by a liberty that was so judiciously granted and bestowed through their efforts. It is hardly necessary to recall the part that freed men played in the patronage of Rome, the mainstay of the aristocracy’s power. But the particular conditions of the High Middle Ages made these advantages more evident than ever, at the same time as they tended to attenuate the inconveniences that had for so long been their counterpart. More and more it was in the indirect form of rents and duties that people were inclined to make use of servile labor. Once free, the tenant farmer did not in

492

Bloch

practice produce less than when he was a slave. Taxation, levies on legacies and on marriages, evidently compensated for the services that one ordinarily only asked of slaves. Did the master sometimes ask more for manumissions than that the freed men no longer cost him anything? Did they perhaps put a price on freedom? The slaves, eager to see the legal and moral barrier between themselves and free men lowered, possessing in addition, if they were casati, a small personal hoard, were probably capable of purchasing a thing that was so precious in their eyes. In fact, charters of manumission do not seem to have ever mentioned any price paid. But the example of what later passed for slavery teaches us that in this kind of action, in which people wanted to preserve the appearance of a pious act, people were extremely reluctant to mention anything so mundane, even though it corresponded to reality. It is, therefore, possible that manumissions sometimes provided their granters the chance to add a supplementary increase to the periodic rents that they levied on the man and his land, to be collected once. However, we could not say for sure if this is how it was, nor could we determine any kind of frequency with which it happened. What is certain is that, from these real or seeming acts of generosity, the master derived benefits of another kind which, even though they had no monetary value, were nonetheless held in great esteem. In the society of these first few centuries of the Middle Ages, which was in many ways anarchic, relations of personal subjection had taken on a great deal of importance. The price of power and prestige was to group around oneself a large number of dependents—not slaves but free men who could sit in juridical assemblies and who were qualified to appear at the host. Manumission “with obedience” became foremost in the greatly varying range of relations between lords and followers, and even gave some of its characteristics to several of these relationships whose original reason for existence was different. Therefore, everything tended to increase their number: ordinary self-interest, the desire to be a leader, and concern about the life beyond. It is no surprise, then, that there were so many of them. Thus, slavery was like a reservoir that constantly emptied itself at the top, at an accelerating pace. For slaves no longer to exist, however, these losses had to stop being replaced, or at least had to be attenuated at the bottom. In other words, the very recruitment of the slave population definitely had to dry up. For make no mistake—if masters no longer had the same interest as before in the maintenance of large groups of slaves, they still employed domestics, household artisans, and boys and girls as farm workers on their lands, not to mention concubines. They were inclined to seek these types of personnel from the slave trade. It is probable that this trade, after the Carolingian era, slowed up its supplies. But only the following centuries were to see, if not its complete

How And Why Ancient Slavery Came To An End

493

cessation (which never took place during the Middle Ages), at least its diminution, to the extent that in most of Europe, slavery was practically to disappear and, even where it remained, was to become reduced to the role of a fairly insignificant supplement as a source of manpower. Now that we have come to this point of the evolution, we must first warn of an ambiguity, to which the sources contribute only too well. Let us leaf through the charters of the eleventh and twelfth centuries. Often we will see this old word servus designating a class of men, a word that dictionaries deservedly have taught us to translate as “slave” insofar as ancient texts are concerned. The vernacular languages, for their part, speak of serf in French and servo in Italian, and their agreement with the Latin of the acts assures us that the scribes, always to be suspected of an excess of purism, while continuing to apply a classical term to the reality of their time, were at least in this instance, only conforming to everyday usage. More generally, under different names—there is nothing so variable as the juridical vocabulary of the Middle Ages—a whole population of humble people, much more numerous than the servi of the Frankish era had been, appear in these texts as if deprived of “freedom,” as if plunged into servitude. They did not exist in certain countries where the development of the social structure had followed completely other ways, such as in the low lands that, from Frisia to Dithmarschen, lie along the shores of the North Sea, or in the Scandinavian peninsula. But they were spread over some vast expanses in Germany and Italy as well as in France and England. Within this population, groups of people from very diverse origins had come to mix. Many of the descendants of free coloni had joined there with the descendants of slave-tenants and of many manumitted “with obedience.” At times this was after the express giving over of oneself, supposedly spontaneously, but which most had consented to only under the pressure of threats, from the need for protection, or more simply from hunger. At other times, and doubtless more often, it was the effect of a slow slipping process. This group, moreover, had by no means an absolute juridical uniformity. In Germany, especially, we can distinguish in it many subclasses, defined by different juridical characteristics. But one common characteristic dominated these shadings of difference: the strongly felt separation from men who were qualified as “free.” Had slavery, then, made immense progress from the ninth century onward? Certainly not. What had changed was the very content of the notion of freedom and servitude, which is only its antithesis—to the extent that a whole crowd of men who would have passed as free came to see their condition thenceforth considered as unfree. Constant warfare, the breakup of groups of ancient lineage or those thought to be so, and the already palpable weakness of public power had greatly

494

Bloch

enhanced relations of personal dependence in societies emerging from the invasions. The definite weakness of the state after the eleventh century made them, for several centuries, the only social adhesive that counted, kinship being thenceforth confined to a small circle of kin. It was natural that in order to establish the place of an individual in society, one was above all concerned about the nature of the particular subjection in which he was inevitably held by another more powerful than he. Now, among the originally manifold human relations that people saw beginning or continuing, two categories came rapidly to be distinguished. One was that of relations established by man, in principle, by his own will, even if he could not break them in his lifetime. Such relations in the upper classes as those of vassalage, whose transmission in practice from generation to generation was very frequent and nearly the norm, were never obligatory according to law. Others were those he found already formed as of his birth from his mother’s womb, to which he could only conform until his death, and whose inescapable burden he then in turn bequeathed to his children. Among the peasants, there were always, in every country, those who while renting their fields from a lord and living on the territory where this chief exercised his powers of command, thus bound to him with duties that were frequently very burdensome, were however his subjects only by dint of circumstances—possession or residence—that did not touch them as people. If they abandoned their tenure—abandonment was legal for them—the attachment was totally broken. These hôtes, manants, and vilains (Gäste, Landsassen were the terms in Germanic law), were considered to be free. In the opposite sense, there was nothing so common as the case of the descendants of freed men. Their fathers had been granted a freedom conditioned upon obedience. Since submission accepted in this fashion was hereditary, their sons would one day cease to be among the ranks of free men. As for the slave-tenants, the erasure of the juridical characteristics of slavery from their condition, which was already apparent at the beginning of the Middle Ages, was nothing but the nearly inevitable result of their actual situation, which put them so far apart from the slave. It is scarcely surprising that people progressively lost the habit of making a place for them apart from the dependents of other origins, whose subordination similarly increased, especially because the institutions of public law, having either disappeared or been profoundly altered, from then on, ceased to maintain the ancient barrier between the free man—the only qualified member of the population, the only subject to recruitment to tribunals and the army—and the slave who was foreign to the city. The current language, which is a faithful mirror of this confusion of ideas where the historian very frequently discovers the struggle between the new and the old, manifested a great loosening of terminology after the Frankish

How And Why Ancient Slavery Came To An End

495

era. Poorly instructed scribes labeled free tenants as servi; even the Carolingian monarchy eventually allowed this. As for Visigothic law, a council in Spain did not hesitate to characterize as “servitude” the condition of the freed man who had a master. But official documents generally kept better track. There was still a semilearned law that with some success could impose respect for the ancient juridical classifications. Later on, the absence of all legislation, of any teaching of the law, of any juridical centralization, favored, on the contrary, their renewal. This extension of the notion of the word “servitude” to all the dependencies that weighed on a man from the day of his birth and because of it, took place, of course, without people’s being aware of it. This can be easily explained because slavery formed in a way the prototype for this kind of relationship. Not that the specific duties of the serf, in their details, dated back to the state of slavery; they were much more often borrowed from that of the man freed with obedience. Heredity itself suggested the analogy. Attached to this condition by a nearly physical destiny, the serf, like the slave, could only expect his liberation from the express consent of his lord. People kept the old term “affranchisement” for this act, and in the details of its formulary it continued to replicate certain of the characteristics of the manumissions of slaves of the Frankish era and even of the Roman age. It was the very social disfavor that was, it seems, insurmountably attached to the epithet “servile” that facilitated its application to any hereditary subject of another man, for this very absence of choice seemed to be a mark of inferiority. All serfs were made to suffer certain civil and ecclesiastical disabilities, by which formerly the slave had been affected, such as the refusal to accept his testimony against free men and the refusal to admit him into the holy orders. The name “serf was a slander that tribunals punished when it was wrongly called. One was above all a serf of a lord, but one was also simply a serf, that is to say, a member of a class placed at the very bottom of the scale of human values. Certainly the subjection that was characterized this way was singularly rigorous. Everyday language gave it full justice when it gave to the serf these names full of significance: in France homme de corps, in Germany, “a man held as property by another,” and in England “a bound man” (bondman). The subjugation was translated as various duties whose total weight was undoubtedly very heavy. However, this serf, so despised and placed in a state of strict dependence, was not at all a slave. He did not present the same legal characteristics, since he could possess land by rental or even as full-fledged property to give, sell, and under certain conditions, inherit; since, also, he served at the host and sat at tribunals; and since above all his obligations were, in principal and except when violated, strictly limited by custom. Even less did he represent the

496

Bloch

characteristics of the slave’s economic order, because his labor power did not belong to his lord. In every country, actually, certain lords persisted for a long time in demanding the right to requisition the labor of their serfs in times of pressing need. After all, among all the dependents of the manor, serfs were bound to a particularly commanding duty to help. On occasion, this pressing need could be formidable, but it did not claim in itself anything more than an implementation that was exceptional by definition, and we cannot see that it was frequently followed by other consequences. Serfs doubtless owed their lords a large part of their time, but—with the specific exception of a serf’s duties—this was most often much less as serfs than as tenants, and at the same rate of levy as their neighbors, when the category of the tenures was similar. Above all, whatever the origin and extent of this obligatory work, it was, like all other duties, fixed in its duration and, at times, in the nature of its use, by the normal customs of each manor. These were customs that, after the Frankish era and perhaps after the Roman age, in practice modified the condition of the slave-tenant. The German Tageschalken, whose work duty was on a daily basis, seem at first sight to have been very close to slaves, and they were in fact very humble people, scarcely casati, almost farm hands, whose holding was seemingly confined to their cottage and its garden. But even there the similarity was only on the surface. Did each Tageschalk, whatever his sex or age, come every day to take his place in the team of workers of the lord? Or rather, was not each household from this class—men, women, and perhaps already grown children—required to furnish one worker every day? The point is not clear, though the second solution, which is the most far from slavery, remains the most probable. Also, such duties were extremely rare. In France, I knew of no such example. Nearly everywhere, the heaviest duties required only a few days of work per week. In a word, serfdom or rather the seigneurial system of which serfdom was only one aspect, scarcely placed a quantity of manpower at the disposition of the lords which remained very appreciable, or even numerous, for very long. The stock of labor that was supplied this way constituted an inextensible supply. The slave had been an ox in the stable, always under his master’s orders; the villein, even if he was a serf, was a worker who came on certain days and who left as soon as the job was finished. The fortunes of words themselves suffice to assure us that serfdom was something quite different from slavery. Servus and the Latin derivatives of the word had, as we have seen, slipped bit by bit into a meaning vastly different from their ancient content. The dissolution of the old social frameworks had also impaired various equivalent terms whose evolution followed fairly different lines anyway: the Germanic Knecht and Schalk, its synonyms, had

How And Why Ancient Slavery Came To An End

497

passed on the Continent to the attenuated meaning of “servant.” In England, the meaning of “knight” had changed to that of “armed follower” and later to “horseman.” The Gallo-Roman term vassus or vassalus, which was originally Celtic and which certainly had formerly designated a slave, came to apply to a free retainer, our term “vassal.” In their original, pregnant meanings, these terms were to be replaced by a new semantic creation: this was, in variable phonetic form, the very word that in modern French provided the word “slave.” The word’s history itself is very unclear and lacking in sufficiently advanced studies. It seems to have appeared in the tenth century, about the same time in Germany as in Italy, and from there to have spread through the rest of Europe fairly slowly, for reasons we shall see. Naturally—and this is one of the reasons that makes this development so difficult to trace—it was adopted much more quickly in everyday dialects than by the Latin of the charters or the chronicles, which was essentially conservative and obstinately attached to classical usage. Its introduction into everyday vocabulary evidently marked the difference between the new model of serfdom and a condition that according to the ancient schema really made a man an object belonging to his master. It also attests that alongside the serfs there still existed slaves in the full sense of the word. Moreover, these two conclusions do not exhaust the questions we can ask of linguistics. The word “slave” became only secondarily the label of class. It was primarily an ethnic name. This, too, we have kept in French, with its original meaning, but under a slightly different form: when talking of peoples we speak of Slavs. Certainly the alteration of the meaning is explained by the origin of a large number of men subjected to this condition. We can say more generally that if there were still some slaves in Western and Central Europe in the eleventh and twelfth centuries, these slaves were nearly all foreigners. Actually, this fact was in no way foreign to the most deeply rooted customs. Aside from some exceptional cases, such as penal slavery, slavery for indebtedness, or, where the power of the father was the strongest, the sale of the child by the head of the family, neither the Mediterranean nor the German world had ever considered the enslavement of a fellow citizen as permissible. Most slaves were either captives or the sons of captives who came from nearby societies that were fragmented by a crowd of little tribes or, under the Roman Empire, who were fetched from the vast reservoir of its barbarian frontier. What was new was that in the Middle Ages the notion of “foreigner” took on a different, a completely confessional meaning. States had been broken up into an infinity of tiny pieces. But above them, and encompassing immense human masses, a new city had been born, the civitas Christiana, Christianity, all of whose members belonged, in a moral sense, to a single nation. Of course, peace did not by any means reign there:

498

Bloch

far from it. However, the law of the Church and, more deeply, the religious conscience, did not allow a conqueror to reduce the conquered to slavery when the conquered was his brother in Christ. Here we are touching on the strongest action that Christianity had ever exercised, in a way that was actually somewhat indirect, upon the progress of human liberty, and perhaps upon the social structure in general. Even while favoring manumissions, it had not condemned slavery; it never ceased to accept its imposition upon pagans, infidels, and even upon schismatics within Christendom who were considered to be deprived of true Christian communion. It tolerated keeping baptized slaves under the yoke, if only because they themselves or their ancestors had been foreigners to the true faith at the time of their enslavement. However, in limiting the area where masters and slave traders could legally supply themselves to the space beyond the boundaries of the Catholic world, if it did not entirely dry up the recruitment for slavery, it at least reduced the source to a very thin trickle. Of course, practice only slowly conformed to these feelings, whose influence naturally came into conflict with the old traditions of the razzia and the urgings of the spirit of gain. There is nothing to indicate that the slaves called barbaricini—probably Sardians—of which there was a great deal of trading in sixth century Italy and some of which Gregory the Great purchased, were not Christians. If Smaragdus, the Abbot of Saint Mihiel, thought he had to urge Charlemagne to forbid “further reductions of people into captivity” in his Empire, it is doubtless because acts of this sort had passed before his eyes. Even when the Anglo-Saxons had become Christian, internal conflicts that often set one of their various kingdoms in opposition against the others continued nonetheless for many years to supply captives for the market, sometimes captives of high birth. Also, England was definitely the country where slavery, properly so called, kept an important place in economic life for the longest time in all of Western Europe. A continual guerilla war raged there between Saxons and Celts, who were Christians of course, but whom adversaries willingly considered foreign to Roman orthodoxy. It is not an accident if we can see so many slaves and freed men there carrying Celtic names or surnames such as “Scot,” or if, according to usage that perhaps goes back to the time when the Germanic conquerors of the island still practiced paganism, the word “wealth” (meaning Welsh) came to take on, in everyday speech, the meaning of “slave.” Thus fed by the very close presence of a population whose enslavement appeared legitimate, Anglo-Saxon slavery also owed its long duration to the particular characteristics of the society’s general evolution. The development of the relations of personal dependence, in a regular system that could be substituted almost

How And Why Ancient Slavery Came To An End

499

entirely for other social ties, was much slower than on the continent, and did not achieve its full development before the arrival of the Norman kings. The institutions of Germanic public law marked a greater power of resistance. The traditional classification of human conditions—maintained, besides, by the continuity of the written law and of legislation—fell less quickly into disuse. Indeed the Anglo-Saxon theow was fundamentally that juridical entity that had become so exceptional on the other side of the Channel: a slave. Serfdom, which there as elsewhere had to absorb so many former slaves, was scarcely established before the Norman conquest, no more than was the regime of the fief and vassalage—and the coincidence is significant. War was not alone in providing the primary material for slavery: we are told that often fathers put their own children up for sale. The trade retained a large scale of activity. At the time when French clerks were editing the Domesday Book, people remembered very well the time when, in various internal markets, people “sold men.” But the island also exported large numbers of boys and girls in the tenth and eleventh centuries to Italy, Ireland (notably from Bristol), and perhaps Spain. The girls were sometimes fattened ahead of time to increase their market value. Similarly, after they were converted, the seagoing hordes of Scandanavia continued to sweep the winding and barbaric coastline of the Nordic countries to gather human booty there. Frisian laws, as late as the eleventh century, considered such misadventures as a normal event whose juridical consequences had to be foreseen. However, the movement of new ideas slowly made its way, here sooner and there later. It is very characteristic that the wars that tore asunder the Western world after the last years of Louis the Pious scarcely seem to have led to reductions in slavery. Without doubt, servile livestock remained among the merchandise most actively traded in this time of fairly infrequent exchanges. People not only traded from country to country inside Latin and Germanic Christendom, they also traded outside, but it was no longer people of Catholic birth who were sold this way. From what countries outside Europe (in the restricted sense in which we use this word) did people draw for these foreign slaves, either to keep them or to make money from them? At the time we are studying, many and probably most arrived by land from the eastern confines of Germany. It is well to remember that Germany, at this time, was far from being its present size. In the East, until the tenth century, it scarcely went beyond a border generally marked by the Elbe, the Saal, and the Böhmerwald. In the course of this century the conquests of the Saxon dynasty only ended by extending the line of defense—definitely not what one would call a border—from the Saal to the middle reaches of the Elbe, and to enlarge slightly the zone of protection for the outposts that held its flanks on the right

500

Bloch

bank of this river. Beyond that began the world, hostile more often than not, of the Slavic-speaking peoples. Two indigenous states had been formed there from the tenth century onward which were quite vast, relatively well organized, and in addition were Christian from quite early on: Bohemia and Poland. Outside of these, though, and for the most part between Germany and Poland itself, large expanses remained, occupied by a scattering of peoples who had remained almost entirely pagan, and who were ceaselessly at war among themselves and especially with Christians. This land, blessed as it was with warring raids, was an admirable reservoir for slaves. The lords of the German markets not only used their captives for their own benefit, making them into agricultural workers, tenantfarmers, domestics, and concubines, they also sold them and drew substantial profits from this trade. The Slavic princes who were already converted were similar: tenth century Prague, according to Arab travelers, was one of the great markets of the slave trade. The supply from these far off countries fed a commerce of wide scope. In fact, it does not seem that within Western Christendom itself the area of expansion spread much beyond the Rhine. In France I do not think I have ever come upon them or seen any mention of them in the texts. We should not be so unwise as to conclude from this that nobody owned any. This would deny the possibility of new discoveries, which are that much less improbable because research is difficult and because people have not sought very extensively. Even supposing that the silence of the documents was not to be remedied, neither their abundance nor their precision, unfortunately, are sufficient for isolated facts not to have passed through their meshes. It is nonetheless striking that the very word “slave”—a juridical term, but still conveying at the beginning a very strong ethnic flavor—penetrated the domain of the French language very late. It scarcely appeared in it until the thirteenth century, and was then applied to people of an unfree condition who were perhaps Slavs, but who certainly had not been born upon the frontiers of Germany. Everything indicates that the captives of this area, if they existed in France—we know nothing about them—were there, at any rate, in very small numbers. Probably this merchandise, which had to be fetched from too far away seemed to be too expensive. The very rarity of slave manpower had gotten most masters out of the habit of using them, and the society of the time provided other means for procuring needed labor. On the other hand, on the Rhine, until about the year 1200—that is, at a moment when the progress of the German conquest and that of Christianity began singularly to restrict the hunting grounds—one still found girls of the Slavic race in noble households, frequently as slaves. Around the year 1000, men and women, servi ancillae,

How And Why Ancient Slavery Came To An End

501

figured among the imported merchandise that was tolled on the Alpine roads to Italy. Were these people Slavs or were they English? How can we decide? What is curious is that if the French castles scarcely sheltered any captives taken in the German territories, certain caravans that crossed the country from one part to another bore them fairly frequently. For as early as the tenth and in the eleventh century a good number were exported toward Muslim Spain, which was wealthy and was used to the employment of slave labor. Verdun, at the halfway point, was one of the active centers of this trade. Before sending the young boys they had bought across the Pyrenees, the merchants frequently castrated them to raise their price in the estimation of the harem-masters. Others were most probably sold in the Levant. Among the slaves that Venice shipped on her vessels, to Byzantium and doubtless also to Egypt, it is hard to believe that there were none who were from this source. The ports of the Western Mediterranean and of the Adriatic were exporters of human flesh; were they also occasionally importers to their own hinterlands? There is no doubt about it after the middle of the twelfth century. The slave, at the time, was one of the ordinary items in the return cargos from the long voyages to the Orient and Africa. Slavs or Tartars snatched from the banks of the Black Sea, “olive-skinned” Syrians or Berbers, and Blacks of the Maghreb were to come during several centuries and populate the bourgeois households of Italy, of Provence, and of Catalonia with their humble presence. These are certainly the kinds of people whom documents around that time, all from the South, call “slaves”—for the first time in France. By then, though, the conditions of the Mediterranean traffic had profoundly changed. From then on the craftsmen of Western cities exported their products to consumers beyond the seas, and the sales allowed in return the acquisition of all kinds of merchandise, including human merchandise. Prior to this revolution in the economic current, exchanges were not only much less active, the heaviest burden for the West was borne of the rarity of export goods. To procure the exotic commodities for which it had an overpowering demand, it had either to sacrifice its gold or, precisely, its very own slaves. It could hardly import any. In spite of everything, here and there the West doubtless succeeded in procuring some few, with raids on the nearby Slavonic coast helping to supply the market in Venice, which evidently did not send all of its booty westward. At the end of the eleventh century, the Crusades made the great lords familiar with slavery. One of the surprises to the Latins had been to find its practice widespread in Syria. Guibert de Nogent was all the more strongly offended that many of these slaves were Christians. Later on, Beaumanoir thought that he should set aside, in his descriptions of juridical statutes, a place for this servus “of foreign lands,” so different from the servus he knew from his own experience.

502

Bloch

Imitating these foreign examples, some lofty personages, such as Gaudri, the Bishop of Laon, enjoyed maintaining Negro slaves in their following. All that was a very little matter. At war, Europe could demand a small amount of slave labor in the Slavic territories that its soldiers were ravaging. Economically still weak, it could not expect any appreciable numbers of them from international commerce. The countries beyond the Elbe, however, were not the only lands where Europe touched civilizations that were foreign to Christianity. We should consider Muslim Spain, that Spain whose slow recon-quest was one of the great facts of history. Of course, the inhabitants of the conquered lands were not ordinarily reduced to slavery. On the other hand, most prisoners made on the field of battle were, on both sides. Did the Christians export some of these cautivos outside the country? It is difficult to know. They definitely kept a large number for themselves, making them work in the household or in the fields, and submitting them to a condition that was very much, in the literal sense, that of slaves. The Iberian kingdoms, which were to do so much to spread slavery in the New World, had known it all the time on their own soil. Thus Western and Central Europe, taken as a whole, was never free of slaves during the High Middle Ages. But from the ninth to the twelfth century, slaves always remained very few in number, even fewer than they were to be later on, after the resumption of large-scale Mediterranean trade. They were unknown in entire regions, such as France. Even where they were relatively numerous, enfranchisements came rapidly to thin their ranks because the same arguments that in the preceding era had increased these grants continued to be heard. Neither the lords of the European markets nor those of Leon, Castile, or Aragon managed great plantations that could use numerous slave gangs. And we know how easily, by the natural force of things, the slave-tenant ceases to be a real slave. Where the particular conditions of the country permitted the enslavement of captives without infringing the prohibition of the Church, slavery allowed well-to-do people cheaply to satisfy the manpower needs of the household, and on occasion, perhaps, of a workshop. In other instances, it provided the merchant with a convenient commodity that would help him make useful exchanges with foreigners. But as a force of production, slavery did not count any more. A profound and slow transformation of the very foundations of the economic structure took place, but it is important to note that some of its consequences went beyond the economic sphere. Whatever the misery of certain classes and the disdain in which the fortunate of this world held them—a disdain preserved for some of them by the very survival of the epithet “servile,” perverted, as we know, from its ancient juridical meaning—it was certainly

How And Why Ancient Slavery Came To An End

503

remarkable that no man, no real Christian at any rate, could thereafter legitimately be held as the property of another. In breaking with slavery, the Middle Ages, whose social customs were not kind, neither destroyed nor intended to destroy inequalities of fact or birth, but it did give them a more humane tenor, so to speak. More easily apprehended are the demographic consequences. The importance of slave labor in the Roman world and the size of the scope of the trade led to a melting pot of peoples, whose importance can hardly be exaggerated. How many free families on the soil of Gaul, in the fourth and fifth centuries, were descended from slaves, slaves from the most far-off and diverse lands? Deprived of this foreign immigration, societies without slaves have had their blood renewed much less often. From this viewpoint, it is as if European civilization, like many others, became stabilized and closed in on itself in the course of these centuries.

Slavery in Early China: A Socio-Cultural Approach Robin D. S. Yates Introduction1 Over the last two decades, slavery has become the topic of numerous studies to such an extent that one might say that slavery has become an academic industry in its own right (cf. Karras 1991; Miller 1985; Parish 1989). The reasons for this interest in slavery stem from, on the one hand, the work of the so-called cliometricians, led by Fogel and Engerman (1974), who developed new historical and statistical techniques to reevaluate the nature and economics of the slave system from the sixteenth to the nineteenth centuries in the area that became the southern United States, and on the other, from studies of ancient slavery by historians of Classical Greece and Rome (Finley 1968, 1980, 1983). In addition, Western Marxist scholars attempted to develop the scattered and rudimentary remarks of Marx and Engels on the theory of precapitalist modes of production into a rigorously scientific formulation. They have tried to apply Marx’s methodology in general theoretical terms as well as to the analysis of particular social formations in which slavery has been found to exist. Perry Anderson (1974) has been among the most prominent scholars in England to devote attention to the issues. In France, the group of anthropologists around Source: Yates, Robin D. S., “Slavery in Early China: A Socio-Cultural Approach,” Journal of East Asian Archaeology, 3(1–2) (2002): 283–331. 1  An earlier draft of this essay was presented at the Conference on Ancient China and Social Science Generalizations, Airlie House, Virginia, 21–27 June, 1986. I would like to thank Perry Anderson and Lin Ganquan for their valuable comments on that draft. A Chinese translation of an earlier version of this article, “Gudai Zhongguo nulizhi bijiao lishi yanjiu 古代 中國奴隸制比較歷史研究,” was published in Zhongguoshi yanjiu 中國史研究 111 (1986), 21–34. This essay is dedicated to my late teacher and mentor Professor Kwang-chih Chang, who throughout my career, even in his illness, unfailingly gave me wise advice and generous support. I am also very thankful for the valuable suggestions and corrections by Lothar von Falkenhausen, Robert E. Murowchick, and David Cohen. David Cohen also solved many software problems at the final stages of the preparation of this essay. To him I am much indebted. I would also like to express my gratitude to the Killam Foundation and the Canada Council for the Arts, the Social Sciences and Humanities Research Council of Canada, and Fonds pour la Formation des Chercheurs et de l’Aide à la Recherche, Québec, for financial support in the preparation of this article.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_020

Slavery In Early China: A Socio-cultural Approach

505

Claude Meillassoux (1975, 1978) have examined much ethnographic material from Africa. In Western studies of modern China, James Watson (1980a, b, c) has written detailed ethnographic papers and developed more general, comparative theories on the relations between Asian and African systems of slavery. Simultaneaously, Orlando Patterson (1979, 1982) developed his own, highly original theories on the topic of slavery in general. The question of slavery in ancient China, however, was passed over by most Western experts on slavery and even by specialists on China, despite this growing interest in the topic of slavery among historians in general. In this essay, I would like to analyze some of the new evidence relating to slavery in ancient China that has only recently come to light, in the hopes that the discussions of the question of slavery will lead to a better understanding of the economic and social basis of the traditional Chinese state. In addition, I hope to bring the ancient Chinese evidence to the attention of those involved in the international debate about the nature of slavery and its role in the development of world history. Since the volume of studies on slavery is enormous, in this essay I intend to use material relating to the slave systems of ancient Greece and Rome to help illuminate relevant issues in ancient China because of their comparable antiquity and their importance for laying the foundations of Western and Eastern civilizations respectively. By doing this, I do not intend to cast any aspersions on the value of drawing comparisons with other slave systems in other social formations, such as those in Africa and America: I merely wish to limit the article to manageable proportions. In addition, I will concentrate upon the period of the Warring States and Qin China, the period when the foundations of the traditional Chinese political, social, economic, and religious system were laid, and therefore I will not enter upon an analysis of slavery in later imperial times to any great extent. In academic circles in China, the notion that there was, and will be, an inexorable unilineal historical progression of modes of production, from primitive communist, to slave, to feudal, to capitalist, to communist, was generally accepted as valid until recently The most influential of these scholars was the late Guo Moruo (1973) (cf. Liu Weimin 1975) who changed his mind several times on the date when the slave mode of production gave way to the feudal. In fact, much discussion centered round this question of periodization and by the 1980’s the consensus seemed to be that the period of the Warring States to the Qin 秦 was the watershed between the two modes of production. The Qin empire ushered in the feudal mode of production which lasted until the end of the traditional period, with the “sprouts of capitalism” beginning in the late Ming 明.

506

Yates

Yet despite this conclusion by their Chinese colleagues, a growing number of Western scholars, both Marxist and non-Marxist, disagree with this analysis in their empirical studies and theoretical essays. Finley for example, states “[a]lthough slaves have been exploited in most societies as far back as any records exist, there have been only five genuine slave societies, two of them in antiquity: classical Greece and Classical Italy.” He continues by declaring that the other three all were to be found in the New World (Finley 1980: 9). Meillassoux contends that “[s]lavery always appears in association with other relationships of production, so that it does not constitute the only or even the dominant mode of production, and the ruling class cannot be defined strictly only in relationship to it” (Meillassoux 1975: 20–21, quoted and translated by Patterson 1979: 47). Marxist analysis should not for that reason be abandoned, but rather the argument suggests one must “reject any proto-Marxist shuffling of prefabricated Modes of Production summarily borrowed from this or that writing or remark of Marx or Engels.” Indeed, he believes that, as of 1975, “no formal criterion has been brought to light which permits the creation of a categorical distinction between slaves and all other forms [of dependent labor]” (Meillassoux 1975: 21, my translation). Anderson, on the other hand, is of the opinion that “[t]he slave mode of production was the decisive invention of the Graeco-Roman world, which provided the basis both for its accomplishments and its eclipse.” On the other hand, slavery existed in many forms in other earlier and contemporaneous societies but in them it was never “the predominant type of surplus extraction” (Anderson 1974: 21). One of the most difficult problems facing theoreticians is the fact that slavery has existed in many different societies at many levels of socio-economic and cultural evolution. The slave plantations of the New World were part of the developing capitalist mode of production, and tied in to the world market, yet at the same time “[s]lavery is clearly compatible with all types of post-neolithic economy” (Goody 1980). What criteria should be used to define slavery as a mode of production, and how did the idea develop that slavery was one stage in the hierarchy of modes of production that all, or almost all, societies must pass through? Robert Padgug (1976) provides the answer to the second question. He notes that “it has been customary among Marxists, especially those in Eastern Europe, to designate slavery as the first form of class division and to consider the early societies which contained it—and most of them indeed did—to represent varieties of a single type of slave society,” which stood at the beginning of a kind of universal slave stage. It was a speech by Lenin on “The State” that first gave the theory currency and it was accepted by Stalin in the 1920’s and 1930’s

Slavery In Early China: A Socio-cultural Approach

507

to be the basis for academic discussions and analyses of world historical development. From the Soviet Union, it passed into Chinese scholarly circles. Padgug notes that the idea of this universal stage “has gradually been abandoned in the USSR and elsewhere.” But, it should be pointed out, in China it still has its adherents, although the issue of slavery and Marxist theory has been of diminished intellectual interest among early historians since the beginning of the 1990’s. Theoretically speaking, there is no reason that, if the notion that slavery formed a universal stage in history is rejected, it is necessary at the same time to reject the theoretical possibility of a slave mode of production. Indeed, Padgug (1976), Anderson (1974), and Hindess and Hirst (1975: 109–77) all attempt to define the structure of this potential slave mode. Patterson (1979: 49–52) provides some pertinent criticism of the latters’ theory. They claim that a defining feature of slaves is that they are separated from the means of production, but Patterson correctly observes that not only slaves but also workers in capitalist society are so separated, such that this cannot be a distinguishing feature. Furthermore, their assertion that there is no separation between necessary labor and surplus labor apparent in the slave mode of production, he argues, simply is not true. The slave-owner is quite well aware of the difference between the labor input of the slaves and other factors of production, and the costs of maintaining slaves is carefully assessed against the value of their output. In addition, Patterson finds fault with their assertion that a slave is not forced to labor as a worker is in the capitalist system, for the slave would not be fed if he did not work. Physical coercion being so common to the slave’s everyday experience, this would not be sufficient to cause him to exert himself: removal of food would have the desired effect. So in capitalist societies, workers are subject to physical abuse if they refuse to work and this forces them to contribute their labor, whereas it is economic force that makes a slave work. Thus Hindess and Hirst’s formulation of the slave mode of production is open to serious criticism. Anderson, in his compelling and detailed analysis of the ancient GraecoRoman world, argues that slavery became the dominant, but not the only, mode of production in fifth and fourth century bc Greece, when the size of the slave population in relation to the free increased substantially. At that point, “the nature (of slavery) became absolute: it was no longer one relative form of servitude among many, along a gradual continuum, but a polar condition of complete loss of freedom, juxtaposed against a new and untrammelled liberty,” freedom entailed slavery and vice versa (Anderson 1974: 23). From that time on, slavery structured each of the micro-economies of the Greek city-state world

508

Yates

and gave Greek civilization its special characteristics. As Rome conquered the states and peoples surrounding the Mediterranean Sea and enslaved huge numbers of the defeated populations, a similar transformation took place. Slaves labored primarily in the fields of the great latifundia estates, and the surplus that was extracted from them was consumed in the cities, enabling the urban elites to indulge in unparalleled luxury. Ultimately, however, slavery and slave relations of production paralyzed technological innovation and the forces of production reached their limits: the slave mode of production gave way to the feudal. At the height of the slave mode of production, however, slaves were one of the most highly movable commodities in the ancient inter-city market exchange system and the surplus they generated provided the economic basis of ancient society. With these preliminary remarks in mind, let us now turn to the question of slavery in ancient China.

Slavery in Warring States China

In China from the Neolithic to the late Western Zhou 西周 and early Eastern Zhou 東周 periods, there was no strict separation in status between the lowest members of the society and the highest. All members were to a certain extent unfree, and it was only in the middle of the Spring and Autumn period that a market economy really developed. I am not claiming that there were no slaves at all in early times. Rather, I wish to argue that there was a radical change in the nature of slavery in the middle of the Eastern Zhou period, comparable to that discussed by Anderson in fifth and fourth century bc Greece, when slaves for the first time came to be bought and sold. I will explain my interpretation of the nature of slavery in ancient China and the role it played in the social formation in the following pages. But before embarking upon this analysis, I would like to point out that I shall draw much of my information from the Qin legal documents, which were only discovered in 1975 at Shuihudi 睡虎地, in Yunmeng 雲夢, Hubei 湖北.2 These documents provide essential information not available to previous scholars of ancient Chinese slavery. In the

2  My references in this essay are to Shuihudi Qinmu zhujian Zhengli Xiaozu (1978) (hereafter abbreviated SHD, but see also Shuihudi Qinmu zhujian Zhengli Xiaozu (1990). The bibliography on the Shuihudi texts is enormous. I bring the attention of the reader only to the studies by Xu Fuchang (1993), Li Jing (1985), Wu Fuzhu (1994), Yu Zongfa (1992), Xing Yitian (1987); Hori Tsuyoshi (1988), and Ōba Osamu (1991).

Slavery In Early China: A Socio-cultural Approach

509

West, the most notable are Martin Wilbur (1943), Wang Yi-t’ung (1953), Edwin Pulleyblank (1958), and Ch’ü T’ung-tsu (1972). In 9 ad, the usurper Wang Mang 王莽 proclaimed his inaugural edict in which he condemned the state of Qin for lacking the Way (Dao 道). Among the greatest of Qin’s faults, according to Wang, was that they established slave markets that were the same as those for oxen and horses. And throughout the Han 漢, it was believed that many hereditary slaves were the descendants of criminals whose families had been enslaved under the laws of mutual responsibility (lianzuo 連坐) promulgated by the infamous legalist statesman Shang Yang 商鞅 or Lord Shang 商君 in Qin in the fourth century bc (Wilbur 1943: 72–79; Tomiya Itaru 1985; Yates 1987).3 The new laws paved the way for Qin’s successful conquest of its rivals and its founding of the Chinese imperial order in 221 bc, and this form of legal enslavement for high treason continued down to the end of the Qing 清 dynasty. Just what role slavery played in Qin’s unification effort, therefore, is of vital importance for understanding the internal dynamics of the Chinese historical process. A final reason for considering the question of slavery is that the nature of slavery goes to the heart of the nature of social status and the social structure not only of the early period in Chinese history, but also in the later traditional period as well. It even has implications for social status in the twentieth century. Slavery was only formally abolished in China itself in the last years of the Qing dynasty (Philip C. C. Huang 2001: 17),4 but survived until the Communist liberation in 1949 (Watson 1980c; Meijer 1980). Recent studies suggest that the practice is still not defunct today, even though outlawed (Watson 1980c: 245). On the mainland, recent newspaper articles have brought attention to the fact that some unscrupulous entrepreneurs are luring young and naive peasant girls from provinces such as Sichuan 四川 with promises of employment in the cities. But instead of fulfilling their promises, they kidnap them and sell them as brides to peasants in far off provinces such as Hubei, Henan 河南, and Hebei 河北. Slavery goes to the heart of social status by a dialectical opposition: slaves are quintessentially not free and not honored. Through an analysis of slaves, the meaning of the concepts “freedom” and “status” in traditional China and

3  The Shuo wen 說文 dictionary is one of the texts in which the opinion on the origin of slaves is to be found (Xu Zhongshu 1981: vol.1, 652). 4  However, the enslavement of ordinary “good” commoners was illegal under the Qing and punished severely, because the Qing were anxious to maintain the proper hierarchy in society (Philip C. C. Huang 1996: 94–95).

510

Yates

the nature of slavery in ancient China can be better understood.5 Of course, by “freedom” here I am not in any way implying that the notion of “freedom” in traditional China was at all the same as that in the West. The present study will help to explain why.

The Definition of Slavery

1 Slaves as Property Most scholars who have studied the question have emphasized the fact that slaves were property. Watson, drawing on the work of Finley, the anthropological work of Meillassoux (1975; 1978) in Africa, and the sociological writings of H. J. Nieboer,6 has noted that this aspect is primary, while the fact that slaves are marginal, liminal people, or outsiders to society, is secondary (Watson 1980b: 8–9). In addition, Watson emphasizes the fact that slaves’ labor is extracted through coercion. Let us first take up the question of slaves as property. Unlike in Rome and generally in Western civil law, property was not easily alienable in China. There was no concept of absolute rights in rem (in the thing itself) (Patterson 1982: 20). This was a special characteristic of Roman private law, which, by a legal fiction, gave a person absolute rights and control over an object.7 In Rome, the two main sources of wealth were land and slaves, and it is 5  I place the term “freedom” in quotation marks deliberately. As Finley (1983: 119–120) points out, “[f]reedom is no less complex a concept than ‘servitude’ or ‘bondage’; it is a concept which had no meaning and no existence for most of human history; it had to be invented finally and that invention was possible only under very special conditions.” “Freedom,” he continues, was scarcely possible to say in nineteenth century Chinese. 6  Nieboer (1900: 7) defines slavery as “the fact that one man is the property or possession of another man, and forced to work for him.” 7  Ownership of a thing (res) could be passed in two different formal acts, either man cipatio or cessio in iuris (“cession in court”). Mere transfer of an object, such as a slave, did not pass the right of ownership (Crook 1967: 141). Things were divided into two categories, res mancipi such as land and houses on Italian soil, slaves, certain animals such as beasts of burden and of draft (e.g., oxen) and one kind of easement, the rustic praedial servitude, and res nec mancipi, which included everything else. The latter could be transferred by simple delivery (traditio), for example a sheep (Jolowicz and Nicholas 1978: 137). Dominium could not be passed unless the person had the rights of dominium himself. In fact, dominium could be achieved by simple usucapio (“acquiring dominium by use”) which required a relatively short period to elapse before the object fell into the user’s hands, two years for real estate and one year for movable property. As Jolowicz and Nicholas state (1978: 140), “[t]he Roman law of classical times is dominated by what is commonly called the absolute conception of ownership which it has evolved and by the action through which this right is asserted, the vindicatio.

Slavery In Early China: A Socio-cultural Approach

511

quite possible that this legal fiction of absolute rights over things derived from the institution of slavery, and resulted in the definition of slaves as property. Indeed, the Romans made a clear distinction between possession and ownership, and much of Roman property law was based on the concept of dominium, even though this term is never defined in the extant sources of Roman law. Nevertheless, F. Schulz has concluded that “a Roman principle must have existed to the effect that the right of ownership was to be as unrestricted as possible and the greatest possible latitude given to individual action and initiative” (Schulz 1936, quoted in Harrison 1968: 202, note 2). The situation was different in Athens, however, where legal thinking was not as highly developed. There was, in fact, “no general term to describe the law of property; they had no abstract word for ownership” (Harrison 1968: 201), and the sources are too fragmentary to permit more than a sketchy interpretation of the data. Although the right and power to alienate property had emerged by the fourth century bc, where it had been previously extremely limited, still this power was restricted to cases in which the owner lacked a legitimate heir, and full rights of ownership were restricted in several other ways, including those in which there existed joint ownership and where there were easements. Under the law of succession, property seems to have been divided into two parts, that which a man inherited (πατρῷα) and that which he had created or accumulated himself (’επικτητά) (Harrison 1968: 125). The former was hard to dispose of at will, while the individual had more freedom of decision with respect to the latter. Ancient and traditional Chinese law seems to have lacked the notion of absolute rights over things, and therefore it is not adequate to define ancient Chinese slaves purely in terms of their position as property (cai 財) as almost all scholars have done to date.8 They were indeed property but were much more than property. Ownership, in the developed law, may be defined as the unrestricted right of control over a physical thing, and whosoever has the right can claim the thing he owns wherever it is and no matter who possesses it.” In Germanic and English systems of law, there existed only a relative right to possession, and as Patterson (1982: 20) points out, Anglo-American law rejects the Roman civil view “[b]ecause, first, in both sociological and economic terms … there can be no relation between a person and a thing. Relations only exist between persons. Second, relations between persons with respect to some object are always relative, never absolute.” 8  Niida Noboru (1944: 905–914) emphasizes this aspect, but it should be born in mind that the evidence of this definition of slaves as property (caiwu 財物) only derives from Tang and later sources. Whether the newly discovered Han 漢 laws from Jiangling 江陵 throw some light on the early definition of slaves remains to be seen. Wilbur (1943: 63) defines a slave as “a person who is owned as actual property by another person, group, corporation, or the

512

Yates

Hilary Beattie, in her study of Tongcheng 桐城 county in Anhui 安徽 in the Ming and Qing dynasties, points out that “[p]rivate landholdings were of course always subject to fragmentation on the owner’s death, though it should be born in mind that property at this time and subsequently was never freely alienable by the individual owner; his immediate family and other relatives had rights of objection and pre-emption which did act as a brake on the rapid dispersion of estates among outsiders” (Beattie 1979: 8). In fact, not only did immediate family members and relatives have rights over such landed property, but so also did neighbors, and this was true into the twentieth century, at least in parts of China.9 It was therefore the traditional Chinese family or lineage that held the landed property jointly and, of course, in essence, this meant the male members of the lineage.10 Two theories have been advanced by Japanese scholars on this matter: one holds that family property was the sole, personal, possession of the living common ancestor, and the other is that it was held by ascendants and descendants in a direct line. Inferiors and juniors may use the property with the permission of the elders, such that it is not the private property of the elders and superiors, but is subject to the management of the latter.11 state, whose services are therefore controlled, and who is accorded a distinct status as one of a group so owned and controlled.” 9  Rev. J. Hutson (1919: 553), for example in the section “Purchase of Property,” notes that Yo chieh hsia pai (yuejie xiabei 約界下碑) is to go round the boundaries of the property along with the middlemen and neighboring landlords, who have to be invited to a feast; their presence is equivalent to their recognition of their new neighbour.” 10  James L. Watson (1982: 594) defines “lineage” as a “corporate group which celebrates ritual unity and is based on demonstrated descent from a common ancestor.” In the conclusion to his article (Watson 1982: 617–18), he suggests the possibility that Chinese lineages as anthropologists know them today may not have developed until the end of the Tang 唐 and the beginning of the Song 宋 dynasties. Before that time, the aristocratic elites were interested in forming alliances and were particularly concerned with matters of pedigree. They possessed lines not lineages. From the Song on, however, “the balance shifted from alliance to descent.” While this may be true for the later period, it would seem that lineages in Watson’s definition of the term did exist in Shang 商 and Zhou China. Elites traced their descent from a common ancestor, even though in some cases of the major lineages descent was claimed from a mythological ancestor, and most lineages held towns or cities that were the ritual centers for their ancestral cults. There have been many studies of this phenomenon, including Du Zhengsheng (1979) and Kwang-chih Chang (1976). For recent scholarship on the changing rights of women to property from Song dynasty times on, see Bernhardt (1999), Birge (2002), Ebrey (1993), and McKnight and Liu (1999). 11  Schurmann (1956) observes that “some form of joint family property has existed in China for at least two millennia,” and that junior members could not be accused of stealing, only of appropriating common property.

Slavery In Early China: A Socio-cultural Approach

513

I would suggest that in the period of the late Warring States, Qin and early Han, Chinese law resembled that of traditional Africa, in which Miers and Kopytoff have argued that an individual can be seen both as a member of the kin group and as part of the group’s corporate property and constitutes lineage wealth. And, they add, “[n]either criterion of property nor that of salability can be useful in separating slavery from simple kinship in African societies,” for kin members are treated in ways westerners would consider property” (Miers and Kopytoff 1977: 12; Watson 1980b: 4–5). Supporting evidence for my view comes from section 5 of the Qin legal documents, no. 4 “Sealing and Guarding” ( fengshou 封守), where a list of a commoner’s property is given as including wife, adult unmarried female daughter, non-adult male child, two slaves, one male adult and one female non-adult, one male dog, clothes, house, and ten mulberry trees (McLeod and Yates 1981: 137–39; Hulsewé 1985a: E3, 181–85). Again, in section 4, the questions and answers on points of law, we find a definition of the distinction between a so-called public domain denunciation (gongshi gao 公室告) and a non-public domain denunciation ( fei gongshi gao 非公室告).12 The former refers to a situation in which a person with murderous intent kills, wounds, or robs another person, and the latter is one in which a son or child robs his father or mother, and they on their own responsibility (shan 擅) kill, mutilate, or shave their children and male and female slaves (nüqie 女妾) (SHD 4.90, 195–96; cf. Huang Zhongye 1991: 143–45).13 In addition, in 4.91, if a child denounces his parents for a crime or if slaves denounce their master, this too was a non-public domain denunciation, and the authorities were not to take up the case. And if a master (zhu 主) were on his own responsibility to kill, mutilate, or shave his children and slaves (chenqie 臣妾) this too was a non-public domain denunciation, and the authorities were not to prosecute. If the children persisted in their accusations, they themselves incurred guilt and if another third party took their side, again the authorities were to ignore him (SHD 4.91, 196; Hulsewé 1985a: D 87, 148–149). Here we can clearly see that the parents ( fumu 父母), and it is important to note that the mother is included, it is not just the father, are alternatively referred to as masters (zhu 主), and that they have certain rights over their children and slaves. I say “certain” intentionally because of the presence of the term shan “on their own responsibility” in the definitions. Generally speaking, the parents were expected to request permission from the authorities before carrying out the punishment. This gave the children and slaves some legal protection from arbitrary savagery, yet since the parents were not subject to 12  Hulsewé (1985a: D 86, 148) translates gongshi as “official.” 13  The first transcriptions of the Shuihudi legal materials found in SHD were presented unnumbered in Yunmeng Qinmu Zhujian Zhengli Xiaozu 1976a–c.

514

Yates

prosecution even if they carried out the maximum penalty, the children and slaves cannot have derived much comfort from the legal provision. We can also see that the wife was in some ways in an anomalous position vis-à-vis her status as property: on the one hand, she is treated as part of her husband’s property, and, on the other, she has property rights over her children. We find further support for this view in the almanac texts (rishu 日書), where we read on the shouri 受日 (“Receiving day”), the tenth in the twelve day Qin week, that one could bring into one’s house people (renmin 人民) (i.e., slaves) horses, oxen, husked and unhusked grain, take in wives and other things (tawu 它物) (Shuihudi Qinmu Zhujian Zhengli Xiaozu 1978: plate 117, slip 752; Rao Zongyi and Zeng Xiantong 1993: 408; Liu Lexian 1994: 33; cf. Kalinowski 1986; Loewe 1994: 221–26; Poo 1998: 69–101). The data available at present make it very difficult to determine with any precision Warring States conceptions of property. Did the ancient Chinese distinguish between ownership and possession, as the Romans did, or were their ideas more vague, like those of the Athenians? Certainly there is no theoretical explication of any such concepts in the legal documents found at Shuihudi. Furthermore, all the references to land in the documents seem to imply that the state was the sole owner: only one reference to the “people’s fields” (mintian 民田) suggests, but does not prove, private ownership of land (SHD 4.138).14 This is not to say that there was no private ownership of land in the Qin. As is well known, Han scholars frequently claimed that the scourge of the accumulation of landed property in fewer and fewer private wealthy hands derived from Qin land tenure practices and specifically from Lord Shang’s destruction of the “well-field” ( jingtian 井田) system (Gao Min 1979; Xiong Tieji and Wang Ruiming 1981; An Zuozhang 1981; Zhu Shaohou 1981; Lin Jianming 1981).15 New 14  Hulsewé (1985a: D 135, 164) translates mintian as “fields of commoners.” I prefer to use “commoner” in a more technical legal sense of those with the commoner status of “member of the rank and file,” who probably could be awarded or attain up to the eighth rank in the ranking system. Min 民 here might refer to members of the aristocratic lineages of daifu 大夫 rank as well as commoners or to members of those lineages that did not belong to the ruling clan of Qin, if Gassman’s (2000) interpretation of the difference between ren 人 and min is correct. 15  Lin Jianming argues that the early Qin state never had a well-field system of land apportionment, because before its enfeoffment with the old rich farming lands of the Western Zhou, it was a nomadic pastoral people. When it was enfeoffed, the well-field system was already breaking down. The earliest land system was, therefore, that of the so-called “changing-field” (yuantian 轅/爰田), in which the land was divided into three grades according to quality, superior, middle, and lower, and each family was given a fixed area of each, and the allotments and residences were changed every three years. Another

Slavery In Early China: A Socio-cultural Approach

515

evidence demonstrates, however, that this system, if it ever was a reality, was breaking down by at least 500 bc: each of the aristocratic lineages which was fighting to seize control of the moribund state of Jin 晉 had its own land division and land tax system (Zhongguo Renmin Jiefangjun Junshi Kexueyuan and Zhanzheng Lilun Yanjiubu Sunzi Zhushi Xiaozu 1977: 162–63; Guo Huaruo 1978: 494–496). Lord Shang evidently adopted the land division system of Zhao 趙,16 while taxing the peasant population to which the land was either given or lent.17 It would appear from the Qin legal documents that a substantial part of the land was exploited directly by the state, using slave and convict labor, and description states that each family was given 100 mu 畝 of land which was not to be changed, 200 mu of land was to be changed once (every three years), and 300 mu of land to be changed twice (every three years). Thus Lin suggests that this system developed into one in which there was a single allotment which was returned to the state when the individual reached the retirement age of 60 years. In another way, too, the “changingfield” system was different from the “well-field” system. In the latter, all that the peasants (slaves, in Lin’s view) were required to perform for the state was labor in the fields, but in the former, they were required also to perform military service ( fu 賦). In this regard, the system of “changing-fields” was the same as the later tuntian 屯田 “colony” system. Yet in neither system, be it the “well-field” or the early and later “changing-field” system, did the workers have rights or powers (quan 權) over the land, and it all belonged to the slaveowners, and not to the workers who were slaves (nuli 奴隸). Because of the military needs of the early Qin state, it did not institute the decaying “well-field” system but the “changing-field” allotment with its military implications. When Lord Shang reformed the laws in Qin in the fourth century bc, he reintroduced the “changing-field” allotment system but without the practice of changing every three years, and thereby he merely redrew the socalled qianmo 阡陌 field boundaries of the “well-field” system to create his new system.  While this analysis has many valuable points to it, I am not in agreement with all of Lin’s and the other scholars’ arguments concerning the early Qin land system. For a recent analysis of bronze inscriptions concerning transfer of title over land and related disputes in the mid-Western Zhou, see Lau 1999; cf. Shaughnessy 1993).  It is worth noting, however, that a stone stela dating from the Eastern Han dynasty was discovered in 1973 near Yanshi 偃師 county, Henan 河南. This land-sale contract provides solid evidence of one form of joint ownership of land: twenty-five individuals, who are named, formed a special group called the fulao dan 父老僤 of Shiting 侍廷 village, where dan is the name of an association at the village or sub-village level, to buy 82 mu of land for 61,500 cash (Du Zhengsheng 1990; cf. Yu Weichao 1988; Huang Shibin 1982; Ning Ke 1982). For other views on the Qin land system, see Li Jiemin (1981), Zhang Chuanxi (1985: 77–96), and Hulsewé (1985b: 215–218). 16  Zhao had an acre (mu) of 240 paces (bu 步), according to the new Sunzi 孫子 (Yates 1988: 218). 17  The Shi ji 史記 (Takigawa Kametarō 1979: j. 15, 42) records that the Qin began to tax the land in the form of grain (chu zuhe 初租禾) in the seventh year of Duke Jian 簡公

516

Yates

there is no evidence at all of buying and selling of land in them, even though other literary sources suggest that by Later Warring States times, land was being bought and sold at least in some areas by some authorized individuals.18 This is a highly complex issue that cannot be discussed here at any length, for it would take us too far afield. The Qin laws do, however, provide us with some evidence regarding movable property that can be interpreted as providing support for the notion that private ownership existed and that ownership was held to inhere in a private individual. First, there are a number of laws specifying the rules relating to the loan of public or state- (gong 公) owned property to private individuals. These included oxen and tools, which would have been used for production, as well as military equipment. By implication, this suggests that there were articles that were privately (si 私) owned. For example, one of the statutes on Artisans (gonglü 工律), in Hulsewé’s translation, states: Government armour and arms are each to be incised or branded with the name of the office concerned; on those that cannot be incised or branded, it should be written with vermilion or lacquer. When armour and arms are loaned to commoners, it is essential to record the brand-mark; they are to be bestowed according to the brand-marks. When loaned (armour and arms) are handed in and they have no brand-marks, as well as when it is not the brand-mark of the office concerned, (such armour and arms) are all to be confiscated by the government; they are to be charged according to the Statutes on Equipment. Hulsewé 1985a: A56, 59; SHD 71–72

Second, as pointed out above, a commoner’s property could include family members, slaves, a dog, clothes and mulberry trees. In the case quoted, the government confiscates or impounds this property when the commoner is under investigation for a crime. Does this mean that the Qin commoner lost his property rights if he was accused of a crime, that property rights were contingent upon his good behavior? Or was this legal maneuver merely intended to ensure that those responsible for his behavior under the mutual liability (408 bc), and instituted the fu 賦 military service tax in 348 bc under Lord Shang. (cf. Xiong Tieji and Wang Ruiming 1981: 71; Huang Jinyan 1981: 61–73). 18  Hulsewé (1988: 173), observes that only one rule in the Qin law refers to “the misappropriation of land,” but at present it is not possible to determine whether land could actually be stolen or not.

Slavery In Early China: A Socio-cultural Approach

517

laws could be held accountable and that fines could be assessed against him? Further research needs to be carried out on this issue. Third, if we review the data in the Qin legal documents, we can see that certain other types of articles were considered property. These include mulberry leaves, oxen, cash,19 artisanal equipment (possibly government property), offerings from the official sacrifices, including kidneys (these would be presumably considered public property), a goat with a rope round its neck, state money placed in a storehouse, pigs, pearls and jade (apparently state property), grain, a horse, clothes, and shoes, and crossbows and bolts. Fourth, we find the interesting principle that “a father stealing from his children is not a case of theft,” but a foster-father is stealing from his fosterchildren (SHD 4.17, 159; Hulsewé 1985a: D 17, 125). Further, a son could steal from his father and mother (SHD 4.90, 195–196; Hulsewé 1985a: D 86, 148; SHD 4.94, 197–198; Hulsewé (1985a) D 89, 149–150),20 and a wife could hold property separately from her husband. Finally, a slave and slave-woman were considered to have robbed their master when they robbed their master’s parents only when the parents were actually living with the master as household members or co-residents. If the parents lived separately, the slaves were not deemed to have robbed their masters (SHD 4.18, 159–160; Hulsewé 1985a: D 18, 125–126).21 From this evidence, we may deduce that property rights in ancient China were based to a large extent upon a) blood relationship and b) co-residence, at least in Qin. The property of the wife is presumably her dowry and as such was kept legally separate from the property of her husband. In the case in which her husband commits a crime and she denounces him before the authorities are aware of its commission, her property is not to be confiscated by the government (shd 4.151, 224; Hulsewé 1985a: D149, 169). Similarly, if a wife commits a crime for which she is arrested, it is stipulated that her maidservants (ying 媵), slaves, clothes and utensils are to be given (bi 畀) to her husband, rather than being confiscated (shd 4.152, 224–225; Hulsewé 1985a: D 150, 169). The father possessed a kind of patria potestas over his children in the matter of property, just as he had the power of life and death over them, as we 19  There are many examples in the Qin laws, but note especially SHD 5.11: 255–56; Hulsewé 1985a: E12, 190–91; McLeod and Yates (1981): 5.11, 142–44, where a gongshi 公士, a person of the lowest rank in the social hierarchy above a commoner, is robbed of 10,000 cash. 20  The latter passage specifies that the father and son are “co-residents” (tongju 同居). 21  These rules relating to foster-fathers and to master’s parents possibly reflect Lord Shang’s attempts to break the power and solidarity of the family by emending the law in Qin to hold that no more than one adult child could live together with its parents under the same roof.

518

Yates

have seen above. But to define slaves in ancient China just in terms of property is insufficient and we must adopt a more satisfactory definition, if we are to understand the nature of slavery and its full social significance in that society. Before we pass on to discuss a more adequate definition of slavery, I should like to note that under Qin law, parents could not arbitrarily kill their children, for we find another item of the law stating that such murder would be punished with tattooing of the face and the chengdan hard labor sentence for both father and mother (SHD 4.58, 181; Hulsewé 1985a: D 56, 139; cf. Reed 2000).22 It is in the case where a child robs its parents that the latter may kill it in revenge (SHD 4.90, 195–196; Hulsewé 1985a: D 86, 148).23 Undoubtedly, however, the authorities would investigate the matter to make sure that a robbery had indeed taken place, and that the parents were not liable for murder. Even infanticide was punished as murder, if the child had been killed solely because the family was already too large. But children who were considered “not whole” (buquan 不全) could be killed with impunity. And parents could kill or punish their children for showing lack of filial piety (xiao 孝). Qin law also stated that a slave (rennu 人奴) would be punished in the same way as non-slaves for killing his child, the only difference being that after he had served his term of hard labor, he would be returned to his owner (SHD 4.61, 183; Hulsewé 1985a: D 59, 140). It is remarkable that even in this case the state claimed the authority to punish the guilty slave and it is not the slave owner who possesses sole rights over his property and has the sole right to punish the murderer. In this way, the child of a slave, has the same protection under the law as a free individual. 2 Slaves as Dominated Non-Persons I would like to pass on to what I consider to be a more adequate definition of slavery for the Chinese case, because it is more culture-oriented, that of Orlando Patterson. He concludes that “property is indeed an important (though secondary) factor in defining both the legal and the socio-economic status of the slave, with this critical difference: the slave was a slave not because he was the object of property but because he could not be the subject of property” (Patterson 1982: 28). In other words, a slave was a slave because he could not own property: all he had possession of was, in fact, the property 22  The length of hard labor was probably five years. The woman’s punishment was called chong 舂 “pounding grain.” 23  Hulsewé actually translates this item, “[w]hen a child robs his father or mother, or when a father or mother unauthorizedly kill, mutilate or shave their children as well as their male or female slaves, these are not cases of official denunciation.”

Slavery In Early China: A Socio-cultural Approach

519

of his master. The master merely allowed him to make use of the property the slave possessed; the slave could never claim to be the owner of that property. The Marxist theoreticians Hindess and Hirst also argue that “slaves cannot own the means of production (or possess them)—ownership is confined to freemen/non-labourers” (Hindess and Hirth 1975: 127). Nevertheless, in late imperial China, it does appear that slaves could own property, as Meijer argues against several scholars, including Boulais (Meijer 1980: 336; Boulais 1924: section 362). Is it appropriate to call them slaves? Niida Noboru observes that in the Yuan 元 dynasty slaves could own slaves and that is possible that the term for such “double slaves” (chongtai 重臺) originates in the list of statuses presented in the Zuozhuan 左傳 (Duke Zhao 昭公 year 5) where the lowest status is given as tai 臺. It is, unfortunately, not possible to determine whether slaves owned slaves in the Spring and Autumn period, since the material relating to the lower orders in those times is extremely meager. Niida also points out that, under the Northern, Sui, and Tang dynasties, slaves were also given fields (tian 田) orchards (yuan 園) and houses by the state, but in my opinion the slaves merely possessed this property at state discretion, they did not own it, for the state continued to maintain the ultimate right of ownership.24 Even if slaves owned slaves, therefore, they did not own the means of production stricto sensu. Nevertheless, it would be remiss not to mention the fact that the Qin slave who is denounced to the government for the crime of being a scold (han 悍) and not obeying his master’s orders, was required to have his possessions sealed and guarded ( fengshou 封守). This procedure was demanded for all those under investigation for a crime, regardless whether they were free or unfree. This may imply that the Qin slave had possessions over which he had control, but since the document does not state what actually belonging to the slave was confiscated by the government, it is difficult to draw conclusions about this matter (SHD 4.14, 259–260; Hulsewé 1985a: E 15, 193–194; McLeod and Yates 1981: 5.14, 146–147). This is a crucial point to which we will return below. On the level of human relations, Patterson offers the following definition of slavery. I will use it as a basis to examine in greater detail the nature of the institution and social process of slavery that can be inferred in the light of present knowledge and from the documentation of late pre-imperial and early imperial China. His definition is “slavery is the permanent, violent domination of natally alienated and generally dishonored persons” (Patterson 1982: 13).

24  For a broad analysis of slaves in the Tang, see Li Jiping (1986), but he does not seem to consider the issue that concerns me here.

520

Yates

What is meant by “natally alienated and generally dishonored persons” in the ancient Chinese context? Let us first consider the evidence from the Discourses on Salt and Iron (Yantielun 鹽鐵論), supposedly a record of a debate held at the Han court in 81 bce between the legalists led by the financial expert Sang Hongyang 桑弘羊 (cf. Wu Hui 1981) and the Confucians on government policy, particularly that of the monopolies on salt, iron, and alcoholic beverages. The debate expanded to cover all topics, both in foreign relations and in domestic matters (Loewe 1974), and in one of the last chapters the matter of punishments was argued. Naturally, Sang Hongyang advocates the use of punishments to stop corruption, whereas his Confucian antagonists prefer moral inculcation. We put aside the details of the debate and concentrate on two statements made on either side, for they are crucial for understanding the way in which slaves and convicts were conceptualized in early China. Sang Hongyang begins his comments by declaring, “In the Spring and Autumn Annals (of Confucius), criminals lack names and appellations (wu minghao 無名號). They are referred to as bandits. The reason was to debase ( jian 賤) those who were punished (xingren 刑人) and to cut them off from human relationships (renlun 人倫)” (Lin Zhenhan 1934, pian 57 “Zhou Qin”: 189). And the Confucians note that gentlemen ( junzi 君子) do not approach criminals (xingren 刑人) because criminals are not human (xingren fei ren ye 刑人非人也). At first glance, both statements may strike one as bizarre and extraordinary, but in fact this is not so. Humans are humans because they are tied into a network of kin relationships which constitute the world of social life and which define their social being, and in China humans are humans because of the direct line and link between themselves and their ancestors or ascendants and their offspring or descendants. A slave and a criminal in early traditional China lost his rights and obligations and relations to his family, both his parents and his children. He was a socially dead person.25 All punishment involved some form of mutilation from the lightest punishment that of shaving the beard and whiskers through to the more serious mutilations of shaving the head, cutting off the nose, a foot, or castration. These were accompanied in Qin and Han times by terms of hard labor service varying from one to five or six years, in my interpretation.26 By this mutilation, even in 25  The term is Izard’s (Meillassoux 1975: 22). 26  There is considerable disagreement on the question of length of the Qin hard labor punishments: some scholars even argue that hard labor was for life. Hulsewé, (1985a: 16–17), believes that the length was the same as that in the Han period because the histories record no changes. Such an interpretation is quite possible.

Slavery In Early China: A Socio-cultural Approach

521

the cutting off of the facial whiskers, the individual was rendered unwhole and therefore polluting and unable to serve at and participate in the religious observances of ancestor worship. Thus the individual was essentially cut off from all family ties. It is for this reason that the Confucians in the Discourses on Salt and Iron say that criminals are not human. Further evidence for this interpretation can be gathered from the item in the Qin law that I have already quoted: a deformed (not whole) (buquan 不全) child can be killed by its parents with impunity because it is, in fact, in Chinese social terms, not human. And we find in the Zuozhuan a case in which a lame elder brother is barred from inheriting his father’s rank for precisely the same reason: he is not human ( fei ren ye 非人也), or considered a member of the clan (cf. Gassman) and so cannot serve at the ancestral altars (Hung 1966, vol. 1: 367; Zuozhuan Duke Zhao year 7, 535 bc).

Slavery as Social Process

What, then, of slaves? Cross-culturally speaking, Patterson has noted that there are generally eight ways in which a free person can undergo the institutional process of enslavement (Patterson 1982: 105): 1. capture in warfare 2. kidnapping 3. tribute and tax payment 4. debt 5. punishment for crimes 6. abandonment and sale of children 7. self-enslavement 8. birth Patterson observes that in China the only legal means of enslavement was the first way, capture in warfare (cf. Yu Haoliang 1985a), and that prisoners of war were first assimilated to the status of convicts before being enslaved. Roman speculation on the origins and nature of slavery also emphasizes this point. A conflict was seen to exist between the ius gentium (law of nations) and the ius naturale (natural law), for the latter posited that all men were naturally equal. Slaves were not equal because they were captured in war between nations. As. W. W. Buckland (1908: 1) trenchantly observes, “[c]aptives, it is said, may be slain: to make them slaves is to save their lives; hence they are called servi ut servati and thus both names, servus and mancipium are derived from

522

Yates

capture in war.” It was, in fact, because the Shang (second millennium bc) obviously engaged in frequent raids on their neighbors, particularly the Qiang 羌, to acquire victims for sacrifice to their ancestors and to accompany the dead Shang kings and aristocrats that many scholars have argued that the Shang was a slave state which used their victims as productive labor in addition to killing them. Perhaps, however, “slaves” in the Shang can be seen as a form of wealth to be destroyed on appropriate ritual occasions, much as the northwest coast American Indians often engaged in orgies of slave killing in their potlatches to gain prestige among their communities (Kan 1989). I do not believe that there is sufficient evidence as yet to compare the number of genuine slaves in the work-force to other categories of forced labor to determine whether there really was a slave mode of production, however that be defined in Marxist terms, in Shang times.27

27  Keightley 1969: Appendix 1 “Slavery in the Shang Dynasty,” 357–79), effectively dispenses with the arguments advanced by Guo Moruo (1973) and others that certain groups of forced laborers in the Shang can be identified with slaves, even though there are some discrepancies between the definitions of the terms employed by him and the ones that I am accepting in the present paper. Keightley analyzes the terms nu 奴, pu 僕, chen 臣, qie 妾, and zai 宰 as they appear in the oracle bone inscriptions, because these terms come to be applied to servile and semi-servile statuses in later times. He shows that Guo has misread the two inscriptions which purport to mention the word nu and that the element in the ancient graph , which is the modern xin 辛, may represent a decorative tail: thus pu probably refers to a figure of relatively high status dressed in ritual garb. Pu therefore has nothing to do with slaves. With regard to zai 宰, Guo argued that they were of the same status as chen 臣 which shows that, since the latter were slaves, so were the former. In fact, Guo has once again misread the relevant graph , which may in fact be the ancient form of kou 寇 and probably refers to high status army officers. As for qie 妾, Shima Kunio (1958: 337, 482) has shown that they were “shamaness participants in the ritual of ancestor worship.” But even if they were sacrificial victims, this does not prove that they were slaves in the strict sense of the term. Finally, with regard to chen 臣, Guo argued that the archaic graph represents an eye in the vertical position, indicating a prisoner of war with bowed head, and therefore chen were slaves. Nevertheless, the inscriptional evidence suggests that the xiaochen 小臣 were rather “members of the royal administration who performed the king’s orders by leading the peasantry in various tasks” (Keightley, 1969: 369). Keightley concludes that there is no evidence that Shang state labor employed slavery, or that people were owned like property, nor any concerning the tasks prisoners of war were forced to perform, even though perhaps some of them became human sacrificial victims. Finally, there is no evidence concerning the percentage of the population that was free as opposed to slave. The conclusion is clear: Shang was based not on slave labor but on forced labor.

Slavery In Early China: A Socio-cultural Approach

523

I would rather concentrate on the Warring States and Qin periods to illuminate the question of the relation between convicts and slaves and to examine the role this forced labor played in the economy of those times and thereby determine whether it is valid to posit the existence of a slave mode of production in ancient China.28 For the Han dynasty Wilbur (1943: 80–85) draws the distinction between convicts (tu 徒) and slaves (nubi 奴婢). Pulleyblank (1958: 201), on the other hand, ventures the conjecture that the distinction “would be, not that the slaves were subject to less severe treatment, but that their bondage was perpetual, extending to their offspring, whereas that of convicts (not presumably those guilty of crimes for which the law of shounu 收奴 (the arrest and enslavement of offenders’ wives and children) was imposed—the principals in those cases would have been executed) was, if we may judge from Han practice, limited to a period of years.” The important point to remember is that enslavement was always thought to originate in being a substitute for death, archetypically death in war, or the death commuted was a capital offense, or from misery, starvation, and exposure. Slavery was a conditional commutation, not a pardon from this death: as Katrina McLeod and I have pointed out elsewhere (McLeod and Yates 1981: 136, note 67; cf. McKnight 1981), she 赦 in ancient Chinese did not mean an amnesty in the sense of a complete forgetting (amnesia) of the crime, only a conditional remission. It is this notion that lies behind the treatment of slaves as outsiders, subject to the total domination, exploitation, and disposition of their masters. This belief is apparent in the richly symbolic ritual of surrender in war in the Spring and Autumn period as recounted in the Zuozhuan of which there seems to be two types, but which both indicate that the defeated considered himself dead (Yang Ximei 1956). In one case, the defeated Baron of Xu 許男 presents himself to his captor, the Viscount of Chu 楚子, bound, and holding a jade bi 璧 disc in his mouth, while his lords (daifu 大夫) are dressed in mourning clothes and his knights (shi 士) bear his coffin (Hung 1966, vol. 1: 97; Duke Xi 喜公, year 6). The jade disc was usually placed in the mouth of a corpse prior to interment. In another case, the Earl of Zheng 鄭伯, vanquished by Chu, with shoulder bare as in mourning and leading a goat, says to his captor that if the latter wishes to enslave Zheng’s people (chenqie zhi 臣妾之), it will be his command, and he admits that he had committed a crime (zui 罪) against Chu by not serving it (Hung 1966: vol. 1, 195; Duke Xuan 宣公 year 12). Liao Meiyun (1995: 19–24) has pointed out that many of the women captured in warfare in Spring and 28  See Li Hengmei (1999) for a recent discussion of slaves in Spring and Autumn times, including their origin, status, and occupations.

524

Yates

Autumn and Warring States times became government musicians and slaves concerned with the preparation and serving of wine and food. But, as will be seen below, those whom the state of Qin captured in war in the fourth and third centuries bc, if they were not massacred, became convicts, not slaves.

Slaves in the Qin Laws

It is now necessary to examine in greater detail the evidence of convict and slave status in the Qin laws, though the data cannot be treated exhaustively, especially that relating to convicts, since there is such an abundance of it. First of all, there is no mention of self-enslavement, kidnapping, abandonment and sale of children, and tribute and tax payment of slaves in these documents. A case of kidnapping, however, is recorded in the contemporaneous Zhuangzi 莊子, where the robbers cut off the foot of their victim, because, so mutilated, he will be easier to sell (Guo Qingfan 1978: j. 8B, 860). In addition, it is highly likely that, in the chaotic social conditions of the third century bc during the wars that led up to the Qin conquest and unification of China, as well as during the collapse of the Qin dynasty ten years later, many robber bands roamed the countryside, preying on the unfortunate populace and selling them into slavery for their own profit. This was a feature of the Han and later dynasties: whenever the central government broke down and the laws against this highly profitable activity were unable or unwilling to be enforced (Wilbur 1943: 90–92). During the wars accompanying the overthrow of the Qin, economic conditions in fact reached such a pass, with severe famine hitting many areas, that Emperor Gaozu 高祖 of the Han issued an order permitting the sale of children and allowing the parents to migrate to safety in Sichuan (Wilbur 1943: 266; Wang Xianqian n.d.: j. 24, 9b; cf. Appendix). It is probable that the Zhuo 卓 family, who were exiled there by the Qin when they captured the state of Zhao, the Zhuo’s homeland, in 228 bc, gained many of the slaves, on whose labor they built their famous iron mining and casting fortune, as a result of this order (Wilbur 1943: 259; Wang Xianqian n.d.: j. 91, 8b). With regard to enslavement for debt, which was a common way in the later period in which poor families fell into bondage in times of great economic hardship,29 the Qin laws had the stipulation that a creditor could not forcibly seize the guarantee (zhi 質) of the debt in lieu of payment. Such creditors and those debtors who of their own accord handed over the guarantee were fined 29  Watson (1980c: 223) notes that in the later traditional period nearly every peasant household was affected by slavery and most of the slaves were sold because of extreme poverty.

Slavery In Early China: A Socio-cultural Approach

525

two sets of armor, a very heavy cash fine, but the precise cash equivalent of which is unknown. Here the guarantee was probably, in fact, a person (SHD 4.129, 214–215; Hulsewé 1985a: D 126:162.30 Such pawns in the Han, if not redeemed within three years, were permanently enslaved, at least in the Huainan 淮南 region (cf. Ru Shun’s 如淳 comment on Wang Xianqian n.d.: j. 64A “Yan Zhu zhuan 嚴助傳,” 4a). The other three methods of enslavement listed by Patterson, capture in warfare, punishment for crimes, and birth, are also to be found in early China. In the Han, the normal general term for convict, as has been mentioned, was tu 徒 and that of male slave nu 奴 and female slave bi 婢 although there are several other terms of lesser importance that need not be of concern here (Wilbur 1943: 64–71; Li Xinda 1978). In the Qin legal documents, however, the terms are different. Tu occurs seven times, as far as I can determine, with only two of them possibly having the meaning of “convict,”31 and there are two ways of indicating slave status: rennu 人奴, renchen 人臣, or chen 臣 for a man, and renqie 人妾 or qie 妾 for a woman. The former means “a person’s male slave or servant” and the latter “a person’s female slave or servant.” The Han term for female slave (bi 婢) does not appear. In addition, the hard labor punishments of “watchman” (hou 候), “robber guard” (sikou 司寇), “collector of firewood for sacrifices to the spirits” (guixin 鬼薪) for males and “sorters of rice” (baican 白粲) for females, and “builders of walls from early dawn” (chengdan 城旦) for males and “pounders of rice” (chong 舂) for females also appear. All these labor terms are known from the Han except the first and lightest “watchman.” Possibly this punishment died with the Qin. There are two more terms li 隸 “bondsman,” probably a general term for “servant” or “slave,” and lichenqie 隸臣妾 “male and female bondservant.” The 30  Hulsewé does not venture an opinion as to whether the pledge is to be understood as a person or as an object. 31   S HD 1 Gong 4: 70–71; Hulsewé (1985a), A 55, 58–59, and SHD 1 Yao 1: 76–80. Hulsewé (1985a: A 46, 63) translates the latter rule, “When levying conscripts for work inside a settlement, order is given to guarantee the earth walls for a full year.” In three other passages, the term refers to the workers at the lowest level in craft workshops and stables, hence it could mean convicts or free laborers. See SHD 1 Jiuyuan 3: 33 (Hulsewé 1985a, A 9, 27–30, especially note 21, 29–30); SHD 3.11 and 3.12, 136–38 (Hulsewé 1985a C 11 and C 12, 110). In another case, it refers to soldiers on guard in the palace grounds who are responsible for killing marauding tigers (SHD 3.16, 140–141; Hulsewé 1985a: C 16, 112–13), and in another, also some type of infantry guards (SHD 3.21, 144–145; Hulsewé 1985a: C 21, 116). Finally, in SHD 4.160, 229, it refers to some kind of guard for state ambassadors. Hulsewé (1985a: D 159, 172–173) finds this item extremely hard to understand.

526

Yates

latter appear with great frequency and the Qin state seems to have relied very heavily on them for every type of productive and non-productive labor, from work in the fields, in foundries and workshops, to tending pasture animals, to fighting in the army and acting as prison warders. The major question that has divided scholars of these new documents is this: are these bondservants slaves or are they convicts? Were they a permanent hereditary status, or were they convicts serving a fixed term, or a life term (cf. Tang Zangong)? If we can resolve these questions, we can determine to a far greater extent than hitherto possible whether there was a slave mode of production in the Qin and Warring States, or whether slavery and/ or convict labor played an important role in the economy of the Qin and Warring States social formation. There is no doubt that the evidence is conflicting, so before I discuss the question, let me review the data on the status that can be definitely understood as slave, rennuchen 人奴臣 and renchenqie 人臣妾. These individuals, Gong Changwei and Song Min (1982) have argued, are private slaves in contrast to lichenqie bondservants, whom they interpret as state slaves. People in this type of status were considered to be property of their owners and not human in the social sense, yet at the same time, they were definitely held responsible for their crimes and punished by the government, and were not considered to be suffering from diminished capacity in the legal sense. Crimes mentioned include beating their masters (zhu 主) (SHD 4.63; Hulsewé (1985a: 169), raping them (presumably this applied primarily to mistresses) (SHD 4.63; Hulsewé 1985a: 169),32 secretly planning to murder their masters (SHD 4.64, 184; Hulsewé 1985a: D 60, 140), planning to steal their master’s ox, selling it, taking the proceeds, and fleeing the state (SHD 4.4, 152–153; Hulsewé 1985a: D 4,122), unauthorizedly killing their child (SHD 4.61, 183; Hulsewé 1985a: D 59,140), beating their child so that it dies of its wounds (SHD 4.62, 163; Hulsewé 1985a: D 62, 140–142),33 robbery (SHD 4.18, 159–160; Hulsewé 1985a:

32  The punishments for beating and raping the master/mistress were the same, but not stated in the documents. The punishments meted out to the slave in Qin times for the crimes of beating, raping, and planning the murder of the master/mistress were probably comparable to those in the Qing dynasty. Immediate decapitation was the punishment for beating the master and fornication with the master’s main wife or his daughter, with strangulation after the assizes for such fornication with a master’s concubine. (Meijer 1980: 335). 33  The slave is tattooed on the forehead and cheekbones and then returned to the master; no hard labor is stipulated.

Slavery In Early China: A Socio-cultural Approach

527

D 18, 125–126; SHD 4.122: 211; Hulsewé 1985a: D 119, 159–160),34 and finally being a scold (han 悍) and refusing to work in the master’s fields.35 It is important to note that the state claimed the right to punish such delinquent slaves, even though their masters could mete out the punishment themselves. This was one more way in which the Qin state tried to dominate all members of society and gain exclusive control over their lives. In addition to being punished, slaves could be sold to the government at the fair market price. In one case, the slave has his foot cut off, and is made a chengdan forced laborer (SHD 5.14, 259–260; Hulsewé 1985a: E 15, 193–194; McLeod and Yates 1981:5.14, 146–147). However, he first has to state that he has never been manumitted, is not sick, and is not liable for any other crime. The lingshi 令使 a low official in charge of records (Yu Haoliang 1985b: 109) makes a medical examination of the slave and certifies his state of health. The shaonei 少內 treasurer and his assistant determine the price and make the transaction.36 It was important to prove that the slave had not been freed and then forcibly re-enslaved, because, as we know from a Han source, if such was the case, the owner had committed a serious offence.37 In the Qin, the authorities ask the slave himself about his status, not trusting the word of the owner. The price of the slave would also, of course, depend upon his health, hence the medical examination. The appearance of the term “fair market price” confirms the fact that the Qin had slave markets, as Wang Mang claimed, a state of affairs also recorded in the Zhouli 周禮 where people renmin 人民 are noted as saleable items (Sun Yirang 1983: vol. 2 j. 27, 11b; Biot 1969: vol. 1, 317; Pulleyblank 1958 192–193). In the Qin, all items on sale above the value of one cash were required to have a price tag (SHD 1 Jinbu 5, 57; Hulsewé 1985a: A 46, 53), and therefore undoubtedly slaves 34  In the latter case, it is stated that it is the state that rewards the individual who arrests a male or female slave for stealing 110 cash, rather than the owner who rewards him. 35   S HD 5.14, 259–260; Hulsewé 1985a: E 15, 193–194; McLeod and Yates 1981: 5.14, 146–147, where a male slave is a scold and refuses to work in the fields. SHD 5.15: 260–61; Hulsewé 1985a: E 16, 194–195; McLeod and Yates 1981:5.15, 147–148, where a female slave confesses that she is a scold, but her occupation is not given. Being a “scold” was one of the seven legitimate causes for divorce in traditional law (Johnson 1997: 167, Article 189, subcommentary). 36  Hulsewé (1978: 199) and (1985a: A 40, 50, note 9) has demonstrated that the shaonei treasuries were part of the metropolitan and regional governmental organization, being attached to all organs that were involved in financial transactions. 37  Su Yiwu 蘇夷吾, marquis of Pu 蒲侯, had his marquisate abolished for this reason in 18 bc (Hongjia 鴻嘉 year 3) (Wang Xianqian n.d.: j. 17: 25b; Wilbur, 1943: no. 102, 419).

528

Yates

wore such tags when they went on sale. A humorous story in the Hanfeizi 韓非 子 records that a female slave in Wei 衛 cost more than 100 rolls of cloth (Wang Xianshen 1956: ce 3, pian 31, 5; W. K. Liao 1959: vol. 2, 8). In the Qin, a roll of cloth, which was an important form of currency, was stated to be the equivalent of 11 cash (SHD 1 Jinbu 3, 56; Hulsewé 1985a: A 44, 52–53). Thus a female slave could have cost, at this exchange rate, over 1100 cash, depending on her age, state of health, and general disposition. The value of slaves in the Han dynasty was much higher. In the record of a law case dating from 196 bc translated in the Appendix to this essay, a woman was bought, probably sometime during the Qin period, for 16,000 cash. This price is confirmed in the fragmentary evidence from the Juyan 居延 wooden slips, where we find that it cost 20,000 cash for one adult female, and a nonadult male was worth 15,000 cash. Other evidence suggests a cost of 20,000 to 30,000 per adult slave (Xu Yangjie 1981: 169). As can be seen from the above, a person could not only be a slave, but she/ he could also be a convict for having committed a crime. In 5.14 “Denouncing a Slave” mentioned above, the slave is sold to the government and made a chengdan forced laborer. In 5.15 “Denouncing a Female Slave,” the female slave has her nose cut off by the authorities and is tattooed. Another case in the slips refers to slaves being mutilated, being made chengdan convicts and then being returned to their masters (SHD 4.4, 152–153; Hulsewé 1985a: D 4, 122). This surely indicates that the forced labor punishment was for a fixed period of time, and therefore convicts cannot have served life sentences, and thus were not structurally equivalent to slaves, as some scholars have maintained. It should also be noted that the Qin permitted their convicts to return to their homes to perform essential agricultural duties at appropriate times of the year. No mention of slaves is made in this connection, and therefore it is unclear whether their masters could petition to have their slaves returned to them on these occasions. That this was possible, I consider to be quite unlikely. Slaves could also be used by their masters in the same way as oxen and horses to be held by the government to pay off their fines and debts. To be the owner of slaves was thus additionally beneficial. Such slaves had to wear red clothes, be shackled, and be supervised, if they were held to work with the chengdan and chong forced laborers (SHD 1 Sikong 12, 84–87; Hulsewé 1985a: A 68, 67–71), just like other chengdan convicts who were originally of free status (SHD 1 Sikong 12, 89–90; Hulsewé 1985a: A 70, 72–73), and like guixin collectors of firewood and baican sorters of grain convicts and a status called qunxiali 群下吏, who had not suffered the punishment of shaving the whiskers (SHD 1 Sikong 12, 84–87; Hulsewé 1985a: A 68, 67–71). On the other hand, individuals of the lowest aristocratic rank gongshi 公士 and commoners (shiwu 士伍), who

Slavery In Early China: A Socio-cultural Approach

529

were being held to redeem mutilating punishments and the death penalty and who were held for chengdan and chong labor, did not have to suffer the indignity of wearing red clothes and being shackled.38 This difference between slaves and regular “free” convicts redeeming their crimes or fines is of utmost importance, in my opinion, to see the difference between slave and convict status. The Qin obviously drew a careful distinction between redeemable crimes and non-redeemable crimes and between regular convicts and slaves. Slaves were treated in the same manner as individuals serving non-redeemable heavy labor sentences, but at the same time the term “non-redeemable” must be understood not as meaning that the convict never ended his term, that his labor was for life, but rather that the redeemable crimes could be paid off in cash all at once if the criminal had the financial resources. If he lacked the means, he was held until he paid off the crime. The rate was fixed at 8 cash per day, 6 cash if the convict received food from the government (SHD 1 Sikong 12, 84–87; Hulsewé 1985a: A 68, 67–71). In other words, in a 365 day year, the convict would pay off 2920 cash or 2190 cash, which would work out as 97.3 piculs (shi 石) or 73 piculs of grain, since the price of grain was fixed at 30 cash per picul. Unfortunately, it is nowhere stated what the cash equivalent of the various hard labor service punishments were, but if, as I believe, the labor terms ran from one to five years for the watchmen through to the chengdan the minimum might be 2920 or 2190 to a maximum of 14,600 to 10,950, a very large sum for the average commoner to pay all at once. We might note, however, that the government of the Qin state must have benefited immensely from this practice of cash payment, for if it did not acquire the forced labor of the criminals, then it received lump sums of cash, which it could then put to good use to increase its military and public works budgets and enable it to field large armies for years at a time and pay for the construction of huge public works, such as the Great Wall, the First Emperor’s tomb and the many palaces in the Wei river 渭河 valley. A further point with regard to slave status is that the wives of slaves were apparently not graced with the term that meant legitimate wife, qi 妻, they are always referred to as qie 妾 “concubine.” This is of great significance, for it is a practice common to other slave systems throughout the world (Patterson 1982: 186–190). Slaves are slaves because they are cut off from rights and obligations of family and their unions are not recognized as legitimate and binding on the rest of society. Their “families” could be separated by their owners entirely at their owners’ discretion, although, in general, slave owners kept such 38  Hulsewé (1985a: 69, note 10) observes that the phrase “gongshi or less 公士以下” probably refers to gongshi and commoners (shiwu), but this would be rather remarkable.

530

Yates

families together to encourage better performance of duties.39 In the Qin, evidently slave unions were recognized, because they were recognized as parents who could kill their children. Yet in the customs of the time only legitimate wives could perform duties in the sacrificial rites to their husbands’ ancestors, concubines could not. It is for this reason that I say that slave unions were recognized, but were not considered legitimate: concubinage was, however, condoned for the free population, and so slave unions were assimilated to this secondary form of marriage. As mentioned above, another important item quoted in section 4 of the legal documents states that if a husband commits a crime and his wife denounces him first, before the authorities learn of the crime, she is not subject to the law of arrest (shou 收) and her male and female slaves and maidservants (ying 媵), who accompanied her to her husband’s home, are also not confiscated. And if the wife commits a crime, her maidservants and slaves, clothes, and utensils (presumably her dowry) are not forfeited either, but given to her husband. This indicates that wives, although in some respects property of their husbands, were able to hold property independent of them and that confiscation of property on the commission of a crime was not automatic. A husband or wife could retain the property of their spouse if they denounced them first. No doubt, however, a number of slaves in the Qin government’s hands were originally private slaves confiscated when people failed to denounce their spouses. It is also clear that the government gave gifts of slaves (chenqie 臣妾) to private individuals for meritorious deeds. In the documents, it is recorded that someone who catches a person sending a written communication to the enemy is given two such slaves, male and/or female (SHD 4.44, 174–175; Hulsewé 1985a: D 43, 134–135).40 In this case, government slaves became private slaves. In addition, under Lord Shang’s laws, a person could receive grades of rank and an appropriate number of slaves (li) as a reward for cutting off heads in battle (Zhu Shiche 1974: 73; Duyvendak 1963: 298). Another passage states that prisoners of war were given to successful officers, which leads us back to one of the modes of enslavement mentioned earlier (Wang Xianqian n.d.: ch. 23, 7b; Hulsewé 1955: 327; Takigawa 1979: j. 68: 9; Duyvendak 1963:15; Zhu Shiche 1974: j. 5 pian 39  Harrison (1968: 177) remarks that in the case of Athens “although relatively permanent unions were in fact formed between slaves it is very doubtful whether these had any juristic validity” (cf. Fogel and Engerman 1974: 49 and 142–144). 40  I adopt Hulsewé’s explanation of the term toushu 投書 because it makes better sense in the context in contrast to that of the SHD editors who interpret the term as “to lodge an anonymous complaint.”

Slavery In Early China: A Socio-cultural Approach

531

19,79; Duyvendak 1963: 299–300). But before we turn to this issue and that of the so-called bondservants (lichenqie 隸臣妾), I want to make a few more remarks about slaves. Another of the major sources of Qin slaves must have been birth. Many Qin chenqie slaves may have been the descendants of the chenqie recorded in Western Zhou bronze inscriptions as being transferred between the Zhou ruling house and other members of the Zhou aristocracy.41 It is impossible, however, to determine the extent to which Zhou slaves reproduced themselves and therefore to make any conclusions at all about breeding practices and the extent to which the slave population of those times reproduced itself. Therefore fundamental questions about the economic viability of the slave system cannot be answered. This is regrettable given the advances that have been achieved in this aspect in studies of the slave system in the Western hemisphere. The data regarding this issue is also lacking for the rest of the imperial period. Quite clearly it must have been relatively easy for free members of Warring States society to fall into servile status by the methods we have already discussed because in the almanac texts (rishu) there are a number of passages which indicate what will happen to children born on a particular day: some days are lucky some days unlucky. The almanac texts express the fears of Qin parents. For example, slips 1138–1139 state that to bear children on the renyin 壬寅 day was unlucky or inauspicious: a girl born on this day would become a doctor (Yunmeng Shuihudi Qinmu Bianxiezu 1981: plate 164), and children born on the jisi 己巳 and dingwei 丁未 days were thought to grow up to be slaves. I am not sure exactly why these days were considered so inauspicious. 41  Keightley (1969:210–216) argues that these chenqie were not slaves in his definition of the term, even though they were originally captured by the Western Zhou aristocratic rulers. The reasons are that chen were never transferred as individuals but in groups or families ( jia 家) and that geographical names were frequently attached to them, indicating “that we are dealing with groups of people whose previous social organization or regional location took precedence over their status as ch’en. In other words, the ch’en’s primary ties were to their family units and to the land” (213). The term chenqie certainly appears in the Zuozhuan and I would argue that there is a strong possibility that, with the development of the beginnings of a commodity market system in the late Spring and Autumn and early Warring States periods, these chenqie began to be sold as individuals. Huang Ranwei (1978) doubts the interpretation of the terms as indicators of geographical origin, rather believing them to be types of servants (chenpu 臣僕) (cf. Chen Hanping 1988: 259–60; Musha 1980). Liu Weimin (1975: 118–22) considers the evidence from bronze inscriptions as well as literary texts and concludes that they together with the various other types of low-grade workers, were some kind of forced laborer. For another analysis based on a comparative perspective, see Hu Liuyuan and Feng Zhuohui (1988: 317–415).

532

Yates

Perhaps it had something to do with beliefs in word magic. But it is not impossible that children born on these days would have been killed if the parents thought that they could avoid prosecution by the authorities for murder (Yunmeng Shuihudi Qinmu Bianxiezu 1981: slips 1139 and 1142; plate 128, slip 874).42 Again, some days were considered inauspicious for bringing slaves and lodgers ( jiren 寄人) into a house, for it was believed that within a short period they would replace their masters.43 It would seem that these disadvantaged individuals could gain rights and obligations quite easily. Slaves could also gain their freedom by having their relatives redeem them with cash or equivalent goods. Indeed, according to Liu Xiang’s 劉向 Shuoyuan 說苑 in the mid-Han, the state of Lu 魯 in Confucius’ day provided funds from the state treasury to redeem Lu slaves held in other states, but whether this was a common practice throughout China is unclear (Liu Xiang n.d.: j. 7: 23ab). Among the Qin documents is also preserved one item of the household statutes of the state of Wei 魏, probably promulgated in the year 252 bc, in which the king proclaims that it is not the state’s fault that sometimes, presumably in those of great economic hardship, the people abandon their homes in villages to go and dwell in the countryside and enter the houses of orphans and widows, demanding their wives and daughters. He therefore proclaims that gate-guards ( jiamen 叚門, i.e., jianmen 監門), migrants or sojourners (nilu 逆呂 = 旅) and pawns (zhuixu 贅婿) and houfu 後父, probably another type of pawn or step-father, may not be allowed to establish households and receive fields and houses from the state.44 And this prohibition would last for 42  Perhaps the reason for these prognostications lies in the beliefs associated with word magic. The graph ding 丁 meant an individual” and wei 未 “not yet”: the complete term dingwei 丁未 might have been thought to indicate that the person would not exist. Actually, trees (mu 木) were not supposed to be watered (澍)—perhaps the graph is an alternative form for shu 樹 “planted”—on wei 未 days because the graph is a representation of a tree with a cut across the trunk. Perhaps they thought that such a tree would come to no good end (Yunmeng Shuihudi Qinmu bianxiezu 1981: plate 126, slip 853). Ji 己, on the other hand, meant “self” and si 巳 sounded like si 死 “die,” indicating that the self would die if born on that day. 43  Yunmeng Shuihudi Qinmu Bianxiezu 1981: plate 120, slip 786, reads, “Do not bring lodgers in on xinyou 辛酉 days; if you bring in lodgers, they will inevitably take over living in the residence.” 44  Hulsewé translates nilu as “innkeepers.” The precise meaning of the terms jiamen and nilu however, has been the subject of some debate. The SHD editors (293, note 4) interpret the former as jiamen 假門 to be read as jiamen 賈門, with the meaning of “traders” or “merchants” (shangjia/gu 商賈). Wu Rongzeng (1995a, 1995b), however, disagrees and makes

Slavery In Early China: A Socio-cultural Approach

533

three generations. Only after this time lapse could they be allowed to gain employment, presumably in the government, although they still had to register as grandsons of such semi-servile people (SHD 292–294; Hulsewé 1985a: F 1, 208–209). Here, very clearly, we have another way in which semi-servile individuals were disadvantaged in relation to the free, common people. Although registration in households entailed payment of taxes, including corvée labor service, yet at the same time it did permit the poor some means of subsistence, because land and houses were also given to such householders by the state. But these migrants, pawns, and gate-guards were prevented from engaging in agricultural work, and this failure to work in the fields was considered a crime in Warring States China. Lord Shang’s laws stipulated enslavement for the idle (Takigawa 1979: j. 68: 8–9; Duyvendak 1963: 15), and the Guanzi 管子 suggested that they be sent to the frontier to perform guard duty, a form of punishment (Guanzi 1934: ce 3 pian 78 “Kuidu,” 90). In another Wei state statute preserved with the household statute in the Qin tomb, the authorities proclaim that such indigents should be pressed into the army where the generals should not show them any mercy (SHD, 294–295; Hulsewé 1985a: F 2, 209–210). The king even implies that they should be killed outright, and, if not outright, they should be used to fill in the moat as human waste when the need arose during the siege of a city, when an assault on the walls was considered necessary.45

a convincing argument in favor of reading jia as jian 監, with the full term jianmen 監門 meaning a low-status “door- or gate-keeper.” Yang Heding (1983) rejects his conclusions and attempts to prove that the first interpretation is correct. Unfortunately, his efforts to show the importance of gates (men 門) for markets have little to do with the meaning of this status or functional/occupational term, and I remain unconvinced at present. Indeed, it is not impossible that jia should be taken as it stands, and the term jiamen should be understood as low-status bondservants or slaves who had been loaned ( jia 假) out by the government to private individuals: Li Jiahao points out that in Lu dialect the two words jia and jian were interchangeable (Wu Rongzeng 1995b: 171). We find in the Statutes on Granaries, for example 1 Cang 17 (SHD, 48–49) slave girls not yet of serviceable age (weishi 未使) loaned out to the common people (baixing 百姓), who are then responsible for clothing and feeding them. Males in Wei could perhaps have also been loaned out. Nevertheless, in this interpretation, there remains the problem of the meaning of men. 45  I follow the rendition of the SHD editors: Hulsewé’s translation does not seem appropriate here.

534

Yates

Bondservants Let us now turn very briefly to the question of the bondservants (lichenqie), which, as I have already mentioned, some scholars interpret as convicts and others as government slaves. I believe that Liu Hainian (1985, 1986) is correct that some bondservants were convicts whose sentences were of limited duration and others were permanent government slaves. This situation gives further evidence that the number of grades in the orders below the “free” commoners (shiwu 士伍) was considerable. First of all, quite clearly some of them were convicts and probably served three-year hard labor terms. Others, however, inherited their status from their father in the following way. Unlike slaves, a male bondservant could have a wife not of bondservant status, and the union of the couple was considered legitimate, for the wife was called qi 妻. A child born of this union while the father was still a bondservant, however, inherited his father’s status. If on the father’s death, the wife tried to conceal the circumstances of the child’s birth, she herself became a bondservant (SHD 4.155, 225–226; Hulsewé 1985a: D 154, 169–170). It is interesting to note that, generally speaking, marriage to a male slave resulted in the free woman accepting his status, and this was true in ancient and modern times both in the East and the West (Patterson 1982: 131). Indeed, in later Chinese custom, the principle of deterior condicio that the rule in the Qin exemplifies, continued in full force, such that the child inherited “the status of the parent with the lower status.” In the Yuan, the rule was changed and the wife took on the status of the husband (Patterson 1982: 141–144). Qin enemy who surrendered were also made bondservants (SHD 3 Za 24, 146–47; Hulsewé 1985a: C23b, 117–118). Were these individual prisoners of war permanently enslaved or did they merely serve a three-year term? One provision of the statutes of army rank ( junjuelü 軍爵律) indicated that bondservants who cut off enemy heads in battle and received the rank of gongshi, the lowest of the aristocratic ranks ( jue 爵), could return that rank to the state and, in exchange, be freed (mian 免) and made commoners. Two ranks could be exchanged for either a father or mother who was a bondservant. And artisan bondservants who gained similar rank could be freed to become artisans. If they were not whole, i.e., they were mutilated in some way, they became socalled “hidden office artisans” (yinguangong 隱官工) (SHD 1 Junjue 2, 93–94; Hulsewé 1985a: A 91, 83). What a strange situation that a bondservant could gain aristocratic rank! Perhaps we should interpret the situation as follows: the bondservant is substituting or exchanging his own social death by or for the death of the enemy. The medium of exchange is rank, which could, as we know, mitigate the severity of all punishments.

Slavery In Early China: A Socio-cultural Approach

535

That the Qin cannot have been alone in recruiting low status individuals into the army can be seen from the military text Liu tao 六韜, where it is noted that debt-slaves (zhuixu) and captives who wish to conceal their tracks and proclaim their names should be gathered into a separate regiment (zu 卒), as should slaves who are seeking to expurgate their crime and wish to expurgate their shame (Liu Yin 1975: j. 3 pian 53 “Lianshi,” 16; cf. Wu Rongzeng 1995a). Unfortunately, the military systems of the other states in competition with Qin are less well known. We do at least know that some of them rewarded cutting off heads in battle with gifts of gold, fields, and houses, but which of them used low-status soldiers, and whether they rewarded meritorious soldiers with rank, is unclear at present. Wilbur (1943: 98–117) devotes a chapter of his book to the question of whether prisoners of war were enslaved in the Han or not. He argues that, in the Han, most of the war captives were non-Chinese, especially Xiongnu 凶奴 and those from the minority regions in southwest and south China. Given the difficulty of absorbing large numbers of non-Chinese-speaking nomadic warriors, he concludes that, although some Xiongnu prisoners of war were enslaved, most were not, and that those from the southern regions were more tractable and easier to absorb by the Chinese as slaves. In the Six Dynasties 六朝 and Sui 隋 periods, captives were also sometimes enslaved and forced to join the conquering army, and, on occasion, the whole population of a defeated city was enslaved en masse. These slaves were either appropriated by the government, given by it to private individuals, often the victorious generals, or forcibly resettled in a new area far from their original homeland (Wang Yi-t’ung). There were, in fact, other ways in which bondservants could be freed. Five years of frontier duty would be sufficient for a female bondservant (SHD 1 Sikong 14, 91–92; Hulsewé 1985a: A 72, 73,46 as could replacement for old and young adult male bondservants by male adult individuals, who would become bondservants themselves. Two adults could redeem one male adult bondservant. Female bondservants who were tailors or seamstresses could, however, never be redeemed (SHD 1 Cang 25, 54–55; Hulsewé 1985a: A 35, 45–46). This certainly indicates how important the textile industry was to the Qin state, for not only would these seamstresses provide clothes for the state’s royal court, 46  The editors place slip 153 at the end of this rule. In Hulsewé’s translation, “If someone for the redemption of banishment wishes to pay money, this is eight cash per day.” It has been suggested to me that if lichen were slaves and the status was considered the equivalent of 5 years at the frontier, to manumit such a slave would be the equivalent of 14,600 cash. This is possible, but it may not be correct to place this slip in this location, and so I reserve judgment on this matter.

536

Yates

convicts, and slaves (soldiers in the army were expected to provide their own uniforms, but probably not their armor), but also because they wove cloth that was used as an important medium of exchange, a kind of money. Given the relative backward state of the Qin economy in comparison with the more advanced states of the central plains, which had been using metal currency for several hundred years, this cloth was crucial for the economic survival and success of the Qin. In that these seamstresses could never be redeemed, I believe that it is only reasonable to conclude that, although they were technically convicts, in fact they were permanent slaves, forever alienated from their natal ties, and whose labor was ruthlessly exploited by the state in its quest for strength and universal dominance. It is in this context that I wish to emphasize that we should not take the ideological attack by Lord Shang and the legalists, and even the Confucians, on trade and merchants too seriously. The Qin recognized the importance of markets for the overall health of its economy and to provide the necessaries to supply its huge armies (Zhu Shiche 1974: j. 1 pian 2, 8; Duyvendak 1963: 181–182). The Qin legal documents provide ample evidence that the state was directly engaged in all sectors of the economy, not only in agriculture, the fundamental occupation, but in the advanced sector too, in iron foundries, in lacquer, pottery, cloth, and leather production, and in the construction of transport vehicles, and possessed strict rules about the functioning of markets (Hulsewé 1985b). As in other states, the Qin seems to have monopolized many, if not most, of these markets, and the government workshops were probably located close by to reduce transportation costs and permit ease of sale (Qiu Xigui 1980). As in the Han, the Qin registered traders on rolls separate from the rest of the population to ensure the greatest control possible of these essential, if ideologically suspect, members of the community (SHD 1 Jinbu 4, 57; Hulsewé 1985a: A 45, 53). Indeed, in a group of newly published texts, discovered in 1973, and probably related to the Warring States text Weiliaozi 尉繚子, it is advocated that in the center of all settlements (yi 邑) there should be a market (Linyi Yinqueshan Zhujian Zhengli Xiaozu 1985).47 This is very different from earlier times, when the centers of the villages or towns were the religious buildings, ancestral temples if the urban agglomeration were large enough to be the center of a lineage. I therefore believe that, although the large number of laws concerning agriculture and grain storage, with their remarkable detail, show that taxes in the form of grain and hay, together with the poll tax, provided the main source of government income, the Qin’s exploitation of slave and convict labor in the 47  Qiu Xigui accomplished the difficult task of deciphering slips 875–890.

Slavery In Early China: A Socio-cultural Approach

537

advanced sector of the economy, not to mention its use of great amounts of corvée labor, may have been one of the decisive factors in their success in defeating the other states and unifying China. Does this mean that in Qin times the basic mode of production was the slave mode? In the current state of research, I would have to say a tentative “no” in response. The reasons for my opinion are that it is likely that, as I have argued elsewhere (Yates 1987), the vast bulk of the population were members of the “free” commoner class, and that it was they who provided most of the agricultural labor and most of the personnel for the Qin armies and aspired to positions in the Qin bureaucracy.48 Nevertheless, there was a slave, or at least forced labor, mode of production. Unlike Rome, where slaves were a far greater component of the total population and where slave labor was employed on a vast scale, particularly in the agricultural estates, the latifundia and also in mining, in certain occupations in the urban economy and in education, and which therefore has a strong claim to be labeled a slave mode of production (Anderson 1974: 53–107, especially 62ff), in China this mass of “free” commoners was a predominant force and certainly owned some of the means of production, such as oxen and tools, even if at present the extent to which they were able to own the land itself is open to question (Hulsewé 1985b: 215–18). As we have seen, these commoners could own slaves themselves, but there was clearly a wide range in financial circumstances between the richest and the poorest of them, and their wealth certainly was not at all equal to that of the aristocrats. Conclusion What other conclusions can be drawn from the data presented above? First of all, it must be recognized that there was a range of low statuses below that of the common people. In non-technical legal language, the common people were the pingmin 平民 “level people” and were known also by the term shuren 庶人 or shumin 庶民. Manumission or freedom from convict status always used this latter term. However, the legal status for a male commoner was shiwu 48  The level to which commoners could aspire in the Qin bureaucracy is unknown. Cho-yun Hsu (1965) gives many examples of low status individuals rising to the highest ranks in other states, but I am not sure if this was always true in Qin. Commoners there may have only risen to the equivalent of the eighth rank in the rank hierarchy. Under Lord Shang’s laws, an individual who was awarded rank for cutting off heads in battle could, if he so chose, return the rank and receive an official position instead, a policy which Hanfeizi deplored (Wang Xianshen 1956: ce 4 pian 43 “Shuoyi,” 22; Liao 1959: vol.2, 215).

538

Yates

a “member of the rank and file.” This status is to be compared and contrasted to lower orders, and to the so-called aristocratic ranks ( jue), which commoners could acquire by cutting off heads in battle, inherit from their fathers, or by performing other meritorious deeds for the government. These we have mentioned above, but there is also the possibility that they could purchase them and win them by producing exceptional quantities of agricultural surplus. Probably commoners, as in the Han dynasty, could not expect to be promoted above the eighth rank in the aristocratic hierarchy, a hierarchy that stretched up to the king and through him to the cosmos as a whole. But the evidence for the Qin on the break between commoners and aristocrats is extremely limited. Below the commoners, and cut off from the ranking system were the gateguards, pawns, migrants, and other low statuses,49 who were denied rights to land, houses, and registration in households. Below them were the convicts, and then slaves. Within the convict status, a distinction was drawn between those who had to wear special red clothing and shackles, on the lowest level, and the lesser hard labor service convicts, those who served up to three years. Although all convicts served for only fixed time periods, yet at the same time, after completion of their sentences, they still did not regain full commoner status or rights. In the first place, if they had been mutilated, they were still considered polluting and could not participate in the ancestral religious rites. Secondly, if they committed another crime after an amnesty (she) had been declared, they could be re-arrested for their original crime. In times of crisis, they could be called up again for service by the government, and if they had been bondservants and mutilated, when they were “freed” they were still registered separately from the rest of the commoner population. Convicts did, however, still retain some rights as commoners—they could still have wives and children who were not taken into government service, provided, of course, that the law of arrest (shou) had not been applied to them. Children of bondservant convicts, however, could inherit their status and become permanent bondservants. Thus there were clearly difficulties in determining short-term bondservants from long-term bondservants. Li itself seems to have been a general term for “slave,” and thus it is true that at least some Han slaves must have been descendants of such criminals. But the actual situation with regard to convicts and slaves in the Qin was much more complex than we 49  One such status was the “man-marmot” (renhe 人貉), whose children were forced to go and take care of their master. If they did not, they were confiscated by the government. If, however, they provided him with food, rather than taking care of him, they were not be confiscated, but given back to their master (SHD 4.175, 235; Hulsewé 1985a: D 174, 177). As Hulsewé observes, this item is extremely obscure.

Slavery In Early China: A Socio-cultural Approach

539

had previously believed. Significantly, if criminals’ families were enslaved by the government, this was known as “drowning” (mo 沒), a word that strongly implies social death, as I have explained. Below the convicts were these slaves, the property of their owners, who had no rights vis-à-vis their families, who were truly socially dead, truly the most liminal or marginal of all the groups we have discussed. The precise status of all these low groups is often difficult to determine accurately because with respect to the common people, they were all socially dead and not human, as the Discourses on Salt and Iron and the Zuozhuan so eloquently put it. But within that death there were degrees, just as within the really dead there were also different statuses, from the powerful dead kings and aristocrats, to the ordinary dead commoners and the ghosts (gui 鬼) (cf. Poo 1998). By Warring States times, it was the state that claimed the right to determine status within the family, not only by giving ranks to adult male members, thus altering the status relationships within the whole family and lineage, but also even by trying to punish parts of the lineage’s corporate property, the slaves and the family members. Such punishment radically affected the organization of such families and lineages, because punishment cut the family members off from their natal ritual ties. With the breaking of those ties, the li 禮 “rites” were destroyed and the individual became polluting not just to his family, but to the entire Chinese conceptual scheme and social order (McLeod 1978; Yates 1997). To take even a manumitted maidservant and make her a legitimate wife who had specific ritual duties to perform in the ancestral rites was thought in the Tang dynasty to disrupt the principles of heaven and earth.50 It would appear that this was already true in Warring States times. From this conceptual ordering of the world, with its emphasis on the preservation of family ties, of the rites, we can understand how crime was considered to be disrupting to the entire hierarchy that held the cosmos together. We can also see why convicts and even more so slaves were considered to be total 50  Wallace Johnson (1997: 155) translates the Tang code and subcommentary on this matter. The subcommentary reads as follows. “The wife is an equal as was said of the states of Ch’in (Qin) and Chin (Jin). Concubines can be bought and sold and so are very different from the wife. Slaves are inferior and basically not of the same class as commoners.  If a person makes the wife a concubine or makes a slave the wife, this violates and sets aside an agreement that was decided upon, injures the true way of husband and wife, and insults the principles of human relationships. If the distinction between upper and lower classes is upset and ritual and principles are scattered, those who commit such crimes are punished by two years of penal servitude.” For a comprehensive study of concubines, female slaves and courtesans in the Tang, see Liao Meiyun (1995).

540

Yates

outsiders, polluting and dangerous (cf. Douglas 1966). And finally, we can note how difficult it was for the Chinese to accept on equal terms peoples on their periphery, such as the nomads on the northern border and the native tribes of the south and southwest, as genuine human beings. They did not accept the rites and the agricultural way of life which defined the social and economic life of the truly Chinese, and thus by their very nature were of the same category as slaves. It was thus no accident that the ancient Chinese felt few qualms in enslaving them when the opportunity or the necessity arose. Finally, we need to note a remarkable implication of the conception of slaves and convicts as marginal, liminal people. Because they were cut off from normal human relationships and were thought of as socially dead, they themselves could touch the really dead, for both they and the corpses were dead, and were polluting to the rest of human society (cf. Yates 1997).51 In this essay, I have not been able to confirm the existence of a slave society or that the slave mode of production was ever the dominant mode in ancient China. Nevertheless, I believe I am correct in asserting that the social institution of slavery, together with its associated penal system, was a crucial factor in the creation of a unified Chinese empire and thus fundamentally affected the course of Chinese and consequently world history. Appendix One of the early Han legal cases found at Zhangjiashan 張家山 documents how a female slave Mei 美 attempted to reject her re-enslavement by her former master Dian 點 and subsequent sale to the Dafu Lu 大夫祿 (Jiangling Zhangjiashan Hanjian Zhengli Xiaozu 1993: 22; Li Xueqin 1993; Peng Hao 1993; Yates 1996). The text of the official report reads as follows.

51  It might be valuable to apply the cultural analysis suggested in this article to the study of eunuchs in Chinese history. The anomalous position of slaves in, or rather outside, society in the conceptual scheme of things might explain why eunuchs, the ultimate slaves, were so hated and yet so necessary. At various points in time, they mediated between the sacred Yang (quintessentially male) emperor and the general mass of the bureaucrats, courtiers, and common folk, who were Yin (quintessentially female in relation to the emperor) because they were degendered, neither male nor female, and were “pure” enough to be able to come into contact with the sacred emperor. During most of Chinese history, however, they stayed within the imperial palace ministering to the emperor’s private needs and to those of his empress and his harem. In this case too, they were permitted to do so because they were degendered males.

Slavery In Early China: A Socio-cultural Approach

541

The eleventh year, eighth month, bingxu day, jiashen being the first day of the month, Ao 驁, Assistant of Jiangling, ventures to submit to higher authority the (following) case: In the third month, jisi day, the Dafu Lu made a statement, saying: “In the sixth year, second month, I bought the slave Mei at the member of the rank and file (shiwu) Dian’s place. The price was 16,000 cash. But then in the third month, dingsi day, she absconded. I sought and captured Mei. Mei said, ‘I should not be considered a slave.’”

• Mei said, “Formerly I was Dian’s slave. In the time of Chu, I went off

and absconded. I surrendered and became a Han. My name was not written down or enumerated. Dian captured me and registered and enumerated me again as his slave. He sold me to Lu’s place. I (originally) had pledged myself; I should not suffer enslavement (again). Accordingly, I went off and absconded. The rest is as Lu (states).” Dian said, “Mei formerly was Dian’s slave. In the time of Chu, she absconded. In the sixth year, in the second month, I captured Mei. Mei did not yet have a name and was not yet enumerated. Then I registered and enumerated her, and sold her to Lu’s place. The rest is as Lu and Mei (state).” I (i.e., Assistant Ao) investigated Mei, (saying), “You, Mei, were formerly Dian’s slave. Although in the time of Chu, you went off and absconded, and surrendered and became a Han, you did not write down or enumerate (yourself). Dian captured (you) and registered and enumerated you. You (Mei) thereby again became a slave. That he sold you was appropriate. But you went off and absconded. What is your explanation?” Mei said, “In the time of Chu, I absconded. Dian then became a Han. He again enslaved me, and sold me. I had pledged myself. I ought not to have been considered a slave again. So then I went off and absconded. There is no other explanation.” I (Assistant Ao) questioned Mei. She is forty years old. The rest is as in the deposition. I (Assistant Ao) judged the case. Mei formerly was Dian’s slave. In the time of Chu, she absconded and surrendered to be a Han, but she did not write down her name or enumerate (herself). Dian captured (her) and registered and enumerated (her). He made her a slave again, and sold her to Lu’s place. But Mei went off and absconded. She is forty years old. She was captured. All (the above-mentioned facts) have been carefully investigated.

• •

• • •

542

Yates

• I suspect Mei is guilty of a crime, but this may have been discussed in another county. I venture to submit (the case) to higher authority; when I referred (it) to higher authority the response was that the post [investigation] was to be similar to the Granary Laws(?). The official (li) matching (dang 當) is: Tattoo Mei’s face and…, and give her to Lu. An alternative states, “She matches (deserves) to be a commoner (shuren).”

•

This document is very interesting on a number of points. It provides information, for example, on the price of slaves in the early Han, and evidence that Han officials closely followed Qin legal forms and procedures.52 It is the latest in the collection from Jiangling: the date of Assistant Ao’s submission corresponds to 196 bc, according to Li Xueqin (1993: 30) and Peng Hao’s (1993: 32) calculations. The slave Mei, being forty years old at the time of the case, must have been born well before the Qin unification, in 236 bc, presumably in the state of Chu, though the “time of Chu” probably refers to the time of the recreation of the Chu kingdom under Xiang Yu 項羽 after the collapse of Qin imperial authority. The crux of the issue, as Li Xueqin points out, is that Mei failed to register herself as a free commoner after she had absconded and submitted to Han authority. She hoped that since her master, an ordinary commoner, Dian, had also surrendered to the Han, and thus both of them had changed their recognition of what was the legal authority, the pledge of her person as debt-slave to Dian under Qin rule had been invalidated and she should no longer be considered a slave. Even though it is recorded in 202 bc that Han Gaozu seems to have manumitted slaves who had fallen into bondage as a result of poverty in the previous period,53 here it appears that one legal opinion was that she should be punished with tattooing of the face and returned to bondage in Lu’s house, thus upholding the validity of Dian’s re-enslavement and subsequent sale of her. I think that the Han official is suspicious about the circumstances of Mei’s failure to register. He suspects that she may indeed have registered, but that she committed some crime in another prefecture. It would appear that the

52  The contention in my article Yates (1987) that shiwu indicated commoner status and was not the name of a punishment is fully born out in the present case, for the shiwu Dian owns the slave Mei. 53   Han shu j. 1B, p. 4b; Dubs 1938 vol. 1: 104, reads, “As to those people who because of famine or hunger have themselves sold their persons to be slaves or slave-girls, let them all be freed and become common people.”

Slavery In Early China: A Socio-cultural Approach

543

Assistant could not send out an inquiry to other prefectures about her liability for crimes committed directly or horizontally. Instead, he had to submit the case up the administrative hierarchy where he hoped that the pertinent records would reveal the true situation regarding her potential culpability. If it was indeed true that Mei had registered, that she chose to conceal the fact, which would have released her from her slave status, might suggest that she had committed some heinous crime that would have resulted in a punishment even severer than enslavement. She may, however, have failed to register herself out of ignorance. Nevertheless, she had once been a commoner before pledging herself to Dian, so she should have known what the Qin system had been earlier on. In this case, it is apparent that both Dian and Lu are availing themselves of the state registration system to enforce their claims on Mei’s person. References Cited An Zuozhang 安作璋 (1981). “Cong Shuihudi Qinmu zhujian kan Qindai de nongye jingji 從睡虎地秦墓竹簡看秦代的農業經濟 (Looking at agricultural economics in the Qin period from the Qin slips at Shuihudi).” Qin Han shi luncong 秦漢史論叢 1: 27–40. Xi’an: Shaanxi Renmin Chubanshe. Anderson, Perry (1974). Passages from Antiquity to Feudalism. London: NLB. Beattie, Hilary (1979). Land and Lineage in China: A Study of T’ung-Ch’eng County. Anhwei in the Ming and Ch’ing Dynasties. Cambridge: Cambridge University Press. Bernhardt, Kathryn (1999). Women and Property in China, 960–1949. Stanford: Stanford University Press. Biot, Edouard (1969). Le Tcheou Li ou Rites des Tcheou. Reprint ed. Taibei: Chengwen Chubanshe. Birge, Bettine (2002). Women, Property, and Confucian Reactions in Sung and Yuan China (960–1368). Cambridge: Cambridge University Press. Boulais, Guy (1924). Manuel du Code Chinois. Shanghai: Imprimerie de la Mission Catholique. Buckland, W. W. (1908). The Roman Law of Slavery: The Condition of the Slave in Private Law from Augustus to Justinian. Cambridge: Cambridge University Press. Chang, Kwang-chih (1976). “Towns and cities in ancient China.” In Kwang-chih Chang, Early Chinese Civilization: Anthropological Perspectives, pp. 61–71. Harvard-Yenching Monograph series vol. 23. Cambridge: Harvard University Press. Chen Hanping 陳漢平 (1988). Xi Zhou ceming zhidu yanjiu 西周冊命制度研究 (Researches on the investiture system of the Western Zhou). Shanghai: Xuelin Chubanshe.

544

Yates

Ch’ü, T’ung-tsu (1972). Han Social Structure. Han Dynasty China, Vol. 1. Jack Dull (editor). Seattle: University Washington Press. Crook, J. A. (1967). Law and Life of Rome. London: Thames and Hudson. Douglas, Mary (1966). Purity and Danger: an Analysis of Concepts of Pollution and Taboo. Harmondsworth: Penguin. Du Zhengsheng 杜正勝 (1979). Zhoudai chengbang 周代城邦 (City-states in the Zhou period). Taibei: Lianjing Chuban Shiye Gongsi. Du Zhengsheng (1990). “‘Dan’ shi gongshe haishi jieshe? Yu Yu Weichao xiansheng shangque 單是公社還是結社?—與俞偉超先生商榷 (Is ‘Dan’ a commune or an association? A review of Yu Weichao’s theory).” Xin shixue 新史學 1990.1: 107–124. Dubs, Homer H. (1938–55). History of the Former Han Dynasty. Vols. 1–3. Baltimore: Waverly Press. Duyvendak, J. J. L. (trans.) (1963). The Book of Lord Shang. Chicago: University of Chicago Press. Ebrey Patricia Buckley (1993). The Inner Quarters: Marriage and the Lives of Chinese Women in the Sung Period. Berkeley: University of California Press. Finley, Moses I. (editor) (1968). Slavery in Classical Antiquity: Views and Controversies. Cambridge: Heffer and New York: Barnes and Noble. Finley, Moses I. (1980). Ancient Slavery and Modern Ideology. New York: Viking Press. Finley, Moses I. (1983). “Between slavery and freedom.” In Brent D. Shaw and Richard P. Sailer (eds.), Economy and Society in Ancient Greece, pp. 116–132. Harmondsworth: Penguin. Fogel, Robert W. and Stanley L. Engerman (1974). Time on the Cross: The Economics of American Negro Slavery. Boston and Toronto: Little, Brown. Gao Min 高敏 (1979). “Cong Yunmeng Qinjian kan Qin de tudi zhidu 從雲夢秦簡看秦 的土地制度 (Looking at the land system of the Qin from the Qin slips at Yunmeng).” In Yunmeng Qinjian chutan 雲夢秦簡初探, pp. 148–170. Xincheng: Henan Renmin Chubanshe. Gassman, Robert H. (2000). “Understanding ancient China and society: approaches to rén 人 and min 民.” Journal of the American Oriental Society 120.3: 348–59. Gong Changwei 宮長為 and Song Min 宋敏 (1982). “Lichenqie shi Qin guan nubi 隸臣 妾是秦官奴婢 (Lichenqie bondservants are Qin government slaves).” Zhongguo shi yanjiu 中國史研究 1: 120–26. Goody, Jack (1980). “Slavery in Time and Space.” In James L. Watson (editor), Asian and African Systems of Slavery, pp. 16–42. Oxford: Basil Blackwell. Guanzi 管子 (1934). Guoxue jiben congshu ed. Shanghai: Shangwu Yinshuguan. Guo Huaruo 郭化若 (1978). Shiyijia zhu Sunzi 十一家注孫子 ( Eleven commentators on Sunzi). Shanghai: Shanghai Guji Chubanshe. Guo, Moruo 郭沫若 (1973). Nuli zhi shidai 奴隸制時代 (The Slave Period). 2nd ed. Beijing: Renmin Chubanshe.

Slavery In Early China: A Socio-cultural Approach

545

Guo Qingfan 郭慶藩 (editor) (1978). Zhuangzi jishi 莊子集釋 (Collected explanations of the Zhuangzi). Beijing: Zhonghua shuju. Harrison, A. R. W. (1968). The Law of Athens: The Family and Property. Oxford: Clarendon Press. Hindess, Barry and Paul Q Hirst (1975). Pre-capitalist Modes of Production. London and Boston: Routledge and Kegan Paul. Hori Tsuyoshi 堀毅 (1988). Qin Han fazhi shi lunkao 秦漢法制史論考 (Studies on the history of the Qin and Han legal system). Beijing: Falü Chubanshe. Hsu, Cho-yun (1965). Ancient China in Transition: An Analysis of Social Mobility 770–222 B.C. Stanford: Stanford University Press. Hu Liuyuan 胡留元 and Feng Zhuohui 馮卓慧 (1988). Xi Zhou fazhi shi 西周法制史 (History of the Western Zhou legal system). Xi’an: Shaanxi Renmin Chubanshe. Huang Jinyan 黃今言 (1981). “Qindai zufu yaoyi zhidu chutan 秦代租稅徭役制度初探 (Preliminary investigation of the tax and corvée labor system in the Qin period).” Qin Han shi luncong: 61–73. Huang, Philip C. C. (1996). Civil Justice in China: Representation and Practice in the Qing. Stanford: Stanford University Press. Huang, Philip C. C. (2001). Code, Custom, and Legal Practice in China: The Qing and the Republic Compared. Stanford: Stanford University Press. Huang Ranwei 黃然偉 (1978). Yin Zhou qingtongqi shangsi mingwen yanjiu 殷周青銅 器賞賜銘文研究 (Researches on reward and bestowal inscriptions in Yin and Zhou bronze vessels). Hong Kong: Longmen Shudian. Huang Shibin 黃士斌 (1982). “Henan Yanshixian faxian Handai maitian yueshu shiquan 河南偃師縣發現漢代買田約束石券 (The Han period stone binding contract for purchase of fields discovered in Yanshi County, Henan).” Wenwu 文物 1982.12: 17–20. Huang Zhongye 黃中業 (1991). Qinguo fazhi jianshe 秦國法制建設(The structure of the Qin state legal system). Shenyang: Liao Shen Shushe. Hulsewé, A. F. P. (1955). Remnants of Han Law vol.1. Leiden: E.J. Brill. Hulsewé, A. F. P. (1978). “The Ch’in documents discovered in Hu-pei in 1975.” T’oung Pao 64 (1978): 175–217, 338. Hulsewé, A. F. P. (1985a). Remnants of Ch’in Law: An Annotated Translation of the Ch’in Legal and Administrative Rules of the 3rd Century B.C. Discovered in Yün-meng Prefecture, Hu-pei Province, in 1975. Leiden: E.J. Brill. Hulsewé, A. F. P. (1985b). “The influence of the ‘Legalist’ government of Qin on the economy as reflected in the texts discovered in Yunmeng County.” In Stuart R. Schram (editor), The Scope of State Power in China, pp. 211–35. London and Hong Kong: St. Martin’s Press, School of Oriental and African Studies, University of London, and the Chinese University of Hong Kong.

546

Yates

Hulsewé, A. F. P. (1988). “The wide scope of tao 盜 ‘theft’ in Ch’in-Han law.” Early China 13: 166–200. Hung, William (editor) (1966). Harvard-Yenching Institute Sinological Index series Supplement no. 11. Combined Concordances to the Ch’un-Ch’iu, Kung-yang, Ku-liang and Tso-chuan. 4 vols. Taipei: Chinese Materials and Research Aids Service Center. Hutson, Rev. J. (1919). “Chinese life on the Tibetan foothills: Book II-Social-continued.” The New China Review 1 (6): 547–561. Jiangling Zhangjiashan Hanjian Zhengli Xiaozu 江陵張家山漢簡整理小組 (1993). “Jiangling Zhangjiashan Hanjian shiwen (yi) 江陵張家山漢簡釋文 (一) (Transcrip­ tion of the Han slips from Zhangjiashan, Jiangling, part 1.” Wenwu 1993.8: 22–25. Johnson, Wallace (1997). The T’ang Code : Volume II, Specific Articles. Princeton : Princeton University Press. Jolowicz, H. F. and Barry Nicholas (1978). Historical Introduction to the Study of Roman Law. 3rd ed. Cambridge: Cambridge University Press. Kalinowski, Marc (1986). “Les Traités de Shuihudi et l’Hémérologie chinoise à la Fin des Royaumes-combattants.” T’oung Pao 72: 175–228. Kan, Sergei (1989). Symbolic Immortality: the Tlingit Potlatch of the Nineteenth Century. Washington: University of Washington Press. Karras, Alan L. (1991). “Of human bondage: creating an Atlantic history of slavery.” Journal of Interdisciplinary History 22: 285–293. Keightley, David N. (1969). Public Work in Ancient China: A Study of Forced Labor in the Shang and Western Chou. Ph.D. dissertation. New York: Columbia University. Lau, Ulrich (1999). Quellenstudien zur Landvergabe und Bodenübertragung in der westlichen Zhou-Dynastie (1045?–771 v. Chr.) Nettetal: Steyler Verlag. Li Hengmei 李衡眉 (1999). “Qiguo nuli kaoshu 齊國奴隸考述 (Report on the investigation of slaves in the state of Qi).” In Li Hengmei, Xian Qin shi lunji 先秦史論集. Ji’nan: Qilu Shushe. Li Jiemin 李解民 (1981). “‘Kai qianmo’ bianzheng 開阡陌辨正 (Identification and correction of errors concerning ‘opening up the qianmo field paths’).” Wenshi 文史 1981.11:47–60. Li Jing 栗勁 (1985). Qin lü tonglun 秦律通論 (Grand survey of Qin law). Ji’nan: Shandong Renmin Chubanshe. Li Jiping 李季平 (1986). Tangdai nuli zhidu 唐代奴隸制度 (The slave system in the Tang period). Shanghai: Shanghai Renmin Chubanshe. Li Xinda 李新達 (1978). “Guanyu Qin Han ‘cangtou’ de wenti 關於秦漢蒼頭的問題 (On the question of Han ‘green heads’).” Wenshizhe 文史哲 1978.2: 61–67. Li Xueqin 李學勤 (1993). “Zou yan shu jieshuo (shang) 奏讞書解說 (上) (Explanation of the Book of Reports on Deliberations of Crimes).” Wenwu 1993.8: 26–31. Liao Meiyun 廖美雲 (1995). Tangji yanjiu 唐伎研究 (Researches into Tang female entertainers and prostitutes). Taibei: Taiwan Xuesheng Shuju.

Slavery In Early China: A Socio-cultural Approach

547

Liao, W. K. (trans.) (1959). Complete Works of Han Fei Tzu. London: Probsthain. Lin Jianhan 林振翰 (editor) (1934). Yantie lun 鹽鐵論 (Discourses on salt and iron). Guoxue jiben congshu ed. Shanghai: Shangwu Yinshuguan. Lin Jianming 林劍鳴 (1981). “Qinguo nulizhi shehui xingtai de tedian 秦國奴隸制社 會形態的特點 (Special points about the social form of the slave system in the Qin state).” Qin Han shi luncong 秦漢史論叢 1: 3–15. Linyi Yinqueshan Zhujian Zhengli Xiaozu 臨沂銀雀山漢墓竹簡整理小組 (1985). “Yinqueshan zhushu Shoufa Shouling deng shisan pian 銀雀山竹書守法守令等十 三篇 (The thirteen bamboo books of the methods of defence, orders for defence, etc., from Yinqueshan).” Wenwu 1985.4: 27–37. Liu Hainian 劉海年 (1985). “Guanyu Zhongguo suixing de qiyuan: jian tan Qin xingtu de xingqi ji lichenqie de shenfen, shang 關於中國歲刑的起源—兼談秦刑徒的刑期 及隸臣妾的身份, 上 (On the origin of punishments in years in China—and a concurrent discussion of time-limitation punishments and lichenqie bondservants in the Qin, part A).” Faxue yanjiu 法學研究 1985.5: 67–76. Liu Hainian (1986). “Guanyu Zhongguo suixing de qiyuan: jian tan Qin xingtu de xingqi ji lichenqie de shenfen, xia 關於中國歲刑的刑的起源—兼談秦刑徒的刑期及隸臣 妾的身份, 下 (On the origin of punishments in years in China—and a concurrent discussion of time-limitation punishments and lichenqie bondservants in the Qin, part B).” Faxue yanjiu 1985.6: 68–73. Liu Lexian 劉樂賢 (1994). Shuihudi Qinjian rishu yanjiu 睡虎地日書研究 (Researches on the almanac books in the Qin slips from Shuihudi). Taibei: Wenjin Chubanshe. Liu Weimin 劉偉民 (1975). Zhongguo gudai nubi zhidu shi 中國古代奴婢制度史 (History of the ancient Chinese slave system). Hong Kong: Longmen Shudian. Liu Xiang 劉向 n.d. Shuo yuan 說苑 (Garden of sayings). Sibu congkan ed. Shanghai: Shangwu Yinshuguan. Liu Yin 劉寅 (editor) (1975). Liutao zhijie 六韜直解 (Direct explanations of the Liu tao). Kambun taikei ed. Tokyo: Fuzambö. Loewe, Michael (1974). “The Grand Inquest—81 B.C.” In Crisis and Conflict in Han China 104 B.C. to A.D. 91, pp. 91–112. London: George Allen and Unwin. Loewe, Michael (1994). “The almanacs ( jih-shu) from Shui-hu-ti: a preliminary survey.” In Divination, Mythology and Monarchy in Han China, pp. 214–35. Cambridge: Cambridge University Press. McKnight, Brian E. (1981). The Quality of Mercy: Amnesties and Traditional Chinese Justice. Honolulu: University of Hawaii Press. McKnight, Brian E. and James T. C. Liu (trans.) (1999). The Enlightened Judgments Ch’ing-ming Chi: The Sung Dynasty Collection. Albany: State University of New York Press. McLeod, Katrina C. D. (1978). “Law and the symbolism of pollution: some observations on the Ch’in laws from Yun-meng.” Unpublished paper, Annual Meeting of the Association for Asian Studies, Chicago.

548

Yates

McLeod, Katrina C. D. and Robin D. S. Yates (1981). “Forms of Ch’in law: an annotated translation of the Feng-chen shih.” Harvard Journal of Asiatic Studies 41.1: 111–163. Meijer, Marinus J. (1980). “Slavery at the end of the Ch’ing Dynasty.” In Jerome Alan Cohen, R. Randle Edwards, and Fu-mei Chang Ch’en (eds.), Essays on China’s Legal Tradition, pp. 327–358. Princeton, N.J.: Princeton University Press. Meillassoux, Claude (editor) (1975). L’Esclavage en Afrique Precoloniale. Paris: Maspero. Meillassoux, Claude (1978). “Correspondence (On slavery).” Economy and Society 7: 321–331. Miers, Suzanne and Igor Kopytoff (editors) (1977). Slavery in Africa: Historical and Anthropological Perspectives. Madison: University of Wisconsin Press. Miller, Joseph C. (1985). Slavery: A Worldwide Bibliography 1900–1982. White Plains, N.Y.: Kraus International. Musha Akira 武者章 (1980). “Sei-Shü sakumei kinbun bunrui no kokoromi 西周冊命金 文分類の試み (An attempt at a comprehensive classification of the Western Zhou bronze inscriptions: the case of the investiture of feudal vassals (ceming jinwen).” In Matsumaru Michio 松丸道雄 (editor), Sei-Shū seidōki to sono kokka 西周青銅器 とその国家 (Western Zhou bronzes and states), pp. 241–324. Tōkyō: University of Tokyo Press. Nieboer, H. J. (1900). Slavery as an Industrial System: Ethnological Researches. The Hague: Martinus Nijhoff. Niida Noboru 仁井田陞 (1944). Shina mibunhō shi 支那身分法史 (History of legal status in China). Tokyo: Zayūhō kankōkai. Ning Ke 寧可 (1982). “Guanyu Han Shiting li Fulao dan maitian yueshu shiquan 關於漢 侍廷里父老僤買田約束石券 (On the Han binding stone contract for the purchase of fields of the Fulao and Dan association of Shiting Village).” Wenwu 1982.12: 21–27. Ōba Osamu 大庭脩 (1991). Qin Han fazhi shi yanjiu 秦漢法制史研究 (Researches into the history of the Qin and Han legal system). Translated by Lin Jianming et al. Shanghai: Shanghai Renmin Chubanshe. Padgug, Robert A. (1976). “Problems in the theory of slavery and slave society.” Science and Society 40: 3–27. Parish, Peter J. (1989). Slavery: History and Historians. New York: Harper and Row. Patterson, Orlando (1979). “On slavery and slave formations.” New Left Review 117: 31–67. Patterson, Orlando (1982). Slavery and Social Death: A Comparative Study. Cambridge, MA: Harvard University Press. Peng Hao 彭浩 (1993). “Tan Zou yan shu zhong de Xi Han anli 談奏讞書中的西漢案例 (Discussion of the examples of cases in the Book of Reports on the Deliberations of Crimes.” Wenwu 1993.8: 32–36. Poo, Mu-chou (1998). In Search of Personal Welfare: A View of Ancient Chinese Religion. Albany: State University of New York Press.

Slavery In Early China: A Socio-cultural Approach

549

Pulleyblank, Edwin G. (1958). “The origins and nature of chattel slavery in China.” Journal of the Economic and Social History of the Orient 1: 185–220. Qiu Xigui 裘錫圭 (1980). “Zhanguo wenzi zhong de ‘shi’ 戰國文字中的市 (‘Markets’ in Warring States writings).” Kaogu xuebao 考古學報111: 285–296. Rao Zongyi 饒宗頤 and Zeng Xiantong 曾憲通 (1993). Chudi chutu wenxian sanzhong yanjiu 楚地出土文獻三種研究 (Three different researches into texts excavated from Chu territory). Beijing: Zhonghua Shuju. Reed, Carrie H. (2000). “Tattoo in early China.” Journal of the American Oriental Society 120.3: 360–76. Schulz, Fritz (1936). Principles of Roman Law. Oxford: Clarendon Press. Schurmann, H. F. (1956). “Traditional property concepts in China.” Far Eastern Quarterly 15: 507–516. Shaughnessy, Edward L. (1991). Sources of Western Zhou History: Inscribed Bronze Vessels. Berkeley: University of California Press. SHD = Shuihudi Qinmu zhujian Zhengli Xiaozu 睡虎地秦墓竹簡整理小組 (1978). Shuihudi Qinmu zhujian 睡虎地秦墓竹簡 (The bamboo slips from the Qin tomb at Shuihudi). Beijing: Wenwu Chubanshe. Shima Kunio 島那邦 (1958). Inkyo bokuji kenkyū 殷墟卜辭研究 (Researches into the oracle bone inscriptions from the wastes of Yin). Hirosaki: Chūgokugaku Kenkyūkai. Shuihudi Qinmu zhujian Zhengli Xiaozu (1990). Shuihudi Qinmu zhujian 睡虎地 秦墓竹簡 (The bamboo slips from the Qin tomb at Shuihudi). Beijing: Wenwu Chubanshe. Sun Yirang 孫詒讓 (editor) (1983). Zhouli zhengyi 周禮正義 (Correct Interpretations of the Rites of Zhou). Sibu beiyao ed. Taibei: Zhonghua Shuju. Takigawa Kametarō 瀧川龜太郎 (1979). Shiki kaichu kōshō 史記會注考證 (Assembled commentaries and textual criticism on the Shi ji). Taibei: Mingyu Chubanshe. Tang Zangong 唐贊功 (1981). “Yunmeng Qinjian guansi nuli wenti shitan 雲夢秦簡官 私奴隸問題試探 (Preliminary discussion of the question of private and government slaves in the Qin slips from Yunmeng).” Zhonghua wenshi luncong 中華文史 論叢 1981.3: 143–63. Tomiya Itaru 富谷至 (1985). “Renzasei to sono shūhen 連座制とその周辺 (The system of linked responsibility and related issues).” In Hayashi Minao 林巳奈夫 (editor), Sengoku jidai shutsudo bunbutsu no kenkyū 戦国時代出土文物の研究, pp. 475–527. Kyōto: Kyōto Daigaku Jinbun Kagaku Kenkyūjo. Wang Xianqian 王先謙 (editor) n.d. Han shu buzhu 漢書補注 (Supplementary commentary on the Han shu). Taibei: Yiwen Yinshuguan. Wang Xianshen 王先慎 (editor) (1956). Hanfeizi jijie 韓非子集解 (Collected explanations of the Hanfeizi). Taiwan: Shangwu Yinshuguan. Wang, Yi-t’ung (1953). “Slaves and other comparable social groups during the Northern Dynasties (386–618).” Harvard Journal of Asiatic Studies 16: 293–364.

550

Yates

Watson, James L. (editor) (1980a). Asian and African Systems of Slavery. Oxford: Basil Blackwell. Watson, James L (1980b). “Slavery as an institution, open and closed systems.” In Watson (1980a), pp. 1–15. Oxford: Basil Blackwell. Watson, James L (1980c). “Transactions in people: the Chinese market in slaves, servants and heirs.” In Watson (1980a), pp. 223–250. Oxford: Basil Blackwell. Watson, James L (1982). “Chinese kinship reconsidered: anthropological perspectives on historical research.” China Quarterly 92: 589–622. Wilbur, C. Martin (1943). Slavery in China during the Former Han Dynasty. Anthropological Series, vol. 34. Chicago: Chicago Field Museum of Natural History. Wu Fuzhu 吳福助 (1994). Shuihudi Qinjian lunkao 睡虎地秦簡論考 (Studies of the Qin slips from Shuihudi). Taibei: Wenjin Chubanshe. Wu Hui 吳慧 (1981). Sang Hongyang yanjiu 桑弘羊研究 (Researches on Sang Hongyang). Jinan: Qilu Shushe. Wu Rongzeng 吳榮曾 (1995a). “Xumi shitan—lun Zhanguo shi de xingtu zhi 胥靡試 探—論戰國時的刑徒制 (Exploration of xumi: a discussion of the convict system of the Warring States period).” In Xian Qin liang Han shi yanjiu 先秦兩漢史研究, pp. 148–61. Beijing: Zhonghua Shuju. Wu Rongzeng (1995b). “Jianmen kao 監門考 (An investigation of jianmen).” In Xian Qin liang Han shi yanjiu, pp. 162–71. Beijing: Zhonghua Shuju. Xing Yitian 邢義田 (1987). Qin Han shi lungao 秦漢史論稿 (Essays on Qin and Han history). Taibei: Dongda Tushu Gongsi. Xiong Tieji 熊鐵基 and Wang Ruiming 王瑞明 (1981). “Qindai de fengjian tudi suoyou zhi 秦代的封建土地所有制 (The comprehensive system of feudal land in the Qin period).” In Yunmeng Qinjian yanjiu 雲夢秦簡研究: 67–78. Beijing: Zhonghua Shuju. Xu Fuchang 徐富昌 (1993). Shuihudi Qinjian yanjiu 睡虎地秦簡研究 (Researches into the Qin slips from Shuihudi). Taibei: Wenshizhe Chubanshe. Xu Yangjie 徐揚杰 (1981). “Hanjian zhong suojian wujia kaoshi 漢簡中所見物價考釋 (Explanations of the price of goods seen in the Han slips).” Zhonghua wenshi luncong 111.3. Xu Zhongshu 徐中舒 (1981). Shuowen jiezi duanzhu 說文解字段注 (Duan Yucai’s commentary on the Shuowen jiezi). 2 vols. Chengdu: Chengdu Guji Shudian. Yang Heding 楊禾丁 (1983). “Jiamen yu jianmen 叚門與監門 ( Jiamen and jianmen).” Zhonghua wenshi luncong 111: 223–237. Yang Ximei 楊希枚 (1956). “Xian Qin zhuhou shoujiang xianjie qianfu zhidu kao 先秦 諸侯受降獻捷遣俘制度考 (Studies on the system of accepting surrender, awarding prisoners of war, and repatriating prisoners among the lords in the pre-Qin period).” Lishi yuyan yanjiusuo jikan 歷史語言研究所集刊 27: 107–116. Yates, Robin D. S. (1987). “Social status in the Ch’in: evidence from the Yün-meng legal documents. Part one: commoners.” Harvard Journal of Asiatic Studies 47: 197–236.

Slavery In Early China: A Socio-cultural Approach

551

Yates, Robin D. S. (1988). “New light on ancient Chinese military texts: notes on the nature and evolution, and the development of military specialization in Warring States China.” T’oung Pao 74: 211–248. Yates, Robin D. S. (1996). “The exercise of discourse: Qin popular culture and the Qin state.” Paper presented at the 48th Annual Meeting of the Association for Asian Studies, April 11–14, 1996. Yates, Robin D. S. (1997). “Purity and pollution in early China.” In Zang Zhenhua 臧振華 (editor), Zhongguo kaoguxue yu lishixue zhenghe yanjiu 中國考古學與歷史 學整合研究 (Integrated studies of Chinese archaeology and history), pp. 479–536. Symposium Series of the Institute of History and Philology, Academia Sinica, No. 4. Taipei: Academia Sinica. Yu Haoliang 于豪亮 (1985a). “Qinjian zhong de nuli 秦簡中的奴隸(Slaves in the Qin slips).” In Yu Haoliang xueshu wencun 于豪亮學術文存, pp. 116–23. Beijing: Zhonghua Shuju. Yu Haoliang (1985b). “Yunmeng Qinjian suojian zhiguan shulue 雲夢秦簡所見職 官述略 (Brief account of the administrative positions seen in the Qin slips from Yunmeng).” In Yu Haoliang xueshu wencun, pp. 88–115. Beijing: Zhonghua Shuju. Yu Weichao 俞偉超 (1988). Zhongguo gudai gongshe zuzhi de kaocha: lun xian Qin liang Han de dan-dan-dan 中國古代公社組織的考察—論先秦兩漢的單–僤–彈 (Investigation into the organization of communes in ancient China: a discussion of dan, dan, and dan in the pre-Qin and two Han periods). Beijing: Wenwu Chubanshe. Yu Zongfa 余宗發 (1992). Yunmeng Qinjian zhong sixiang yu zhidu gouzhi 雲夢秦簡 中思想與制度鉤摭 (A probe into the thought and system of the Qin slips from Yunmeng). Taibei: Wenjin Chubanshe. Yunmeng Shuihudi Qinmu Bianxie Zu 雲夢睡虎地秦墓編寫組 (1981). Yunmeng Shuihudi Qinmu 雲夢睡虎地秦墓 (The Qin tomb at Shuihudi, Yunmeng). Beijing: Wenwu Chubanshe. Yunmeng Qinjian Zhengli Xiaozu 雲夢秦簡整理小組 (1976a). “Yunmeng Qinjian shiwen (yi) 雲夢秦簡釋文 (一) (Transcriptions of the Qin slips from Yunmeng [1]).” Wenwu 1976.6: 11–14. Yunmeng Qinjian Zhengli Xiaozu (1976b). “Yunmeng Qinjian shiwen (er) (Transcrip­ tions of the Qin slips from Yunmeng [2]).” Wenwu 1976.7: 1–11. Yunmeng Qinjian Zhengli Xiaozu (1976c). “Yunmeng Qinjian shiwen (san) (Transcriptions of the Qin slips from Yunmeng [3]).” Wenwu 1976.8: 27–37. Zhang Chuanxi 張傳璽 (1985). Qin Han wenti yanjiu 秦漢問題研究(Researches into questions concerning the Qin and Han). Beijing: Beijing Daxue Chubanshe. Zhongguo Renmin Jiefangjun Junshi Kexueyuan 中國人民解放軍軍事科學院 and Zhanzheng Lilun Yanjiubu Sunzi Zhushi Xiaozu 戰爭理論研究部孫子注釋小組 (1977). Sunzi bingfa xinzhu 孫子兵法新注 (New commentary on Sunzi’s Art of War). Beijing: Zhonghua shuju.

552

Yates

Zhu Shaohou 朱紹侯 (1981). “Qin Han shidai de nuli yifu nongmin he qita laodongzhe: Qin Han tudi zhidu he jieji guanxi yanjiu 秦漢時代的奴隸依附農民和其他勞動 者—秦漢土地制度和階級關係研究 (Slaves, dependent peasants, and other workers in the Qin period: the Qin and Han class and land system).” Qin Han shi luncong 1: 227–242. Zhu Shiche 朱師轍 (1974). Shangjunshu jiegu dingben 商君書解詁定本(Final version of the explanatory notes of the Book of Lord Shang). Hong Kong: Zhonghua Shuju.

Empire, Monotheism and Slavery in the Greater Mediterranean Region from Antiquity to the Early Modern Era* Jeffrey Fynn-Paul I Introduction Nearly a decade ago, the American Historical Review held a forum entitled ‘Crossing Slavery’s Boundaries’, in which David Brion Davis called for a longerterm and geographically broader view of the Atlantic slave system, one that went beyond paying lip service to its antecedents in the Canary Islands, and which delved into the structures of the European, African and western Asian slave networks from which the Atlantic slave system was calved.1 Whereas until quite recently it was believed that the Atlantic slave trade was a largely selfcontained phenomenon, it is now acknowledged by most writers that this slave trade is part of a much wider picture, which includes traditional African slave systems and the advance of Islam in Africa.2 Recent volumes such as Davis’s Source: Fynn-Paul, Jeff, “Empire, Monotheism and Slavery in the Greater Mediterranean Region from Antiquity through the Early Modern Era,” Past and Present, 205(1) (2009): 3–40. By permission of Oxford University Press. * I would like to thank Melanie Newton of the University of Toronto, whose course on Caribbean history inspired her teaching assistant to think more deeply about the antecedents and longterm dynamics of the Atlantic slave system. I would also like to thank Derek Mancini-Lander of the University of Michigan, whose insights into Islamic historiography have saved me from much error. And additional thanks to Ewout Frankema of Utrecht University, for his advice on economic theory. 1  David Brion Davis, ‘Looking at Slavery from Broader Perspectives’, contribution to the forum ‘Crossing Slavery’s Boundaries’, Amer. Hist. Rev., cv (2000). 2  For the older view, see, for example, Vincent Bakpetu Thompson, The Making of the African Diaspora in the Americas, 1441–1900 (London, 1987); Herbert S. Klein, Slavery in the Americas: A Comparative Study of Virginia and Cuba (Chicago, 1967). For an intermediate phase, which acknowledges the impacts of Islamic and European slavery on Africa, but still sees them as essentially independent, see Paul E. Lovejoy, Transformations in Slavery: A History of Slavery in Africa (Cambridge, 1983), ch. 1. For a more recent synthesis, see Akosua Adoma Perbi, A History of Indigenous Slavery in Ghana from the Fifteenth to the Nineteenth Century (Legon, 2004). See also Paul E. Lovejoy (ed.), Slavery on the Frontiers of Islam (Princeton, 2004), which comparatively addresses the impacts of the Muslim–pagan trade within Africa and the transatlantic slave trade. © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_021

554

Fynn-Paul

own Inhuman Bondage and the important essay collection Slave Systems: Ancient and Modern have invited us to find new parallels in what were formerly seen as discrete problems in the history of slavery.3 As the present article will show, our increasing understanding of the relations between Christianity and Islam in the Middle Ages can now enable us to take a more synthesized view of the history of what I shall call ‘Greater Mediterranean slavery’ from antiquity to the end of the early modern era.4 To do so, I shall locate Atlantic and African slavery within a Greater Mediterranean slave system, one which has been evolving since the rise of the Egyptian and Sumerian empires. In order to grasp better how this system worked, this article introduces the concepts of ‘slaving zones’ and ‘no-slaving zones’. A slaving zone is defined as the geographical area impacted by a given society’s demand for slaves, and a no-slaving zone is the area considered off limits for slave raiding by that society. In the light of these concepts, we shall see that the rise of the first empires inaugurated a new era in the history of slavery, since the new empires constituted the first large-scale no-slaving zones, even as they created the first large-scale slaving zones in the lands beyond their borders. However, the no-slaving zones of the first empires were imperfect at best, since many zone inhabitants could be reduced to slavery via economic or judicial means. But the creation of the Christian and Islamic monotheistic blocs was a major turning point in the history of the Greater Mediterranean slave system, since these empires came to adopt a religio-ethical taboo against the enslavement of the majority of their inhabitants. In the process, they created what might be called the world’s first ‘perfect’ no-slaving 3  David Brion Davis, Inhuman Bondage: The Rise and Fall of Slavery in the New World (Oxford, 2008); Enrico Dal Lago and Constantina Katsari (eds.), Slave Systems: Ancient and Modern (Cambridge, 2008). The latter volume, to my mind, still shows a tendency for writers on slavery to move too quickly from the ancient to the early modern period, without exploring many medieval continuities; this is in part due to the fact that the bulk of these continuities lie south and east of western Europe, but such a situation should increasingly prove to be less of a hindrance to exploration and synthesis. 4  By ‘Greater Mediterranean’ is meant all of Asia west of the Indus, most of Africa, and all of Europe. It will be shown that these were all drawn into the civilization nexus which evolved around the Mediterranean (and the Fertile Crescent, which can be viewed as periMediterranean) well before the early modern era. The Atlantic world, and particularly its slave system, should in this view be seen as an extension of this Greater Mediterranean slave system. Michael McCormick’s The Origins of the European Economy: Communications and Commerce, AD 300–900 (Cambridge, 2001) provides a true ‘missing link’ between slavery in the classical and high medieval periods, and also opens up a new understanding of the relationship between Islamic and Christian slaving activity during the early medieval period.

EMPIRE, MONOTHEISM and SLAVERY

555

zones. Perfect no-slaving zones largely eliminated many of the means by which traditional slaving societies maintained slave supplies, including the economic and judicial paths to enslavement of zone inhabitants mentioned above. Both of the new monotheistic blocs therefore had to rely more strongly on external sources of slave supply, and the result was an increased pressure on the slaving zones at their peripheries. The benefits of such a view are many. It reveals that Greater Mediterranean slaving zones have shifted according to very specific patterns of political and religious organization ever since the time of Egypt and Sumer. Over time, almost every part of the Greater Mediterranean region has been a slaving zone, exploited to a greater or lesser degree as the boundaries of civilization crept ever outward from their river-valley epicentres. In terms of African slavery, this model suggests that profound structural changes occurred after the rise of civilization in Egypt and, later, after the rise of Christian and Islamic monotheism. Also, my focus on monotheism as a structural force serves to highlight the many, too often overlooked, similarities between the African and Russian slaving zones. For over a thousand years these areas were impacted in similar ways by virtue of their location on the frontiers of the Christian and Islamic no-slaving zones. More concretely, the rise of Christian and Islamic monotheism created a hyper-exploitation of African and Russian populations for slaving purposes which has continued until quite recently in both regions. The final vestiges of this system still linger in parts of Africa today, and can be seen as the twilight of a Greater Mediterranean slave system which, taking its pre- and post-monotheistic phases together, has been continuously evolving for over five thousand years. From the perspective of historians of New World slavery, this model is beneficial in so far as it provides greater agency to African elites than many older views, while making clear that African elites played a role structurally quite similar to the roles played by other elites in the history of the Greater Mediterranean slave system.5 But this type of structural analysis, for all its value, should not be seen as in any way downplaying the enormous damage done by racist attitudes which have been magnified by the timing and volume of the African slave trade, or by the associations with slavery which still so

5  The classic discussion on the varieties of indigenous slavery in Africa remains Suzanne Miers and Igor Kopytoff (eds.), Slavery in Africa: Historical and Anthropological Perspectives (Madison, 1977), esp. 12–14. See also the newly re-edited study by Humphrey J. Fisher, Slavery in the History of Muslim Black Africa (New York, 2001), for an extensive discussion, albeit one which concentrates on the nineteenth century.

556

Fynn-Paul

strongly influence modern perceptions of Africanness.6 Nor, for that matter, should this structural analysis be seen as relieving of guilt any person of any society who, confronted with the realities of a slave system, chose to behave less humanely or idealistically than they could have done. It does, however, explain why people as disparate as ancient Egyptians, Vikings, Muslims of Córdoba and early modern Africans of many disparate tribes and nations found themselves born into a society which was at that time playing a particular role in a long-established edifice of slavery and empire. It has long been recognized that European colonialism did not ‘create’ a demand for slaves in Africa ex nihilo, but rather operated on a pre-existing system. The present model now enables us to examine the effects of this imperial demand more precisely, in so far as we can see that a similar demand has played corresponding roles in every Greater Mediterranean society which has found itself drawn into an imperial slaving zone. It will be shown that wherever this outside pressure occurred, and in proportion to the strength of this pressure, imperial demand could gradually overburden traditional eco-social relationships, including traditional slavery relationships, until they became unsustainable. Thus, imperialism has been causing ‘unnatural’ slave demand since the dawn of the imperial age over five thousand years ago. Monotheistic imperialism made this demand even more acute. II

Slavery in the Ancient World: ‘Imperfect’ No-Slaving Zones

Slavery was an integral feature of almost every ancient society of Europe, Africa and Asia.7 The reasons for this ubiquity are not difficult to deduce: prior to the rise of the first empires, no ruler monopolized violence in an area 6  An excellent synthesis on the origins of racism against blacks can now be found in Davis, Inhuman Bondage, ch. 2. See also David M. Goldenberg, The Curse of Ham: Race and Slavery in Early Judaism, Christianity, and Islam (Princeton, 2003); Benjamin Braude, ‘Cham et Noé: race, esclavage et exégèse entre islam, judaïsme et christianisme’, Annales: histoire, sciences sociales, lvii (2002). For one of the first attempts to trace the origins of racial prejudice back into Graeco-Roman society, see Benjamin Isaac, The Invention of Racism in Classical Antiquity (Princeton, 2004). 7  Michael Bush, in the introduction to M. L. Bush (ed.), Serfdom and Slavery: Studies in Legal Bondage (London, 1996), somewhat misleadingly implies that slave systems have only been evolving over the past 2,500 years. In fact, of course, slave systems existed from the earliest times of Sumer and Egypt. For an ageing but still valuable list of studies on ancient slavery in all these areas, see Bernard Lewis, Race and Slavery in the Middle East: An Historical Enquiry (Oxford, 1990), 3 n. 2.

EMPIRE, MONOTHEISM and SLAVERY

557

greater than a few square miles, and so continual warfare was an ineluctable fact of life. The question of what to do with conquered neighbours can entail only four possible solutions: one can ignore or pardon the vanquished, one can slaughter them, or demand tribute, or enslave them. Most of the time, some combination of these options has been employed; all of them carried dangers and benefits that varied from situation to situation. Agricultural peoples might have found slavery to be more beneficial than would hunter-gatherers, since agriculturalists could effectively house and thus restrain slaves, and there was always additional farm labour to be done. It may well be that hunter-gatherer societies practised slavery on a smaller scale than agriculturalists because of smaller and less certain food supplies. With war so widespread, and trade networks yet primitive, it is likely that skirmishing and raiding were major sources of slaves in this early period.8 The rise of the first empires would have had a novel and significant effect on this traditional pattern of slave-taking and slave-keeping, since an empire is an area where one ruler’s monopoly of violence extends over a large territory, and which incorporates previously self-governing peoples. When the union of the upper and lower Nile created the Egyptian Old Kingdom, all the inhabitants of the Nile valley were now, for the first time in history, nominally ‘off limits’ to each other as potential slave stock, at least in terms of active raiding. Thus the creation of an empire in effect created the first large-scale no-slaving zone. Ancient no-slaving zones, prior to the rise of monotheism, were not, however, perfect, in so far as those born within a given zone could still become the slaves of others who were native to the zone. There were many channels by which Egyptians, Greeks, Babylonians, Romans, Celts and ancient Germans, for example, could become the slaves of their own countryfolk. The fact that slaves came to be regarded throughout the ancient world as just another economic commodity ensured that money and property concerns, rather than any ethical code, determined one’s odds of becoming a slave and also one’s chances of escaping from slavery. Thus, in every ancient no-slaving zone, unfortunate people had the opportunity to sell children as slaves, governments found slavery to be an effective form of judicial punishment, and debtors could be compelled to sell themselves into slavery, although the prevalence of these methods naturally varied over time and place.9 8  But on the many ways to acquire slaves in traditional Africa, see Suzanne Miers and Igor Kopytoff, ‘African “Slavery” as an Institution of Marginality’, in Miers and Kopytoff (eds.), Slavery in Africa, 12–14. 9  For this phenomenon in general, see Orlando Patterson, Slavery and Social Death: A Comparative Study (Cambridge, Mass., 1982), esp. pt 2. On the evolution of debt slavery in

558

Fynn-Paul

This suggests that there were several ways in which most societies, from prehistoric times up to the present, have created and maintained slave populations. The most obvious way is through capture in war or by kidnapping; the second is purchase through a more or less organized market system; the third is by judicial enslavement for crimes or debts; the fourth is by voluntary sale of one’s self or one’s family into slavery; and a fifth way is through the legal enslavement of the offspring of slave women. It is important to notice that slave populations, once the institution was established, were always maintained by some combination of these factors.10 Until quite recently most historians wrote as though conquest was the primary means for ancient societies to acquire slaves, but this has now been discredited, and it is generally accepted that all five options were employed by most slave societies. Furthermore, it is now established that imperial societies created a demand for slaves in neighbouring areas.11 These areas which became the focus of imperial slaving efforts can be called ‘slaving zones’. Babylonia, see Muhammad A. Dandamaev, Slavery in Babylonia: From Nabopolassar to Alexander the Great (626–331 BC) (DeKalb, 1984), 157–80. For a brief mention of debt slavery amongst the ancient Celts, see H. D. Rankin, Celts and the Classical World (Portland, Oreg., 1987), 132. For the poor submitting themselves to slavery in fifth-century Gaul, see R. Samson, ‘Slavery, the Roman Legacy’, in John Drinkwater and Hugh Elton (eds.), Fifth-Century Gaul: A Crisis of Identity? (Cambridge, 1992), 223. For debt slavery and voluntary submission to slavery in high medieval Dalmatia, see Neven Budak, ‘Slavery in Late Medieval Dalmatia/ Croatia: Labour, Legal Status, Integration’, Mélanges de l’École française de Rome: Moyen Âge, cxii (2000), 748. For economic and judicial slavery in Viking Scandinavia, see Eljas Orrman, ‘Rural Conditions’, in Knut Helle (ed.), The Cambridge History of Scandinavia, i, Prehistory to 1520 (Cambridge, 2003), 308. On Tatar parents selling their children in the fourteenth century, see Peter Spufford, Power and Profit: The Merchant in Medieval Europe (New York, 2002), 340. 10  Orlando Patterson discourses on these paths to slavery in Slavery and Social Death, esp. 105; see also ibid., 147. There are of course many other variations. In some African cultures, certain children were considered supernaturally dangerous and would be left to die by their parents; these children, as well as naturally orphaned infants, would occasionally be brought up as slaves by those who found them: Miers and Kopytoff, ‘African “Slavery” as an Institution of Marginality’, 13. Duby notes that Carolingian society utilized a similar combination of means for maintaining slave supplies: Georges Duby, The Early Growth of the European Economy: Warriors and Peasants from the Seventh to the Twelfth Century (Ithaca, 1974), 32. 11  For a decisive argument against the assumption that Roman society relied on warfare as its primary means of replenishing slave supplies, see Keith Bradley, Slavery and Society at Rome (Cambridge, 1994), 31–56. It is now accepted that the Islamic Empire also used purchase as a primary means of gaining slaves, especially after the initial conquest was

EMPIRE, MONOTHEISM and SLAVERY

559

Because certain economic laws were at work in all the slave societies under discussion, the creation of slaving zones on the borders of no-slaving zones occurred according to a discernible pattern. A society which had begun to rely on large pools of slave labour, such as Egypt, Greece or Rome, naturally found that it had to replenish its slave capital, just as must be done with most other forms of capital. The demand for slaves was continuous and regular, and so slave markets were a routine feature of ancient societies, one whose trade routes naturally overlapped with routes for other commodities.12 As with any other trade good, supply lines began in regions where demand could be met by supply. Outside the boundaries of empire, in areas where other empires had not yet grown up, monopolies of violence were still maintained in small spheres by local chiefs. This pattern was universal, whether one speaks of Palestine in 2000 bc (on the fringes of Egypt and Sumer), or Gaul in 200 bc (on the fringes of Rome). Local rulers in these border regions continued to engage in frequent disputes with their many neighbours, and the fruits of successful disputes would include fresh supplies of slaves. People who remained in small states or tribal areas on the borders of empire thus remained especially vulnerable to enslavement, while those within imperial bounds were far less so.13 But a new and sinister force was at work with the rise of empire, since these border regions would now begin to feel the effects of an increasing economic demand for slaves within the empire. Mechanically speaking, prices for slaves within the empire would always be higher than prices for slaves within the completed. Slaves were bought from agents who travelled throughout central Asia for the purpose of acquiring them from local chiefs and raiders: Daniel Pipes, Slave Soldiers and Islam: The Genesis of a Military System (New Haven, 1981), 146–8. On a slave trade amongst Celtic tribes (and perhaps between these tribes and Rome) in pre-Roman Britain and Gaul, see Rankin, Celts and the Classical World, 132. On Gaulish slaves being traded for Roman wine, see Bradley, Slavery and Society at Rome, 36. 12  For slave market mechanisms in the Roman Empire, see Bradley, Slavery and Society at Rome, 41–3; for the early Islamic Empire, see Pipes, Slave Soldiers and Islam, 142–8. For similar mechanisms in the early modern era, see William D. Phillips Jr, Slavery from Roman Times to the Early Transatlantic Trade (Minneapolis, 1985), 141. For the overlap of slave and other trading networks in West Africa, see Philip Curtin, ‘Africa North of the Forest in the Early Islamic Age’, in Philip Curtin et al. (eds.), African History: From Earliest Times to Independence (London, 1995), 94. For the overlap of fur- and slave-trading networks in medieval Russia, see H. R. Ellis Davidson, The Viking Road to Byzantium (London, 1976), 99. 13  This argument is made by Dmitrij Mishin in ‘The Ṣaqāliba Slaves in the Aghlabid State’, Annual of Medieval Studies at the CEU, 1996–1997 (1998), 237.

560

Fynn-Paul

tribal economy, and the result was a net export of slaves.14 Socially speaking, this meant that merchants came looking to buy slaves, and local chiefs were encouraged to engage in wars and raids in order to boost their wealth and prestige. A new class of ‘slave raider’ was created as a side effect of empire. These raiders were armed brutes with a following of like-minded individuals, who lived from thievery and the proceeds of selling kidnapped victims to imperial merchants or their suppliers. Often, such raiders would be natives of a slaving zone who routinely sold their own countrymen into slavery. Such men are attested as recently as the nineteenth century in Africa, and their presence can be traced at least as far back as medieval times. A tenth-century description of the effects of slave demand on a slaving zone in the Sudan runs thus: ‘Merchants from Egypt come to this region … they steal children … they castrate them and import them to Europe, where they sell them. Among the Sudanese there are people who steal other people’s children to sell them to the merchants when they come’.15 As this excerpt attests, the preferred victims of slavers have generally been children.16 It seems clear that the disruption caused by imperial demand for slaves in these slaving zones could be very considerable, depending on the level of demand and the ability of local chieftains to monopolize violence within their territories. Prior to the intrusion of imperial demand, traditional patterns of warfare and enslavement would naturally have been maintained at sustainable levels. In areas where imperial demand was very high, however, these traditional patterns frequently became unsustainable, with devastating effects on entire regions, whose economic and political development might be severely retarded.17 We can state this as a formula: empires 14  Early evidence for the price differential has now been presented for the eighth and ninth centuries in McCormick, Origins of the European Economy, esp. 754–9. 15  Hadud al-Alam, quoted in Claude Meillassoux, The Anthropology of Slavery: The Womb of Iron and Gold, trans. Alide Dasnois (Chicago, 1991), 44. Where these slaves may have been headed is an interesting question. 16  For raiders preying on children in ninth-century Sicily, see McCormick, Origins of the European Economy, 246. Olaudah Equiano was famously captured along with his sister despite the fact that their parents had attempted to barricade them within the family’s walled compound as a precaution against slave-taking kidnappers: Olaudah Equiano, The Interesting Narrative of the Life of Olaudah Equiano (1789; New York, 2001), 32. 17  For an argument regarding the impact of slaving on political centralization in early modern Africa, see Philip Burnham, ‘Raiders and Traders in Adamawa: Slavery as a Regional System’, in James L. Watson (ed.), Asian and African Systems of Slavery (Oxford, 1980), 59–64. First-hand accounts of the devastation caused by imperial demand for slaves on slaving zones are comparatively rare, owing, amongst other things, to the fact that

EMPIRE, MONOTHEISM and SLAVERY

561

which rely on slave labour have a tendency to hinder the development of their slaving zones, at a rate proportional to their effective demand. To return to our imperial narrative, very gradually the techniques of statecraft spread from Egypt and Mesopotamia to the surrounding regions. The next states to evolve in the Greater Mediterranean–Persian Gulf region appeared on the borders of long-established empires: these include the Hittite state in Asia Minor, the Mycenaean states in Greece and the Hebrew state in Palestine. But it is important for the understanding of later developments that these states did not arise until more than 1,500 years after the neighbouring river valleys had evolved into empires. Amongst other things, that means that these areas remained actual or potential slaving zones for the neighbouring empires for very long periods of time. As the Phoenicians and then the Greeks spread out along the southern and northern Mediterranean in the first millennium bc, they extended the effective slave demand of the imperial centres of the eastern Mediterranean to the west and north. Thus the Etruscans, Iberians and Celts, who already practised slavery amongst themselves, came to feel the pressure of an increased demand for slaves, along with other commodities. Spain, Italy and France now became slaving zones for the more developed economies of the eastern Mediterranean. By about 500 bc, the techniques of civilization had spread around the Mediterranean to an extent that empires began to clash with one another more frequently. Thus we have an era of epic struggles between Persians and Greeks, and Romans and Carthaginians. The creation of multiple empires within the Greater Mediterranean region meant that, from now on, not all imperial frontiers were automatically available as slave zones. Slave traders naturally continued to concentrate their activities in areas where tribal states still predominated. Eventually, by the time of Augustus, the entire Mediterranean area was incorporated into a single empire under Rome. Demand for slaves was to a large extent met from distant areas at or beyond the frontiers of the empire.18

slaving zones have usually been pre-literate. Anglo-Saxon descriptions of Viking raids might count as slaving-zone destruction narratives, since it would seem as though Viking raids were in part motivated by Byzantine and Islamic demand for slaves. On devastation caused in parts of Africa by slave demand during the later nineteenth century, see Ronald Segal, Islam’s Black Slaves: A History of Africa’s Other Black Diaspora (New York, 2001), 157. For a sense of the insecurity experienced by all slaving-zone societies, see Equiano, Interesting Narrative, 32–8. 18  Bradley, Slavery and Society at Rome, 42–3.

562 III

Fynn-Paul

The Advent of Monotheism and the ‘Perfect’ No-Slaving Zone

The Christianization of the Roman Empire opened a new epoch in the history of slaving and no-slaving zones. Christian and Islamic monotheism created a new type of no-slaving zone which differed from any classical zone in so far as religio-ethical considerations now led to the effective prohibition of enslavement of any member of a given zone by any other member.19 This phenomenon could be termed a ‘perfect’ no-slaving zone, since, in theory, all members were protected from intra-zone slavery, though of course the reality was usually less than perfect. Several interrelated processes contributed to the evolution of the Latin Christian no-slaving zone, each of which should be dealt with in turn. The first is the development of the intellectual ethos which forbade enslavement of Christians by Christians. A second factor is the extent to which this ethos was applied in reality. A third consideration, which will not be discussed here in detail, is the growing separateness of Western and Eastern Christianity during the early medieval centuries, which caused the Latin and Greek spheres to split into somewhat distinct no-slaving zones. The fourth major influence, which affected all the above developments from the time of its genesis in the seventh and eighth centuries, was the rise of Islam. It is argued here that the rise of Islam played a key role in strengthening the ideological foundations of the Latin no-slaving zone. Finally, it remains to explain why Latin Europe actually moved beyond creating a perfect noslaving zone within its borders, and phased out slavery altogether in many regions by the eleventh century. It was one thing to forbid the enslavement of the dominant religious group, as well as protected minorities: this was done throughout the Dar al-Islam and Latin Christendom. But why did the Europeans cease to enslave even pagans, when this practice had rarely been forbidden by the Church? The absence of any large-scale slaveholding system in high medieval Europe is almost unique in the history of civilization, and our model suggests a new answer to the problem. It will be argued that the creation of two perfect no-slaving zones around the Mediterranean led to a 19  Note that the situation within the Dar al-Islam was somewhat complicated in so far as large numbers of Christians and Jews, who were usually protected as ‘people of the book’, were still liable to judicial slavery. The same went for Muslims living under Christian rule in the Middle Ages. For an analysis of the manifold meanings and uses of slavery in a multi-religious setting, see Mark Meyerson, ‘Slavery and the Social Order: Mudejars and Christians in the Kingdom of Valencia’, Medieval Encounters, i (1995), and its companion piece, ‘Slavery and Solidarity: Mudejars and Foreign Muslim Captives in the Kingdom of Valencia’, Medieval Encounters, ii (1996).

EMPIRE, MONOTHEISM and SLAVERY

563

unique and fierce competition between Christianity and Islam for the slave resources of the remaining slaving zones. Western Europe, with its less developed economy, was hopelessly outclassed in this economic war, and as a result the region became starved of slaves. This lack of external supply, in addition to ongoing internal processes discussed below, rendered the continuance of the European slave system untenable. Prior to the spread of the Christian world view, ancient society felt no overriding moral reason why one citizen might not become the slave or the master of a fellow citizen. Aristotle in his Politics famously justified the institution as indispensable to the political and economic well-being of the polis. More troublingly, he also attempted to justify slavery on moral grounds, arguing that some people, owing to a deficiency in the reasoning parts of their souls, could best fulfil their natural potential by serving a master as a slave.20 He even went so far as to suggest that, while the line between man and animal is usually firmly drawn, that between animal and slave is not. It is generally believed that race or nationality did not factor into Aristotle’s analysis: certain Greeks were evidently just as bestial in Aristotle’s eyes as certain members of any other society.21 Aristotle’s view, despite its obvious contradiction of reality (since, for example, many of the most intelligent ancient teachers were slaves), was seldom challenged by other ancient thinkers, and ancient elites showed little inclination to object to his teachings.22 From the very beginning of Christianity, however, there emerged a tension between that religion’s egalitarian ideals and the deeply embedded mores of the society into which it was born.23 The extent to which certain strains of Christian egalitarianism actually served to discourage or undermine ancient 20  Peter Garnsey, Ideas of Slavery from Aristotle to Augustine (Cambridge, 1996), 37–8, 108–10. 21  Ibid., 111. But see Benjamin Isaac, ‘Proto-Racism in Graeco-Roman Antiquity’, World Archaeology, xxxviii (2006); Benjamin Isaac, The Invention of Racism in Classical Antiquity (Princeton, 2004). 22  Garnsey argues that, contrary to what is often thought, the Stoics did not reject the idea of natural slavery: Garnsey, Ideas of Slavery from Aristotle to Augustine, 138–51. 23  There is still debate on how comfortable emergent Christianity was with the idea of slavery. Pierre Bonnassie helped to ignite the debate in his article ‘Survie et extinction du régime esclavagiste dans l’Occident du haut moyen âge (IVe–XIe s.)’, Cahiers de civilisation médiévale, xxviii (1985). Also crucial is Peter Brown, The Body and Society: Men, Women, and Sexual Renunciation in Early Christianity (New York, 1988), esp. 44–57, whose discussion of ideas of the body in the early Church touches on many aspects of slavery. Jennifer Glancy reminds us that Christianity was born into a world where slavery was entirely normative: Jennifer A. Glancy, Slavery in Early Christianity (Oxford, 2002), esp. 3, 156. Others have argued that the Christian ethos and slavery were more immediately at odds. See,

564

Fynn-Paul

slavery is debatable, although twentieth-century historians of late antique and medieval slavery have tended to assign Christian morality a minimal role. An oft-quoted but clearly fallacious argument suggests that centuries of collusion by the Church as an institution capture an essential ‘Christian’ thinking on the matter. Certainly it is true that very few Christian arguments against slavery, even the enslavement of Christians by Christians, can be found in the Christian writers of late antiquity.24 Also, it is true that slavery was not phased out in many parts of western Europe until about the year 1000, some six centuries after the Christianization of the empire.25 But it has more recently for example, Rosemary Morris, ‘Emancipation in Byzantium’, in Bush (ed.), Serfdom and Slavery, 130–1. 24  The principal exception seems to be Gregory of Nyssa: Garnsey, Ideas of Slavery from Aristotle to Augustine, 80–5. 25  Despite the knocking that many of Bonnassie’s points have taken since the publication of From Slavery to Feudalism, there is still little reason to doubt his conclusions on the broad chronology of medieval slavery: Pierre Bonnassie, From Slavery to Feudalism in South-Western Europe (Cambridge, 1991), 51–6. Even William Chester Jordan, who is highly critical of the book, does not appear to argue with Bonnassie’s chronology: see his review of Bonnassie in Slavery and Abolition, xiii (1992), 97–102. For the consensus that the transition to serfdom in the Frankish heartlands had occurred by the ninth century, see Alice Rio, ‘Freedom and Unfreedom in Early Medieval Francia: The Evidence of the Legal Formulae’, Past and Present, no. 193 (Nov. 2006), 7–11. Much of the new consensus relies on Wendy Davies, ‘On Servile Status in the Early Middle Ages’, in Bush (ed.), Serfdom and Slavery. Charles Verlinden long ago pointed out that slavery did persist during the otherwise slaveless period c.1000–1363, in the form of relatively small numbers of Muslim captives held in regions of Italy, southern France and eastern Spain: Charles Verlinden, L’Esclavage dans l’Europe médiévale, i, Péninsule Ibérique, France (Bruges, 1955); ii, Italie, colonies italiennes du Levant, Levant latin, empire byzantin (Ghent, 1977). This phenomenon has been examined for Barcelona in Stephen Bensch, ‘From Prizes of War to Domestic Merchandise: The Changing Face of Slavery in Catalonia and Aragon, 1000–1300’, Viator, xxv (1994). Susan Mosher Stuard has also found evidence of continuing domestic slavery throughout this period: Susan Mosher Stuard, ‘Ancillary Evidence for the Decline of Medieval Slavery’, Past and Present, no. 149 (Nov. 1995), 4–5; and see a reply by Jean-Pierre Devroey, ‘Men and Women in Early Medieval Serfdom: The NinthCentury North Frankish Evidence’, Past and Present, no. 166 (Feb. 2000). The institution of slavery persisted in England into the early twelfth century, where ecclesiastical institutions were some of the last holdouts: see David A. E. Pelteret, Slavery in Early Mediaeval England: From the Reign of Alfred until the Twelfth Century (Rochester, NY, 1995), 251–6. In Sweden, where the process of Christianization and abolition happened later and can be traced more easily in written sources, farm slavery disappeared first, but household slavery lingered until all slavery was banned by royal decree in 1335. See Joan Lind, ‘The Ending of Slavery in Sweden’, Scandinavian Studies, 1 (1978). Although slavery persisted in

EMPIRE, MONOTHEISM and SLAVERY

565

been contended that historiography has overlooked the ethical operations of Christianity on the slave system.26 The primary cause for this oversight is that for many decades the most influential historians of early medieval slavery were either Marxists or highly influenced by Marxist paradigms.27 Consequently, these historians concentrated on the end of slavery rather than on its actual evolution, because they were concerned to show how Marx’s concept of ‘feudalism’ replaced his concept of ‘ancient slavery’.28 These same historians have wanted to view slavery as an economic issue first and foremost, and this has made them prejudiced against social or intellectual factors.29 What is indisputable, however, is that between 500 and 1100, it became increasingly difficult for Christians to enslave fellow Christians within Latin Europe. This rule came to be enforced by a combination of church authorities and secular rulers acting with the approval of the Church, as we shall see.30 Nor many parts of Europe up to about the year 1100, it is thought that the great Carolingian estates underwent a process of transformation from slavery to serfdom beginning perhaps as early as the third century, and which was certainly complete by the tenth century. See Jean-Pierre Devroey, Économie rurale et société dans l’Europe franque (VI e–IX e siècles), 2 vols. (Paris, 2003); Chris Wickham, Framing the Early Middle Ages: Europe and the Mediterranean, 400–800 (Oxford, 2005). 26  Bonnassie was one of the first to see Christianization as a significant force in the decline of slavery in western Europe, and his view is still arguably not the majority opinion: Bonnassie, From Slavery to Feudalism, 37 ff. Devroey refines the argument when he writes: ‘Without fundamentally challenging the institution of slavery, Christianity made a significant contribution to an increase in the subjective rights of the unfree, by restoring to slaves their human condition, and by recognizing in canon law their capacity to contract a legitimate marriage’: Devroey, ‘Men and Women in Early Medieval Serfdom’, 8. Lind argues that the coming of Christianity led to an almost immediate cessation of slavery in Denmark and Norway, while being directly responsible for a more gradual erosion in Sweden: ‘Ending of Slavery in Sweden’, 60. 27  Although Bonnassie followed Marxist paradigms in much of his work, it is thought that his consideration of the influence of Christianity in the ending of early medieval slavery was distinctly post-Marxist. See Rio, ‘Freedom and Unfreedom in Early Medieval Francia’, 9. 28  McCormick makes this point in Origins of the European Economy, 734. See also Ross Samson, ‘The End of Early Medieval Slavery’, in Allen J. Frantzen and Douglas Moffat (eds.), The Work of Work: Servitude, Slavery, and Labor in Medieval England (Glasgow, 1994), 96. 29  Even the less Marxist Duby, for example, maintains that the Church dealt slavery ‘barely a glancing blow’: Duby, Early Growth of the European Economy, 32. 30  McCormick, Origins of the European Economy, 748–52. This is especially true of enslavement of Latin Christians by Latin Christians; the early medieval incidence of Byzantines trading in Latins and vice versa has not been studied enough to allow any meaningful

566

Fynn-Paul

does it matter much if it was the Church or the secular rulers who were the primary enforcers on this or that occasion; the essential fact is that enslavement of Christians by Christians gradually came to be considered taboo. Likewise, the fact that many prelates held slaves throughout the early medieval period is largely irrelevant to my main point, since my concern is to show that, over a period of centuries, the Latin Church’s increasingly anti-slavery ideology encouraged the abrogation of the slave system, and discouraged it from continuing or reappearing, even if the slave system largely broke down for the economic or political reasons suggested below.31 Owing to the sparseness of the evidence, exactly when and how this anti-slavery ethos was applied remains contentious, but it seems that many purely ideological strictures against slavery were observed from an early period. We know from Tacitus and other sources that the ancient Germans practised slavery amongst themselves, and there is no reason to doubt that some of the late empire’s demand was satisfied by slaves brought from Germania.32 We would therefore expect that, once they had conquered western Europe, German rulers would bring with them the custom of reducing captured enemies to slavery. And in fact we do know that pagan leaders showed no qualms about reducing their enemies, whether pagan or Christian, to slavery.33 Thus the famous appearance of Bede’s Angli on the sixth-century Roman slave market. It is well known that Merovingian and Carolingian armies enslaved generalizations. Warren Treadgold suggests in passing that by 1200 most Byzantine slaves were pagans from the non-Christianized parts of Russia: Warren Treadgold, A Concise History of Byzantium (New York, 2001), 188. 31  The argument that points to slave-holding churchmen as proof that ‘the Church’ was either pro-slavery or effectively silent on the issue of slavery is, while still widely held, a gross oversimplification. For a cautious treatment which allows for a great deal of church influence in the decline of slavery, see Marc Bloch, Slavery and Serfdom in the Middle Ages, trans. William R. Beer (Berkeley, 1975), 9–17. For a more recent dismissal of ecclesiastical influence on abolition, see Samson, ‘Slavery’, 223. 32  Bjørn Myhre, ‘The Iron Age’, in Helle (ed.), Cambridge History of Scandinavia, i, 73. Ammianus found the Alans remarkable for the fact that they did not practise slavery: Bradley, Slavery and Society at Rome, 19–20. Procopius records that the Slavs, for their part, were occasionally carrying large numbers of Byzantines into slavery by the sixth century: P. M. Barford, The Early Slavs (Ithaca, 2001), 58–9. The Franks were raiding and probably trading for slaves by the early seventh century: McCormick, Origins of the European Economy, 739. 33  Ellis Davidson, Viking Road to Byzantium, 99 ff., provides a good discussion of pagan Scandinavian slave raids, referencing Adam of Bremen’s Gesta Hammaburgensis and the Vita Rimberti. See also the older study by A. Mez, The Renaissance of Islam (London, 1937), 157 ff.

EMPIRE, MONOTHEISM and SLAVERY

567

conquered pagans as a matter of course. Under Charlemagne, the Frankish kingdom turned pagan Saxony into a very active slaving zone. It is therefore interesting to note that, by the eighth and ninth centuries, even Michael McCormick’s vast researches could unearth no evidence that Christian leaders of the western kingdoms reduced other Christians to slavery during time of war.34 Though earlier Christian rulers were less scrupulous than their descendants would become, it seems clear that it was Christianity itself, rather than any political force, which began to maintain most of western Europe as a no-slaving zone akin to what had existed under the empire.35 Thus, through the spread of Christianity, the majority of western Europeans were spared the resumption of pre-Roman slaving norms despite the empire’s fall. This is remarkable in so far as it was the first time in history that a philosophical system, rather than political force, maintained the integrity of a large-scale no-slaving zone. In practical terms, we can say that the development of the Christian noslaving zone led to the gradual extinction of most of the ancient channels through which citizens could become the slaves of fellow citizens.36 The significance of this development for the future slaving zones of Russia and Africa can hardly be underestimated, especially since a similar process was under way within the Dar al-Islam. Whereas in pagan Rome all the traditional operations of enslaving debtors, selling children into slavery, and judicial slavery were commonplace, by the eleventh century all these avenues to slavery had been shut down within Latin Christendom. Note that this occurred more quickly in

34  McCormick, Origins of the European Economy, 746–7. 35  See ibid., 747–52. The power of the Church’s teaching on marriage and sexuality should not be underestimated as a mechanism for reining in many of the excesses that traditionally accompany slave ownership. In the tenth century, for example, the king of the Bulgars could claim one in ten slaves brought to his country as a concubine. The Christianization of Bulgaria was accompanied by the demise of this system. Ellis Davidson, Viking Road to Byzantium, 66. 36  Bonnassie argues that Christianity slowly shut down the slave system, but, oddly, he writes about the various means of enslavement of Christians by Christians as though these processes did not diminish with increasing Christianization: Bonnassie, From Slavery to Feudalism, 35–6. Other evidence, however, suggests the pattern which probably occurred in most of Europe, that is, a gradual elimination of most of the traditional means of replenishing the slave supply. For this process in early medieval France, see Duby, Early Growth of the European Economy, 32–3, and in high medieval Scandinavia, Lind, ‘Ending of Slavery in Sweden’, 68. Lewis shows that a similar process happened under Islam: Lewis, Race and Slavery in the Middle East, 6–11.

568

Fynn-Paul

areas which did not border on slaving zones.37 Already by the sixth century in Italy, gradations of status between slaves and free people were beginning to appear. By the seventh century the Frankish Church had won acceptance of the slave’s right to marry. This was a considerable step towards the humanization of slaves. Later on, slaves won the ability to marry freeborn women, and the process of obliterating slave and free categories was well under way.38 Furthermore, church leaders became increasingly vocal about the virtues of manumitting slaves. It became generally accepted that manumission was a pious act that hastened one’s soul along the path to salvation, and the importance of this fact to early medieval rulers, the great patrons of Western monasticism, should not be underestimated.39 It should also be remembered that by the eleventh century, the growing strength of the reform papacy would have given episcopal and papal sanctions a coercive force throughout Latin society to a degree that had not previously been felt. The effects of the reform papacy on slaving issues has not to my knowledge been investigated in detail. Certainly Gregory VII conducted a strenuous anti-slavery campaign as part of his own reforms, which would set the stage for Urban IIs Crusade.40 It has been argued 37  As has been stated, in frontier societies such as Spain, however, Jews and Muslims living under Christian rulers could be reduced to slavery through judicial means, even though enslavement of Christians by Christians was now prohibited. See Meyerson, ‘Slavery and the Social Order’. In Dalmatia, a perfect no-slaving zone was not constituted until the early fourteenth century; this was then undermined by the post-Black Death resumption of a European slave-owning culture, when Dalmatia was partially incorporated into the Italians’ slaving zone: Budak, ‘Slavery in Late Medieval Dalmatia/Croatia’, 752–7. 38  Duby, Early Growth of the European Economy, 32–3. See also Devroey, ‘Men and Women in Early Medieval Serfdom’, 8. 39  For the conflict between protecting church property and the pious ideal of manumission in the Iberian Visigothic Church, see Carles Buenacasa Pérez, ‘Un exemple de la caritat cristiana a l’església primitiva: la manumisió dels esclaus a Hispània segons les fonts dels segles IV–VII’, Annals de l’Institut d’Estudis Gironins, xxxviii (1996–7). For a recent discussion of manumission in early medieval Europe, see Wickham, Framing the Early Middle Ages, 564–5. 40  Mohamed Talbi suggests that the decrees of Gregory VII, and his pressure on the Venetian doges, played a significant role in the cessation of the Latin slave trade with the Islamic Mediterranean. See Mohamed Talbi, ‘Law and Economy in Ifrīqiya (Tunisia) in the Third Islamic Century’, in A. L. Udovitch (ed.), The Islamic Middle East, 700–1900: Studies in Economic and Social History (Princeton, 1981), 214. Stephen Bensch describes two instances of high and late medieval popes rebuking Catalan kings and clergy over issues of slavery. In 1206, Innocent III rebuked the clergy of Barcelona for turning a blind eye while the city’s burghers refused baptism to their Saracen slaves. By this time the Church taught that sincere conversion placed a strong moral obligation on owners to manumit their

EMPIRE, MONOTHEISM and SLAVERY

569

that the Church’s anti-slavery campaign, which McCormick detects from the eighth century, increased during the ninth and tenth centuries until it became common to speak for the total abolition of any distinction between slaves and free men.41 That the Byzantine Church also considered slavery to be immoral, and manumission to be a step towards personal salvation, can be seen from emancipation documents dating back to the mid ninth century.42 Justinian himself had decreed that Christian marriage was tantamount to manumission, a ruling which set up an intractable centuries-long conflict between the Church’s teaching on marriage and the economic interests of slave-owners.43 While the influence of church teaching now appears more important than formerly in helping to discourage the continuance of European slavery, it is important to note that this teaching did not develop in a cultural vacuum. Nor is it likely that these anti-slavery sentiments would have produced many practical results, if economic and political factors had not conspired to bring these anti-slavery strains in Christian thought to fruition. The evolution and demise of slavery within western Europe has traditionally been treated as an endogenous problem, but the slaving-zone model proposed here, combined with Michael McCormick’s new researches on Mediterranean travel in the Dark Ages, enables us to suggest that exogenous factors were of paramount importance. Most important of all was the rise of Islam, which occurred in the seventh and eighth centuries. During this time, the entire southern half of the former Roman Empire, which had been Christianized for some three centuries, was co-opted into a new monotheistic empire whose borders stretched from Spain to the Indus valley. The Muslims, modelling themselves on the exclusionist monotheistic beliefs of the Jews and Christians, soon actively discouraged the reduction of co-religionists, as well as those ‘people of slaves, at least at the time of an owner’s decease: Bensch, ‘From Prizes of War to Domestic Merchandise’, 83. In another instance, in 1337 Pope Benedict XII rebuked King Pere III for the impropriety of having Saracen slaves serve at table: ibid., 67. 41  Bonnassie, From Slavery to Feudalism, 54. 42  Morris, ‘Emancipation in Byzantium’, 133–4. 43  As Charles Brand has pointed out, Justinian upheld the classical notion that slaves, as things, were not permitted to marry. But he also turned this law on its head, decreeing that anyone who had received the Christian sacrament of marriage was now legally free. This law was upheld until the time of Alexius I, who decreed that slaves could marry without altering their status. It seems as though Justinian’s law had generally had the effect of denying marriage to slaves, while Alexius felt that all Christians, of whatever status, should be eligible for the matrimonial state. Charles Brand, ‘Slave Women in the Legislation of Alexius I’, Byzantinische Forschungen: internationale Zeitschrift für Byzantinistik, xxiii (1996), 21.

570

Fynn-Paul

the book’ who lived within Muslim territories, into slavery.44 At the same time, for reasons that have not been satisfactorily explained, Islamic rulers began to rely very heavily on the use of slaves for the maintenance of their states, especially from the beginning of the ninth century.45 These were used in great numbers to staff Islamic bureaucracies, to fill military levies and to staff the large households of the wealthy.46 For several centuries, the lands of the Islamic Empire were far more economically developed than the territories of western Europe, and so the same forces which had operated on all peri-imperial zones since the days of Egypt and Sumer now created a demand for slaves within Europe that was resisted only with difficulty. During the eighth to the tenth centuries, then, we can say that Europe itself became a slaving zone of the new and powerful Islamic Empire, together with Africa and Russia. As is usual in slaving systems, those areas of Europe where states were least developed now proved the most susceptible to 44  The argument presented here suggests that Islamic societies might have enforced the taboo against enslaving co-religionists before this became a common ideal within Christian lands. 45  Patricia Crone has made the most influential case for the advent of the slave armies in particular, arguing that the system was implemented by al-Ma‌ʾmun (813–33) and greatly expanded by his successor al-Muʿtasim (833–42). It was soon widely imitated. See Patricia Crone, Slaves on Horses: The Evolution of the Islamic Polity (Cambridge, 1980), 74–81. Matthew S. Gordon, in The Breaking of a Thousand Swords (Albany, 2001), narrativizes most of what is known about the importation of the first Turkish soldiers to Samarra and Baghdad under the caliphs al-Ma‌ʾmun and al-Muʿtasim. Daniel Pipes suggested that Muslim citizens began to turn away from civil institutions for religious reasons, necessitating the recruitment of slaves: Pipes, Slave Soldiers and Islam, 70–5. Domestic slavery, and the concomitant restriction of women’s freedoms, seems to have become fashionable from the time of al-Rashid (786–809): Albert Hourani, A History of the Arab Peoples (Cambridge, Mass., 1991), 32–7. 46  As is well known, the notion of ‘slavery’ in Islamic societies encompassed people of all social classes; provincial governors might technically be slaves or freedmen, while many slaves were agricultural drudges. Women slaves could range from politically powerful courtiers to the poorest maids or prostitutes. For an overview of the types and conditions of slavery in Islamic countries, see Murray Gordon, Slavery in the Arab World (New York, 1989), 48–78. For the unusual case of the Zanj, black land-clearing slaves who revolted and successfully waged a guerrilla war in ninth-century Iraq, see Alexandre Popovic, The Revolt of African Slaves in Iraq in the 3rd/9th Century, trans. Leon King (Princeton, 1999). For the large-scale presence of agricultural slaves in Iraq and North Africa, and an indication that the historiography concentrates perhaps too much on the exclusively domestic nature of Islamic slavery, see Michael G. Morony, ‘Landholding in Seventh-Century Iraq: Late Sasanian and Early Islamic Patterns’, in Udovitch (ed.), Islamic Middle East, 165; Talbi, ‘Law and Economy in Ifrīqiya (Tunisia)’, 221.

EMPIRE, MONOTHEISM and SLAVERY

571

the new imperial demands. Beginning in the eighth century, slaves from pagan Britain, Scandinavia and Germany most frequently found their way into the markets of the caliph.47 It has now been shown that the Venetians took advantage of their position as Latin Christians with Byzantine political ties to sell Latins into slavery.48 This trade was resisted only with extraordinary effort on the part of Latin rulers. Thus, from the eighth century, the prohibition of the sale of Christians to pagans became one of the main preoccupations of ecclesiastical legislators.49 It is probable that the rise of Islamic demand for non-Islamic slaves effected a counter-reaction which spurred Christian leaders to enforce greater strictures against the sale of Christians by Christians. In the light of the strengthening Christian resolve, the Venetians opposed moves to Christianize the pagan Slavs during the ninth century, since they knew this would effectively remove their most lucrative source of slaves.50 Lopez notes that many Italian contracts contain slave transactions up to the end of the tenth century, but the number decreases significantly after that time. So the early Venetian slavers’ fears were indeed realized, and the Venetian economy was forced to find new, socially healthier avenues to explore.51 In France, Bonnassie sees an increasing tendency for Frankish slaves to escape their bonds from the early eighth century, inaugurating a trend which by the tenth century had developed into what he calls ‘a general attack on the very concept of servitude by the slaves themselves’.52 The slaves’ new boldness may well be a secondary effect of the Church’s own anti-slavery campaign, which it now can be argued was greatly intensified as a reaction to increased Islamic demand. The eighth and ninth centuries were thus even more difficult for western Europeans than has been realized, since the new demand for non-Islamic slaves encouraged non-Christians on all sides of Latin Christendom, including Slavs, Arabs and Scandinavians, to raid the lands of their Christian neighbours in search of valuable human plunder. Italian and other Christian merchants 47  See map in McCormick, Origins of the European Economy, 762; Mishin, ‘Ṣaqāliba Slaves’, 237–8. Note that McCormick finds a pattern of slave imports to the Caliphate arising prior to the reign of al-Ma‌ʾmun (813–33). 48  The Venetian slave trade is summarized in Michael McCormick, ‘New Light on the “Dark Ages”: How the Slave Trade Fuelled the Carolingian Economy’, Past and Present, no. 177 (Nov. 2002), 46–51. See also McCormick, Origins of the European Economy, 759–77. 49  McCormick, Origins of the European Economy, 748–52. 50  Ibid., 754. 51   Medieval Trade in the Mediterranean World: Illustrative Documents, ed. Robert S. Lopez and Irving W. Raymond (New York, 2001), 115. 52  Bonnassie, From Slavery to Feudalism, 48–51.

572

Fynn-Paul

seem likewise to have played a major role in incorporating western Europe into the Islamic slaving zone, just as native merchants in all slaving zones had been doing for millennia and would continue to do for centuries to come. If the influence of Christianity had not maintained a large no-slaving zone throughout the Frankish kingdom, the Frankish nobles themselves would have become slaving agents of their own people on a much larger scale than they were, with predictable results. As it was, our model suggests that the Caliphate’s effective demand for slaves was one of the major causes of Viking and other pagan raids on Europe during the Dark Ages.53 The process is similar to what the Germans would have experienced at the hands of the Roman Empire’s slaving agents, only the ‘perfect’ quality of the Islamic no-slave zone meant that demand on their slaving zones was increased proportionally. During these centuries, Europeans in many parts of the continent lived in continual fear of the slave raider; those who were captured, even in the far north of the continent, would end up in the southern lands of the caliph. The life of the ninth-century Greek saint Elias of Sicily, quoted by McCormick, paints a story of children who could never stray beyond the walls of their city for fear of slave raiders. After

53  McCormick was one of the first to assert the central importance of the slave trade in eighth- and ninth-century Europe, and to show its main directions. The suggestion that the Viking raids were spurred by slave demand from the Caliphate has not, to my knowledge, been made openly, although many authors hint at this conclusion. That the Vikings were active slave takers and slave traders is wholly established: see Orrman, ‘Rural Conditions’, 308. H. R. Ellis Davidson long ago acknowledged that slaves were a central part of Viking trading with Constantinople: Ellis Davidson, Viking Road to Byzantium, 99–100. Peter Sawyer notes that Baltic slaves were being exported to the Caliphate from about 790. He also suggests that the increasing sophistication of the Russian state probably led to the decrease in Viking slaving activity from the eleventh century: Peter Sawyer, ‘The Viking Expansion’, in Helle (ed.), Cambridge History of Scandinavia, i, 115–17. Thomas Noonan writes: ‘Between ca. 800 and 1020, millions of Islamic silver coins … were imported into European Russia from the Islamic world. A significant part of these dirhams were then re-exported from Russia to the lands around the Baltic Sea and especially Sweden … the overwhelming majority were the result of an active trade involving the Vikings and Russia’: Thomas Noonan, ‘The Vikings and Russia’, in Ross Samson (ed.), Social Approaches to Viking Studies (Glasgow, 1991), 202. For Noonan’s explanation of why the Viking-Islamic trade began in the eighth century, see Thomas S. Noonan, The Islamic World, Russia and the Vikings, 750–900: The Numismatic Evidence (Aldershot, 1998), esp. ch. 2. Note that the period of peak trade, which according to Noonan began about 840, corresponds with al-Muʿtasim’s expansion of al-Ma‌ʾmun’s mamlūk (slave soldier) programme, and its subsequent adoption by most Islamic rulers. See Crone, Slaves on Horses, 76–85. See also Michael Cooperson, Al Ma‌ʾmun (Oxford, 2005), 107–9.

EMPIRE, MONOTHEISM and SLAVERY

573

experiencing several nightmares whose theme was capture by slavers, the future saint was indeed taken in a raid and carried off to slavery in North Africa.54 By the tenth and eleventh centuries, however, both Latin Europe and Byzantium had managed to resist assimilation into the Islamic slaving zone, and both Christian regions succeeded in creating and maintaining slaving zones of their own. A crucial event was the seizure of Crete in 965 by the Byzantine emperor Nicephorus Phocas, which proved a turning point in Arab naval domination of the Mediterranean. By the eleventh century, decrees by the Venetian doges and Gregory VII marked the end of unfettered slave trading between Latin Europe and the Islamic Mediterranean.55 In sum, we can say that three distinct processes cemented the formation of a perfect no-slaving zone through most of Latin Christendom by 1100. First, the Latin Church in conjunction with secular rulers developed and gradually enforced a prohibition against reducing fellow Christians to slavery, especially by capture. It seems as though this prohibition had a definite and increasing effect on Christianized barbarians from the earliest migration period, and added greatly to the security of Christians within their own no-slaving zone, while it increased slave demand on pagans who lived outside the zone. Secondly, the states of western Europe eventually became powerful enough to maintain integral boundaries against all their neighbours, including Muslims, Slavs and Scandinavians. Islamic raiders now found it largely unprofitable to raid for Christian slaves. As Christian and Islamic states began to enter into relatively stable, long-term political and economic relationships, ransoming became much easier to arrange, and so slave raiding between Christian and Muslim lands became less a means of procuring slaves, and more a means of producing short-term war captives. Even this type of raiding gradually became more dangerous because of increased naval activity and a growing likelihood of diplomatic retaliation.56 At the same time, Latin rulers gained control over their own merchants, to the extent that they could enforce prohibitions on slave trading from within their own borders. As a result of these factors, the Islamic demand for slaves remained most effective in the non-Christianized lands east of the Christian territories, those whose less organized states and 54  McCormick, Origins of the European Economy, 245–8. 55  Talbi, ‘Law and Economy in Ifrīqiya (Tunisia)’, 224. 56  This tendency is already noted by McCormick for the ninth century: McCormick, Origins of the European Economy, 769. See also Mishin, ‘Ṣaqāliba Slaves’, 237. For a study on the redemption of Spanish Christian captives taken by Muslims in the later Middle Ages, see Andrés Díaz Borrás, El miedo al Mediterráneo: la caridad popular valenciana y la redención de cautivos bajo poder musulmán, 1323–1539 (Barcelona, 2001).

574

Fynn-Paul

lack of monotheistic ethos still encouraged local elites to lead slave-raiding parties into neighbouring territories, and where merchants could sell slaves of almost any origin without fear of rebuke.57 Islamic society also bordered on sub-Saharan Africa, where Ethiopian Christians and various groups of animists further augmented available slave supplies. A final factor which contributed to the emergence of a perfect no-slaving zone by the high Middle Ages was the ongoing conversion of formerly pagan peoples. Britain, Scandinavia, Germany, central Europe, the Balkans and most of Russia were all added to the Christian no-slaving zone between the sixth and eleventh centuries, greatly enhancing the security of those within the core areas of the zone.58 This meant that Latin Christians, who had no access to sub-Saharan Africa during these centuries, had only the increasingly distant pagans of the Russian steppes from whom to draw their slave supply.59 IV

The End of Slavery in Latin Europe

So much for the evolution of a perfect European no-slaving zone, which was in force by 1100 or 1200 at the latest. Latin Christians were now prevented by a combination of taboo and legislation from enslaving other Latin Christians. In this they were, belatedly it seems, following the Muslim states, which had put the theory of a perfect no-slaving zone into practice earlier than the Christians. It remains to answer the rather distinct question of why slavery had died out altogether in the European heartland by the tenth and eleventh centuries. In other words, I have shown how Latin Christians generally ceased to enslave 57  Talbi notes a paucity of slaves entering Tunisia from the tenth century onward, in contrast with a superabundance of slave imports during the ninth century: Talbi, ‘Law and Economy in Ifrīqiya (Tunisia)’, 214–27. 58  The split between Latin and Orthodox Christendom led to an increased danger of enslavement of one type of Christian by another, but the danger to a Christian of enslavement by a Christian of the opposed Church was generally less than the danger of enslavement by Muslims or pagans, since diplomatic recourse could often be had, at least for wealthy or influential captives. For a ninth-century pope protesting against the sale by Byzantines of Latin Christians they had supposedly ‘liberated’, see McCormick, Origins of the European Economy, 745. On ‘Greek’ slaves in fourteenth-and fifteenth-century Spain, see Verlinden, L’Esclavage dans l’Europe médiévale, i, 321–30. 59  Marc Bloch long ago noted the receding frontier with paganism as a possible source for the decline of western European slavery: Bloch, Slavery and Serfdom in the Middle Ages, 28. However, he and those following him have greatly preferred economic explanations: Bonnassie, From Slavery to Feudalism, 37–8.

EMPIRE, MONOTHEISM and SLAVERY

575

fellow Latin Christians by this time, thus creating a perfect no-slaving zone. But why did the Latin Christians go a step further and more or less cease to enslave non-Christians soon thereafter? This is especially interesting when we remember that enslavement of non-Muslims continued to thrive in the perfect no-slaving zone created by the Islamic world. If the new monotheistic world views did not by their nature cause a phasing out of slavery, what they did ensure, on both sides of the Mediterranean, was that slave supplies could not now be maintained by many of the means which had traditionally been employed by pagan slave societies. Both Christianity and Islam held that manumission was a pious act, and although the historiography is hesitant to compare rates of manumission in pre- and post-Christianized Europe, it seems logical to conclude that they were higher in the sixth and seventh centuries than in pagan times.60 In both Christian and Islamic societies, different degrees of unfreedom were introduced, and these tended to hinder the intergenerational continuity of slave populations.61 Likewise, in both cultures, judicial slavery, debt slavery and the sale of children were severely curtailed by religiously inspired law and custom. As a result, both the Christian and Islamic monotheistic blocs (understanding Christendom to be somewhat divided between East and West) had to rely increasingly on only two of the mechanisms by which slave populations had been maintained through classical times, namely auto-reproduction, and the purchase or capture of slaves from outside the no-slaving zone. There is much disagreement on the extent to which various slave populations can auto-reproduce, even for slave societies as modern and well studied as American antebellum slavery. In general, it seems as though populations in which newly captured slaves predominate will tend towards a lower level of auto-reproduction, and older, more established populations will tend towards a higher level of auto-reproduction, but regional 60  Bradley argues that manumissions were not excessive under the empire, and that the practice of manumission was employed primarily as a means of control. He also notes the continuing collection of the manumission tax in the fourth century as an indication that there were still substantial numbers of slaves at this time: K. R. Bradley, Slaves and Masters in the Roman Empire: A Study in Social Control (Brussels, 1984), 81–112. For Jewish slave-owning and manumission under the Romans, see E. Leigh Gibson, The Jewish Manumission Inscriptions of the Bosporus Kingdom (Tübingen, 1999), esp. 56–94. Again, for a recent discussion of manumission in early medieval Europe, see Wickham, Framing the Early Middle Ages, 564–5. 61  For gradations of unfreedom in the Maliki school of Islamic thought (which was prevalent in North-West Africa and Spain), see Cristina de la Puente, ‘Entre la esclavitud y la libertad: consecuencias legales de la manumisión según el derecho mālikí’, Al-Qanṭara, xxi (2000).

576

Fynn-Paul

and temporal variances will change this equation greatly.62 Leaving aside for the moment the debate on whether or not various European slave populations could sustain their numbers at various places and times, we can state that, other things being equal, the external supply mechanism, which is usually very important to maintaining slave economies, was becoming increasingly unavailable to Latin Christians as the early Middle Ages progressed, while the Islamic Empire generally found external supplies to be adequate or even abundant. And, interestingly under these circumstances, Latin Christendom effectively phased out slavery altogether in favour of serfdom, while Islam went the opposite route, and expanded the role of slavery all the more. Islam began to exert a tremendous demand on its surrounding slave zones, beginning shortly after the year 800.63 This occurred because internal mechanisms for maintaining slave supply became ideologically unacceptable, at the same time that Islamic rulers adopted the customs of keeping slave armies, slave bureaucracies and large slave households. In fact, part of the reason why the caliphs adopted slave armies and bureaucracies was to cir62  For this view of auto-reproduction in the New World slave regions, see Herbert S. Klein, The Atlantic Slave Trade (Cambridge, 1999), 166–8. For pagan Rome, see Bradley, Slavery and Society at Rome, 31–56. Bradley maintains that, generally speaking, ancient slave systems relied on external sources of supply as a significant maintenance mechanism. 63  For the purposes of the present argument it does not much matter whether early Islamic slave-holding patterns were continuous with or fundamentally different from Sassanid or Byzantine patterns, since we are primarily concerned with the nature of slave supply and demand itself, and secondarily concerned to show that demand for slaves from non-Islamic areas probably increased from the late eighth and early ninth centuries. As noted above, scholars know that there was some continuity in agricultural slave-holding between pre-Islamic and Islamic periods, probably more than is usually acknowledged. See Morony, ‘Landholding in Seventh-Century Iraq’, 165. Mohamed Talbi quite unequivocally states that slaves played a major role in Tunisian agriculture in the ninth century: Talbi, ‘Law and Economy in Ifrīqiya (Tunisia)’, 215. Certainly there is precedent for the institution of harem slavery in Sassanid sources, even though it did not become a dominant social institution until the time of al-Rashid; and, anyway, the institution of the mamlūk from the ninth century constitutes an entirely novel system which can only have increased the slave demand of Islamic societies. The high prices commanded for slaves from the early ninth century, as recorded by McCormick, are in any event evidence of a high demand within the Caliphate. On the increasing importance of the harem and household slaves from early Abbasid times, see Hugh Kennedy, The Prophet and the Age of the Caliphates: The Islamic Near East from the Sixth to the Eleventh Century (London, 1986), 138–9. On increased numbers of slaves, see Claude Cahen, ‘Tribes, Cities and Social Organization’, in R. N. Frye (ed.), The Cambridge History of Iran, iv, The Period from the Arab Invasion to the Saljuqs (Cambridge, 1975), 328.

EMPIRE, MONOTHEISM and SLAVERY

577

cumvent the entrenched power of some great Arab noble families. This was an effective solution to the problem that would dog European monarchs for more than a millennium: how to maintain an effective bureaucracy without continually losing royal prerogatives to a hereditary nobility. But it is interesting to note that, had Islam not forbidden the enslavement of co-religionists, it is unlikely that the unique Islamic institution of the slave army would have come into being. By the time that this external Islamic demand was developing, many Carolingian estates had completed a transformation from a slave-based mode of production to a serf-based mode of production, though slavery did continue to exist in many parts of the continent.64 By the ninth century, and perhaps much earlier, it does seem as though the bulk of European servile populations were maintained by auto-reproduction, rather than replenishment through market mechanisms. But in the previous centuries, there had been many slave markets throughout Europe, and it is likely that Christianity itself had played a role in shutting these down.65 And we can state that by the eighth and ninth centuries, to whatever extent European landlords wished to continue purchasing slaves, they were discouraged from doing so by the extreme competition for external slave resources from the Islamic economy. Thus the new Christian attitudes, coupled with Islamic competition for external slave supplies, worked against any return to a slave-based economy in Europe, and hastened the demise of those sectors of the economy which still relied on replenishment of the slave supply. In his Origins of the European Economy, Michael McCormick is concerned with discussing the nature of the slave trade routes which ran through Europe and the Mediterranean, but it is beyond the purview of his already massive 64  See Wickham, Framing the Early Middle Ages, 562–3. In Chris Wickham, ‘The Other Transition: From the Ancient World to Feudalism’, Past and Present, no. 103 (May 1984), he in fact argues that the bulk of this transformation occurred in the third century, prior to Christianization. Just how many slaves remained in various parts of Europe in each century under scrutiny is still a thorny issue, in part because the vocabulary employed in the documents is often loosely applied. For the debate between A. H. M. Jones and others on the number of slaves in the late empire, see Wally Seccombe, A Millennium of Family Change: Feudalism to Capitalism in Northwestern Europe (London, 1992), 45–6. For a good synopsis of the state of the early medieval slavery debate, see Rio, ‘Freedom and Unfreedom in Early Medieval Francia’, 8–12. See also the conclusions of Devroey in ‘Men and Women in Early Medieval Serfdom’. 65  For the abundance of slave markets in early medieval Europe, see McCormick, Origins of the European Economy, 748. For the role that Christianity might have played in suppressing these markets, see Devroey, ‘Men and Women in Early Medieval Serfdom’, 7–8, 15–19.

578

Fynn-Paul

study to dwell on the reasons for their rise and decline. Our model suggests that McCormick’s data showing the differential between European and North African slave prices reveals a strong economic cause for the disappearance of slave markets in early medieval Europe.66 Between the eighth and tenth centuries, the price of slaves in North Africa was three to five times higher than the price of slaves in Europe. While economists might baulk at McCormick’s findings on the grounds that, in an integrated market, price differentials of this magnitude should not exist for very long, we need only to look at slave markets in eighteenth-century Africa and the Caribbean to find an analogous situation. At that time, differences in the productivity of labour between Africa and the Caribbean were so great that slave prices were highly disparate in the two regions, resulting in a net export of slaves to the Caribbean. It seems plausible, on these grounds, to suggest that the productivity of slave labour in Europe was substantially lower than that of North Africa during the eighth to the tenth centuries, thus resulting in the observed price disparity, and a net export of slaves. When we look more closely, McCormick’s series show that prices in the Caliphate spiked during the late eighth and early ninth centuries, just as Islamic society began to adopt large slave households and large-scale slave armies. If, as some authors have argued, the ninth century witnessed the bulk of the transition from slavery to serfdom in western Europe, the timing of these two developments might not be coincidental.67 Leaving aside questions of exact timing, a long-term price differential of this magnitude means that there would have been precious little incentive for any slave merchants to sell their wares within Europe between the late seventh and tenth centuries. Merchants who had gone to the trouble of purchasing slaves from war leaders in Scandinavia, central Europe or Britain would have found it well worth their while to dispose of their cargoes in Venice or other ports with ships bound for North Africa, where Africans would pay a far handsomer price than any

66  See table 25.1 and its explanation in McCormick, Origins of the European Economy, 756 (reprinted and described in his ‘New Light on the “Dark Ages”’, 44). Note also that the presumed impact of Islamic demand on Byzantium suggests a whole other avenue of research, even though McCormick’s data suggests that the price differential between Byzantium and the Caliphate was virtually non-existent by the tenth century at the latest. 67  Rio, ‘Freedom and Unfreedom in Early Medieval Francia’, 8. (But note, again, that the chronology of slavery differed in the various regions of Europe. Bonnassie, for example, found slavery in Catalonia up to the end of the tenth century: Bonnassie, From Slavery to Feudalism, 51–6.)

EMPIRE, MONOTHEISM and SLAVERY

579

European along the way.68 By the time the European economy had developed to the extent that it could compete with North African demand, the Europeans had learned how to do almost entirely without slaves.69 It has already been suggested that church proclamations against the enslavement of Christians by Christians were a response to increased Islamic demand. It is now possible to suggest that the Church’s increasingly vocal campaign against slavery was made practicable by the fact that the system was on its way out anyway, owing in large measure to the economic difficulties of maintaining slave supplies. V

The Latin Christian No-Slaving Zone c.1000–1500

For approximately three and a half centuries, that is, between roughly 1000 and 1363, Europe became the world’s largest society which did not, for all practical purposes, practise slavery.70 The researches of Charles Verlinden show that the numbers of slaves in thirteenth-century western Europe were very small indeed.71 Slaves were found only in a few border areas, and almost all of them were Muslims who had been captured in war. Many of these were later redeemed by relatives or institutions set up for this purpose.72 Another pattern 68  From the early modern period up to the nineteenth century, African slaves were sometimes marched over a thousand miles to reach coastal slave ports. Some slavers seem to have cared little if as many as four-fifths of their captives died on the way; the profits obtained at port made even losses of this magnitude economically sustainable. See below at n. 98. 69  Bensch notes that by 1300, Mediterranean markets had become far more integrated, and slave prices were comparable and relatively stable throughout the Mediterranean: Bensch, ‘From Prizes of War to Domestic Merchandise’, 78. 70  Peter Spufford writes similarly: Spufford, Power and Profit, 338. Some slaves were, however, still kept in various parts of Europe throughout the high Middle Ages. See the sources cited in n. 25 above. 71  Verlinden, L’Esclavage dans l’Europe médiévale, i. Neven Budak provides a good chronology of the disappearance of early medieval slavery in his study of Dalmatia. Slave sales became quite rare in Dalmatia by the twelfth century, and slavery was almost entirely replaced by indentured servitude during the period of superabundant labour at the beginning of the fourteenth century. A late medieval revival of slavery, in both Dalmatia and Italy, began only in 1363, as a direct result of the second wave of plague. Split made its first late medieval laws restricting the export of slaves in 1375, after the third wave of plague. See Budak, ‘Slavery in Late Medieval Dalmatia/ Croatia’, 750–7. For Italy, see Iris Origo, ‘The Domestic Enemy’, Speculum, xxx (1955), 324–5. 72  On the continual possibility that Saracen slaves would be ransomed, see Bensch, ‘From Prizes of War to Domestic Merchandise’, 72–4. Bensch notes that on the Catalan-Islamic

580

Fynn-Paul

can also be pointed out: generally speaking, slavery tended to disappear more quickly from Christian lands which did not border an active slaving zone, and less quickly from regions which bordered slaving zones. This explains the longer persistence of slavery in Spain, Dalmatia, Britain, Scandinavia and Russia, for example.73 But the extent to which slavery stayed moribund during the period from about 1000 to 1363 still requires more explanation than the historiography offers at present. Throughout the high Middle Ages, northern and Mediterranean Christians continually purchased cargoes in ports where pagan and Christian slaves were available for purchase. Why then did no Tatar slaves, for example, appear on Verlinden’s radar during these centuries, as they would appear so prominently after 1363? Why, in other words, did Latin Christian traders not bring cargoes of these slaves back to their home ports? It is especially noteworthy that the Genoese, who became the middlemen in the Black Sea slave trade between pagan Russia, Byzantium and the Islamic states from the early thirteenth century, did not bring back large numbers of slaves to their own homes. A few non-Muslim slaves were, in fact, imported: a Russian slave is mentioned in Genoa in 1275.74 But for the most part the keeping of domestic slaves remained the affectation of a few, and many cities, including Florence and Venice, actually put tight legal restrictions on the importation of slaves.75 Did the Europeans, under the influence of their ascendant Church, truly develop an aversion to slavery as an immoral institution, and did European fashion therefore frown upon the importation of foreign slaves of whatever religion? There might be some truth to this. Whatever the economic situation may have been, if it had been at all socially acceptable for Europeans to import slaves during this period, upper-class people would have seen to it that they had slaves. If we thus view the absence of slavery from much of Europe between about 1000 and 1363 as due to something of a moral aversion bolstered by the frontier, ransom procedures had been regularized by the mid twelfth century at the latest. By the thirteenth century, the ransom of Christian captives, often using captured Saracens as a medium of exchange, was the legally sanctioned business of the Mercedarian Order: ibid., 82. 73  See Richard C. Hoffman, ‘Outsiders by Birth and Blood: Racist Ideologies and Realities around the Periphery of Medieval Culture’, Studies in Medieval and Renaissance Hist., vi (1983), 14–20. This is also cited by Stuard in ‘Ancillary Evidence for the Decline in Medieval Slavery’, 15 n. 44, who observes the co-ordination between European frontiers and lingering slavery. 74  Phillips, Slavery from Roman Times to the Early Transatlantic Trade, 104. Bensch notes that, prior to the mid fourteenth century, Russian and Tatar slaves cannot be located in Catalonia with certainty: Bensch, ‘From Prizes of War to Domestic Merchandise’, 81. 75  Phillips, Slavery from Roman Times to the Early Transatlantic Trade, 104–6.

EMPIRE, MONOTHEISM and SLAVERY

581

Church, this must strengthen the ‘church morality’ case made above for the previous centuries. But just as I have argued that fashion and morality had strong economic forces on their side during the early medieval centuries, it is also the case that any European aversion to slavery during the high Middle Ages was strongly aided by the economic reality of dirt-cheap labour. It should be remembered that the period from 1000 to 1300 corresponds with a steep and continual increase in European population. Between 1066 and 1300, for example, the population of England had mushroomed from just over one million to more than four million people.76 Then, in 1348, the Black Death struck, carrying off perhaps a third of the population of Europe. Indeed, the severe labour shortage that afflicted Europe from the mid fourteenth century, especially after the second major wave of plague struck in 1362, caused some southern Europeans to employ domestic slaves in their households once again.77 Whatever scruples these southern merchants once felt about importing slaves now crumbled in the face of economic hardship. That Florence felt a keen labour shortage, specifically beginning with the second wave of plague, is indicated by the fact that the commune allowed the unrestricted importation of slaves from 1363.78 A few numbers will help us to realize the extent of the labour crisis. In the cities of Catalonia, it can be calculated that for every five domestics that a wealthy household could keep prior to the plague, that same household could afford to employ only a single domestic by 1380.79 If these figures are indicative, we should not be surprised at Boccaccio’s famous complaint about the cheek of the labouring classes during these decades.80 76  On the fourteenth-century figures, see Barbara F. Harvey, ‘Introduction: The “Crisis” of the Early Fourteenth Century’, in Bruce M. S. Campbell (ed.), Before the Black Death: Studies in the ‘Crisis’ of the Early Fourteenth Century (Manchester, 1991). 77  For the idea that a true labour shortage did not start until the plague of 1362–3, see A. R. Bridbury, ‘The Black Death’, Econ. Hist. Rev., 2nd ser., xxvi (1973). For evidence of the same trend in Catalonia, see Jeffrey Fynn-Paul, ‘Tartars in Spain: Renaissance Slavery in the Catalan City of Manresa, c.1408’, Jl Medieval Hist., xxxiv (2008). See also J. FynnPaul, ‘The Catalan City of Manresa in the Fourteenth and Fifteenth Centuries: A Political, Social, and Economic History’ (Univ. of Toronto Ph.D. thesis, 2005), 72–3. 78  Phillips, Slavery from Roman Times to the Early Transatlantic Trade, 105; Origo, ‘Domestic Enemy’, 324–5. 79  For interest rates on interpersonal loans in Catalonia, see Fynn-Paul, ‘Catalan City of Manresa’, 138–9. For a threefold decrease on state interest rates in Florence and Catalonia, see ibid., 149–50. 80  For servants’ demands for high wages, in return for little service, during the course of the Black Death, see Giovanni Boccaccio, Decameron, trans. G. H. McWilliam

582

Fynn-Paul

It must be said, however, that only in southern Europe did domestic slavery take off from the 1360s.81 That northern merchants and their customers did not begin to import slaves at this time is testament to the effectiveness of the Christian no-slaving zone in restricting the influence of slaving culture and slaving economies. And in the south, although the increase in slavery was dramatic, it represents a dramatic increase over virtually nothing. During this phase of ‘Renaissance slavery’, which lasted into the eighteenth century, southern European slave-owning did not move too far beyond the circles of wealthier households in international trading centres. And the number of slaves in any given city was never very high. In Venice, a census of 1563 listed 7 to 8 per cent of the population as servants, and slaves were only a portion of this figure.82 In the town of Manresa in Catalonia, outside Barcelona, slaves made up about 1 per cent of the population in 1408, and only about 2.5 per cent of Manresan households owned slaves. Despite the fact that slaves had become an economically sensible investment, they were an investment whose high capital outlay barred all but the wealthiest householders. The average Manresan slave-owning household was worth about £1,350 in the money of Barcelona, which was over thirteen times the worth of an average household.83 In the north, there is some indication that slave-owning remained so foreign that it was actively frowned upon. This was helped very much by the fact that by the later Middle Ages the active Russian slaving zone was relatively inaccessible to northern merchants. An English case from 1559 witnesses a cartwright who had bought a slave from a Russian market. After neighbours complained about the cartwright’s tendency to scourge his slave, the court record goes on to state: ‘it was resolved that England was too pure an air for slaves to breathe in’.84 (Harmondsworth, 1972), 54. For a discussion of the 1351 English Statute of Labourers, which suggests that the lower classes had stepped beyond their former social and economic bounds, see Joseph P. Byrne, Daily Life during the Black Death (Westport, 2006), 251. 81  Spufford, Power and Profit, 338. Phillips, in Slavery from Roman Times to the Early Transatlantic Trade, ch. 5, does not address any northern European slavery in the later Middle Ages. 82  For the English diarist John Evelyn’s too-brief description of a seventeenth-century slave market in Livorno, complete with ‘Turks, Mores, and other Nations’ (albeit we are left guessing as to who is selling whom), see The Diary of John Evelyn, ed. E. S. de Beer (London, 2006), 99. The end of ‘Renaissance slavery’ in southern Europe is a topic that stands very much in need of further research. For the Venetian figures, see Phillips, Slavery from Roman Times to the Early Transatlantic Trade, 106. 83  Fynn-Paul, ‘Tartars in Spain’, 354. 84  This is quoted in Richard Hellie, Slavery in Russia, 1450–1725 (Chicago, 1982), 22 n. 32. There is a continuing debate on why and how early modern northern Europeans seemed to

EMPIRE, MONOTHEISM and SLAVERY

VI

583

The Slaving Zones of the Monotheistic Bloc: Russia and Africa

Russia Considerations of space allow us to say only a little about the impact of the monotheistic empires on their slaving zones in Russia and Africa, but something should be said here so as to provide a test case for our model. Suffice to say that a comparative history of slavery in these two regions very much needs to be written. We begin with Russia, where the slave trade peaked some centuries earlier than it did in Africa. The conversion of most of northern and eastern Europe in the ninth to eleventh centuries led to a greatly changed geography of slavery. Whereas at the beginning of the tenth century the frontier between Christian and pagan lands extended in a line from just east of Venice through Prague to the mouth of the Elbe, by the eleventh century, missionaries from both the Roman and Orthodox churches had extended the frontier nearly a thousand miles further east. The Balkans had been almost entirely pagan in the mid ninth century, but these were converted to Orthodoxy beginning shortly thereafter. The greatest gains for Christianity in terms of sheer territorial expanse came with the conversion of Vladimir the Great of Russia shortly before 1000.85 A glance at the map will show why no more is heard of trans-European slaving routes after 1000: the Christian-pagan frontier was now on a line extending north from the Crimea. Accordingly, the most active slaving zone in Europe now shifted far to the east, and the Black Sea became the most convenient slaving terminus for the entire Mediterranean region. Thus it remained throughout the later medieval period. The invasions of the Mongols, which began with the Golden Horde in the mid thirteenth century and continued through the reign of Timur in the late fourteenth, ensured that the region north of the Black Sea remained turbulent throughout these centuries. The Mongols began to adopt Islam from about the mid thirteenth century, but the conversion of the various Mongol and Tatar peoples was not completed for some time thereafter, and the presence of pagan, Muslim and Christian peoples in close proximity made for a fecund slaving market. All the same, the continuing assimilation of formerly pagan peoples into Christian and Islamic states meant that, by the later sixteenth century, there were very few ‘pagans’ left to enslave in the Russian steppe. Furthermore, the capture of Constantinople by the Turks in 1453, and the fall of the last Genoese outpost at Kaffa on the Black Sea in eschew slavery on their own soil, while becoming some of the greatest architects of the Atlantic slave system. See Davis, ‘Looking at Slavery from Broader Perspectives’, 458. 85  Janet Martin, Medieval Russia, 980–1584 (Cambridge, 1995), 5–12.

584

Fynn-Paul

1474, effectively blocked European access to the remainder of the Russian slave market.86 After this time, the Turkish sultans encouraged their own merchants, including Greeks and Armenians, to man the Black Sea routes. Any slaves that could be captured from remaining pagan areas or Christian lands of eastern Europe after the 1450s were generally handled by Ottoman merchants, and almost all of them ended up in Islamic countries. The severing of Europe’s direct access to the Black Sea, along with continuing conversion to Christianity and Islam, meant that Tatar slaves were no longer cost-effective after the mid fifteenth century. It is well known that the Europeans began to explore the coast of Africa just as their direct access to eastern markets was being diminished by Turkish conquests. One of the many commodities which Europeans found when they reached the Niger river was a fresh supply of ‘pagan’ slaves who had been captured in war by rival chieftains; the native slave economy in many parts of sub-Saharan Africa had already been impacted by centuries of Islamic slave demand.87 These slave channels could now be used to fill the demand for domestic slaves which Europeans had developed in the post-Black Death era. The textbook date for the arrival of the first shipment of Africans in Europe is 1441; however, most of the slaves up to the late 1440s seem to have been Berbers, since the ethnically black regions of Africa were not reached by the Portuguese until 1448.88 Henceforth, most European domestic slaves were black. Meanwhile, Islamic merchants and raiders now monopolized access to Russian slave markets, although we must imagine that the Russian slave exports dwindled from the fifteenth century, with the rise of the Muscovite state and the more thorough Islamization of the formerly pagan peoples on its borders. Still, it was only with the Russian capture of the Khanate of the Crimea in 1783 that the last major source of slaving incursions into Russian territory was halted.89 As one further reflection on the evolution of the eastern European slaving zone, it is interesting to note that the practice of slavery, including that of Christians by Christians, continued in Russia right up to the early eighteenth 86  Spufford notes Pedro Tafur’s statement that ‘in this city [Kaffa] they sell more slaves, both male and female, than anywhere else in the world’: Spufford, Power and Profit, 340. For the closing of European access to the Black Sea markets, see Phillips, Slavery from Roman Times to the Early Transatlantic Trade, 106. 87  One of the first collections of essays which integrates the combined impact of Christian and Islamic slavery in West Africa from about 1450 right up to the twentieth century is Lovejoy (ed.), Slavery on the Frontiers of Islam. See the introductory essay by Lovejoy, ‘Slavery, the Bilād al-Sūdān and the Frontiers of the African Diaspora’. 88  Phillips, Slavery from Roman Times to the Early Transatlantic Trade, 138–9. 89  Lewis, Race and Slavery in the Middle East, 12.

EMPIRE, MONOTHEISM and SLAVERY

585

century.90 It could be argued that this was due in part to the Russians’ Orthodoxy, which was generally less anti-slavery than Latin Christendom throughout most of its history. Certainly the Orthodox Church placed fewer restraints on its monarchs than the Catholic hierarchy tended to place on theirs.91 The ‘neighbouring countries’ factor should also be taken into account: we have seen that in Latin Europe, at least, proximity to a slaving zone usually encouraged the persistence of slavery. Russia itself was bordered by many different slaving societies until the late nineteenth century; in fact, St Petersburg did not manage to suppress the central Asian slave systems which had been incorporated into the Russian Empire until a few decades before the 1917 revolution.92 Perhaps the famous Brenner debate on the long continuance of serfdom in eastern Europe could be productively revisited by investigators who took the history of slavery in Russia more fully into account.93 Africa Had European slave demand not expanded beyond the domestic slavery of the Renaissance, the number of slaves exported by Europeans from Africa would probably have remained relatively low. However, a combination of factors led to the re-employment of chattel slavery by Europeans, after a lapse of nearly five centuries. In the mid fifteenth century, it was discovered that sugar plantations could be profitably run on the Cape Verde Islands using slaves which had been imported from the nearby African coast. These plantations were never very large, and did not employ more than a few thousand slaves in total.94 It was the exogenous factor of the discovery of the New World, and the continuing development of its economy, which was to cause such a dramatic increase in Christian–African slave trading. But for this discovery, and the implementation of the Cape Verde sugar plantation system in the New World on an entirely unprecedented scale, the story of African slavery in the history of early modern Europe might have remained a marginal footnote. 90  Hellie, Slavery in Russia, 695–708. 91  A legacy of Byzantine Caesaropapism: see ibid., 3. 92  Ibid., 708–10. 93  T. H. Aston and C. H. E. Philpin (eds.), The Brenner Debate: Agrarian Class Structure and Economic Development in Pre-Industrial Europe (Cambridge, 1985). 94  Phillips notes that most of the roughly 1,000 slaves brought to the Cape Verde Islands from Africa in the early 1500s were en route to other markets: Phillips, Slavery from Roman Times to the Early Transatlantic Trade, 141. Philip Curtin estimates that a total of 25,000 African slaves were imported to the Atlantic islands between 1450 and 1600: Philip D. Curtin, The Atlantic Slave Trade: A Census (Madison, 1969), 268.

586

Fynn-Paul

Meanwhile, as slave sources in Russia and the Caucasus slowly dried up, the only slaving zone which remained to the Ottomans and other Muslim states was the pagan areas of Africa.95 The nineteenth century was to become the greatest era of African slave exports to Muslim polities, as the eighteenth century had been the era of greatest slave exports to European-controlled territories.96 But mushrooming technology ensured that the process of civilization speeded up dramatically during the nineteenth century, and it is likely that in one way or another most African peoples would soon have become able to resist incorporation into any slaving zones, since, as this article has shown, political organization has always been one key for any society to escape the snares of imperial slave demand, while the adoption of Christian or Islamic monotheism with their slave ethics has been another. As it turned out, European colonialists, particularly the British, were to have a large hand in the final extinction of the African slaving zone, because of their continuing competition with Islamic states for African labour resources.97 After their own government halted slave exportations in 1808, the British began to put diplomatic pressure on various Muslim states to cease slave imports. This worked in the British best interest, since slave-trading networks were severely disrupting the economy and public order of the emergent British African empire. The accounts of nineteenth-century observers, which describe the effects of slaving raids hundreds of miles inland from the nearest slaving ports, provide us with perhaps the only sustained portrait of the devastation which imperial centres had been inflicting on their slaving zones since the beginning of the imperial cycle about 3000 bc. One of the most notorious slave raiders of the nineteenth century was Tippu Tip. Ronald Segal quotes from the memoirs of H. M. Stanley, who describes the effects of one of Tip’s raids using the language of the concerned colonialist: ‘Every three or four miles we came in view of the black traces of the destroyers. The scarred stakes, poles of once populous settlements, scorched banana groves, and prostrate palms, all betokened ruthless ruin’. Another observer, a missionary, described his encounter with Tippu Tip’s slave caravan. He observed that slaves were chained by the neck in long files, women carried babies, and many slaves were cut by whiplashes. Some slaves, he said, were forced to journey in this fashion for over 1,200 miles. Those who could not keep up were left to die of starvation or were shot. He estimated 95  See Fisher, Slavery in the History of Muslim Black Africa, 18–32, 40–5. 96  Segal, Islam’s Black Slaves, 147–76. 97  For an economic historian’s take on the connection between British abolitionism and British expansionism, see David Eltis, Economic Growth and the Ending of the Transatlantic Slave Trade (Oxford, 1987).

EMPIRE, MONOTHEISM and SLAVERY

587

that perhaps four in five captives perished on these journeys.98 When they finally arrived at a trading centre, many slaves were on the brink of death from starvation, although other slavers saw fit to protect their commodities by providing them with adequate nutrition. All was left to the whim of the individual dealer. Had anti-slavery observers been present in the interior of eighteenthcentury West Africa, we would possess many similar accounts of the devastation wrought by slavers whose hub was the Christian-run ports of Ghana. We should imagine that similar scenes occurred in Spain and Gaul at the time of the Punic and Gallic wars, as when Suetonius casually describes Caesar as currying the favour of some Gallic chiefs by sending them newly captured slaves ‘by the thousands’.99 But it is important to note that the devastation caused by raiders was only one extreme of the slave-supply spectrum. Muslim slave imports were often conducted under far less harsh conditions. Ottoman slaves from the Caucasus were sometimes procured by an agreement with local village families, who consented to sell their children into slavery on the understanding that their male offspring, at least, might achieve high administrative posts within the Ottoman Empire. The cases of Tippu Tip and the Caucasus peasants thus represent endpoints on the spectrum of impact which slaving empires have had on their respective slaving zones since the genesis of the imperial system.

98  Segal, Islam’s Black Slaves, 157–9. 99  Suetonius, The Twelve Caesars, trans. Robert Graves, revised Michael Grant (New York, 2003), 13.

Medieval Slavery in a New Geopolitical Space Youval Rotman

The Provenance of Slaves

According to the legal definition of slavery in Byzantium, there were three ways to become a slave: first, by being born of a slave mother (slavery by birth); second, by selling oneself (which, by law, only persons older than twenty could do); or third, by being taken into captivity. The first two cases applied to the population within the empire, the third to slaves of foreign provenance. In addition to these three legal means for procuring slaves, there were the slave trade and the forced enslavement of a free person, which could be legal or illegal depending on the circumstances. Note that these five ways of acquiring slaves, which date from the late period of the Roman Empire, were still valid and applicable to the medieval Mediterranean world throughout the entire period under study. To grasp how this system operated, I take as my starting point the law, and in particular the institutions of the Roman Empire. Roman law determined that the free status of a child came from the mother.1 If she had the status of eleuthera during her pregnancy (even if she was later reduced to slavery), the child received the status of eugenēs.2 In the same Source: Rotman, Youval, “Medieval Slavery in a New Geopolitical Space,” in Youval Rotman (Jane Marie Todd transl.), Byzantine Slavery and the Mediterranean World, Cambridge, Mass.: Harvard University Press, 2009, 25–81. Copyright © 2009 by the President and Fellows of Harvard College. 1  Dig. 1.5.5.2; Proch. Nom. 34.5–7. 2  Proch. Nom. 34.5–7. There were several terms for designating a free man. But unlike the words relating to a slave, they defined different legal statuses, as indicated in the Inst. Just. (1.3), “De iure personarum,” and then in the Prochiros Nomas (34.3). There were no legal categories for douloi, as there were for eleutheroi: (1) eleutheros/a (liber/a); (2) apeleutheros/a (libertus/a); (3) eugenēs (ingenuus/a). The word apeleutheros referred to someone with the status of an emancipated slave (it is translated here as “freedman”), eugenēs to a man who was born free (“freeborn”). Nevertheless, eleutheros, a freeman in general, was also used for an emancipated slave. Legally, freeborn was a different status from freedman. In cases where no distinction was made between these two statuses, the law generally spoke of eleutheros, in the Novellae of Leo VI for example, when it was not necessary to specify whether one was freeborn. As a result, eleutheros, translated here as free, designated the status of an emancipated person falling within the status of a free person.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_022

Medieval Slavery In A New Geopolitical Space

589

way, the children of a slave mother and her master were slaves until they and their mother were freed,3 whereas the children of a doulos and his mistress were eleutheroi.4 The situation of free persons who sold themselves to acquire money or who were obliged to become slaves because of a debt had been known since antiquity. An important moment in the evolution of that matter was the Novella of Leo VI that declared the act of selling oneself no longer valid.5 The same law, however, permitted the voluntary enslavement of a person who wanted to marry a slave.6 The act of selling one’s own children can also be included here, since it too entailed reducing a eugenēs (in Latin, ingenuus) to slavery. Although children did not have a legal personality, third-century Roman law recognized their freeborn status.7 All the same, cases of parents selling, abandoning, or prostituting their children were not uncommon. Trade, though not included in the legal definition, remained the privileged method for circulating slaves, whether they originated in the empire or were imported. Of course, the alienation of a formerly free person who had been illegitimately enslaved was still illegal. Normally, however, in the case of an imported slave (as opposed to one born in the empire), it was less important to know whether he or she had been enslaved legitimately because, from the standpoint of Roman law, the imported slave’s original status was defined by the ius civile of another state. Within the category of forced reduction to slavery, Roman and Byzantine legislation distinguished between the forced enslavement of a freeborn person and the enslavement of a freedman (under certain conditions, it was the legal right of a master to reclaim a freedman as a slave). Byzantine legislation, which followed the prohibitions of Roman law, banned the capture of a free person to be sold as a slave.8 But that act was legally permitted if the state itself reduced the person to slavery. In Roman civilization, the reduction to slavery was traditionally a legal punishment. But the only authority that could impose it was the emperor, by virtue of his judicial power.

3  Nov. Just. 18.1; App. Eklo. 10.2. According to classical Roman law, the status of the child after birth was independent of its mother’s status. Changes in the law were introduced by Justinian. 4  C J 9.11; App. Eklo. 4.4. 5  Nov. Leo. 59. 6  Nov. Leo. 100. 7  W. W. Buckland, The Roman Law of Slavery, pp. 420–422; CTh 4.8.6. 8  C J 9.20.7; Ekloga 17.16; Proch. Nom. 39.5, 39.22.

590

Rotman

Among the illegitimate means were the abduction of children and the sale of one’s own freeborn children. These two scenarios preoccupied lawmakers from the sixth to the eleventh century.9 During the entire era concerned, the Byzantine Empire was in a constant state of war and not only in the border regions. Wars affected the civilian populations in several regions of the empire, bringing about significant geopolitical shifts. The situation was particularly critical in the sixth and seventh centuries, but it continued to exist later on. The Prochiros Nomos cites ancient legislation stipulating that “according to the law of war, the conquered belong to the conquerors.”10 As a result, the empire had an unlimited supply of slaves. The obverse of that phenomenon was that many Byzantines were conquered by the enemy, becoming its captives and then its slaves. For the Byzantine population living within the empire, the delimitation between slaves and free persons was clearly marked by law.11 Things were less clear, however, for Byzantines taken into captivity or for populations outside the empire. It would be helpful to examine how the delimitation between the free person and the slave came about in the most complex situation, namely, slavery by captivity.

Civilian Captives, Prisoners of War, and Christian Marriage

Between Slave and Captive In a world where the conquered belonged to the conquerors, the winners literally held the losers’ lives in their hands. It was altogether legitimate to kill the conquered, and if the conquerors decided to let their captives live they could dispose of them as they wished. For that reason, captives had since antiquity become the conquerors’ slaves. That rule also made war a legitimate means for procuring slaves. There is no dearth of examples from the Roman era. The practice continued on a large scale in the Middle Ages, but it was no longer suited to the new geopolitical map of the Mediterranean world. Two problems arose as a result. On the one hand, enslaving prisoners of war was the general rule among both the Arabs and the Byzantines. When the Byzantines could not defeat their enemy, they became victims of that custom. On the 9  Nov. Just. 134.7; 153; Nov. Post. Just., coll. 4, nov. 35. 10   Proch. Nom. 34.2–3. 11  For the moment I shall apply the term “Byzantine population” only to inhabitants born in Byzantium and shall not address the question of whether, for example, a slave imported from a foreign country belonged to that population.

Medieval Slavery In A New Geopolitical Space

591

other hand, the evolution of Christian marriage introduced a new definition of that institution and of the captive’s marital status. Within the context of the Roman legal system, these two factors were in contradiction. Christian marriage, indissoluble by its very nature (except in special cases), made maintaining the free status of both spouses obligatory, even if one was enslaved following a war. The emperors of the Syrian dynasty in the eighth century undertook to resolve that contradiction by changing the law; they introduced new definitions regarding the status of the freeman and of the slave. The expression “war captives” groups together warriors—particularly soldiers—and the conquered civilian population. In fact, throughout the era concerned, it is possible to distinguish four types of war captives in the sources: first, soldiers taken prisoner during a military operation; second, the residents of a city or region taken by the enemy after a military defeat; third, the population captured by the enemy during a raid; and fourth, persons kidnapped by pirates. Although there was a difference between a raid by the enemy and one conducted by pirates, both were criminal acts and therefore had to be distinguished from the others. The aim of these raids was neither military nor political; they were directed exclusively at taking captives and at making a profit from their sale. But is a raid not itself a military operation? The distinctions between these categories of captives are necessarily conventional. This is made clear by the Greek terminology used to refer to them. All sorts of captives were designated by a single term: aikhmalōtos (parallel to aikhmalōtizō, “to take into captivity”). The matter at hand was to determine whom the captives belonged to and, more precisely, who had the right to draw a profit from their market value. In other words, were “the conquerors” victorious soldiers, pirates, or the state? Closely linked to the subject of prisoners of war was the phenomenon of exchanging them (in Greek, allagia). Such an exchange by adversaries could occur during a truce or upon the signing of a peace treaty. This was an innovation in the Mediterranean world and developed especially during the ArabByzantine wars. Exchanges of prisoners, whether soldiers or civilians, played a key role at that time. Simultaneously, however, sources continued to cite cases of prisoners of war who were not held for exchange but were sold as slaves.12 12  On this question I rely on the following sources, cited in chronological order: A. A. Vasiliev, Byzance et les Arabes; M. Canard, “Deux épisodes des relations diplomatiques arabobyzantines au Xe siècle”; M. Canard, “Quelques ‘à-côtés’ de l’histoire des relations entre Byzance et les Arabes”; H. Kennedy, “Byzantine-Arab Diplomacy in the Near East from the Islamic Conquests to the Mid-eleventh Century”; R. Khouri, “’Ανεπίσημες ἀνταλλαγές, ἐξαγορές καί άττελενθερώο εις Βνζαντινών καί ’Αράβων αἰχμαλώτων”; A. Tibi, “Byzantine-Fatimid

592

Rotman

The practice of exchanging captives is linked to my primary concern: the evolution of their slave status. The development of the practice in the case of prisoners of war has been studied for the most part from the Arab perspective,13 which corresponds to a difference between Arab and Greek historiography. Despite the absence of references on the Byzantine side, however, Byzantium not only played an important role in developing the practice but in fact initiated it. Thus far I have spoken of the dichotomy between the conquerors and the conquered. That is, I have adopted an international vantage point from which to view the captives. From the Byzantine perspective, however, there were two types of captured slaves: Byzantines captured by the enemy and enemies captured by the Byzantines. (I shall not call the inhabitants of the Byzantine Empire “Greeks” because Greeks also lived in territories beyond the reach of Byzantine imperial power and because the Byzantine population was not exclusively Greek.) This is an important distinction, because it conditioned the practice of exchanging captives. Those taken by the Byzantines were legally defined as slaves. But free Byzantines who became slaves in enemy territory raised a real problem for the Byzantine legal system, one not raised in Roman law. During the classical and the late Roman era, a free person captured by the enemy and made a slave was recognized as such in Roman territories as well.14 As a captive, hence a slave in another state, that person no longer had the right to own property and thus lost his usucapio.15 Although his property did not Relations in the Reign of al-Mu’izz Li-Din Allah (R. 953–957 ad) as Reflected in Primary Arabic Sources”; D. Letsios, “Die Kriegsgefangenschaft nach Auffassung der Byzantiner”; C. Verlinden, “Guerre et traite comme sources de l’esclavage dans l’empire byzantin aux IXe et Xe siècles”; T. G. Kolias, “Kriegsgefangene, Sklavenhandel und die Privilegien des Soldaten: Die Aussage der Novelle von Ioannes Tzimisikes”; M. Campagnolo-Pothitou, “Les échanges de prisonniers entre Byzance et l’Islam aux IXe et Xe siècles”; A. KoliaDermitzaki, “Some Remarks on the Fate of Prisoners of War in Byzantium (Ninth-Tenth Centuries).” 13  Campagnolo-Pothitou, “Les échanges de prisonniers entre Byzance et l’Islam”; Khouri, “’Ανεπίσημες ἁνταλλαγές, ἐξαγορές καί ἁπελευθερώσεις Βυζαντινῶν καί ’ Αράβων αἰχμαλώτων.” 14   Inst. Just. 1.3. The status of captured Romans is addressed at length in Buckland, Roman Law of Slavery, pp. 291–317 (though not a recent work, a more extensive and exhaustive study of the Roman legal institution of slavery could hardly be hoped for; it remains irreplaceable). See also M. Kaser, Das römische Privatrecht, 1:290–291. 15  This was the case only for a Roman held captive in an enemy country. Romans captured by pirates, bandits, or during a civil war retained their free status (Buckland, Roman Law of Slavery, pp. 291–292). See also Dig. 28.1.13, “qui a latronibus capti sunt, cum liberi manent, possunt facere testamentum,” in contrast to Dig. 28.1.12, which refers to those

Medieval Slavery In A New Geopolitical Space

593

revert to the fisc, it could not be claimed by an heir: a slave, not having a legal personality, could have no heirs. The property therefore remained in abeyance (those of slave status also could not draw up wills).16 Because the captive was assigned the status of a slave, his familial status and his ownership rights were affected: he lost his potestas over his own children, and his marriage was dissolved.17 In short, all his rights that in the empire had been granted him as a person of free status were rendered void. Here Roman law, which—though no longer in force—continued to be cited throughout the Byzantine era, had the character of “universal law” in that it stipulated the same fate for Roman captives and for captives of the Romans. For this reason, Roman law attributed that norm not to ius civile—Roman civil law—but to ius gentium, a Roman definition of “the law of peoples.” The exchange of prisoners of war was very rarely mentioned in the Roman world.18 In fact, the definition that reduced the Roman captive to slave status in his native state posed a difficulty for the argument that his citizenship should be restored to him by public means. A good example is the story told by the historian Malchus regarding the strategos Heraclius, who was captured by the Goths. Emperor Zeno offered to ransom him from the Goths but asked Heraclius’s family to supply the ransom money; had someone other than a family member ransomed Heraclius, he could still have been considered a slave.19 captured in hostium potestate but nevertheless allows heirs to inherit from such Romans who die in captivity. 16  The property he had assigned to his own slave or to his son by actio de peculio before being taken captive did not fall into that category. In fact, the subject of the peculium belonging to the captive’s slave or son was an object of debate among third-century jurists (Buckland, Roman Law of Slavery, pp. 294–295). 17   Dig. 49.15.12.4; 15.8; Buckland, Roman Law of Slavery, pp. 295–298. 18  The Romans sometimes asked that their captured citizens be returned, however. See J. Kolendo, “Les Romains prisonniers de guerre des Barbares au Ier et au IIe siècle.” The only reference given by Dio Cassius is the treaty of 172. The return of Roman captives was requested from the Quadi (Dio Cassius 72.13). P. Ducrey indicates two other examples in the treaties the Romans concluded with their defeated enemies, especially with the kings of Macedonia and Syria, from whom they demanded the return of their prisoners and refugees. These treaties contain no reciprocity clause (P. Ducrey, Guerre et guerriers dans la Grèce antique, p. 242). In addition, the fate of these Romans returned by the enemy is unknown. 19  Malchus, frag. 6.2 (in Blockley, The Fragmentary Classicising Historians of the Later Roman Empire); frag. 4 (L. R. Cresci, ed., Malco di Filadelfia Frammenti, p. 77). Cresci writes that Heraclius was probably made Magister Militum per Thracias by Zeno (Cresci, ed., Malco di Filadelfia Frammenti, p. 176).

594

Rotman

As for captives who had returned to the Roman Empire, their status was complicated.20 According to classical and imperial Roman law, a captive who succeeded in escaping the enemy or who was liberated did not automatically recover his original status upon returning to his native country. All the same, through the institution of postliminium, he received a transitional status that allowed him to recover the ownership rights he had had before his captivity.21 Things were even more complex if the captive was ransomed by a third party (in Latin, redemptus ab hostibus). Historians have attempted to reconstruct a coherent theory of the different imperial regulations (CJ 8.50) and of Roman jurisprudence (Dig. 49.15). That is impossible, since, as Marco Melluso correctly argues, a complete theory of the subject never existed.22 The debate has to do with the status of the captive ransomed by a third party. That captive did not benefit de jure from the ius postliminii so long as he did not ransom himself from his purchaser—in other words, so long as he did not reimburse his purchase price. At the time, there was an ius pignoris that bound him to his buyer.23 But since he could not use his former ownership rights, he did not have the means to purchase himself.24 The only solution was for him to be ransomed by his family or friends, as is indicated in the case reported by Malchus. So long as the captive had not reimbursed his purchase, his status and his former property remained in abeyance. Third-century jurists spoke much more often of the status of the slave captured and ransomed by a third party.25 By ius postliminii, such a slave would be returned to his or her former master. But for that institution to enter into force, 20  Buckland, Roman Law of Slavery, pp. 304–317 and 597; E. Lévy, “Captivus redemptus”; L. Amirante, Prigionia di guerra: Riscatto e postliminium; R. Mentxaka, “Sobre la existencia de un ius pignoris del redentor sobre el cautivo redimido en el derecho romano clasico”; A. Maffi, Ricerche sur postliminium; M. F. Cursi, La struttura del ‘Postliminium’ nella repubblica et nel principato. See also Kaser, Das römische Privatrecht, 1:291 and 2:129–132. 21  Provided that the state applied the ius postliminii on his behalf. On this subject, historians often mention the fate of Hannibal’s captives in Cannae, whom Rome refused to ransom, as well as the case of Attilius Regulus. The postliminium was created, in fact, for Romans who fell outside the limes of the Roman ius civile and not specifically for Romans captured at war. As for captured slaves, in returning to the empire they once more became the slaves of their former masters. 22  M. Melluso, La schiavitù nell’età giustinianea, pp. 18–23. 23  See esp. Mentxaka, “Sobre la existencia de un ius pignoris del redentor”; and CJ 8.50.3. 24  Mentxaka, “Sobre la existencia de un ius pignoris del redentor,” p. 331; Dig. 49.15.12; Buckland, Roman Law of Slavery, pp. 304–317; Kaser, Das römische Privatrecht, 1:291; 2:129– 130; Dig. 49.15.5ff. 25  M. V. Sanna, Nuove ricerche in tema di postliminium e redemptio ab hostibus.

Medieval Slavery In A New Geopolitical Space

595

the purchaser had to be reimbursed. Jurists were much less concerned by the subject of Romans of free status who were captured and ransomed by a third party because there was no dispute over property, as there was in the case of the ransomed slave. It is true that the case of the ransomed slave aptly demonstrates that the ransoming of a person from the enemy was also recognized within the empire. As for captives of free status who returned to the empire as slaves, I believe there is another explanation for why Roman law did not automatically restore their free status. From the Roman point of view, defeat was shameful. Roman jurists say so explicitly, explaining why weapons, unlike horses, were not restored by ius postliminii.26 Traditionally, therefore, the Roman soldier was supposed to fight until victory or his death. Those who became captives of the enemy were believed to have preferred slavery to death.27 Hence a Roman captured by the enemy who returned to the empire as chattel did not attract a great deal of attention from lawmakers because his case did not raise any legal problems. Roman law allowed him to be restored to his former status provided that his purchaser was reimbursed. It was understood, of course, that such a course of action by the captive or his family was possible only if they had the financial means. For that reason, jurists allowed a captive in that situation to draw on an inheritance, which was considered a potential ransom provided by a member of his family.28 If the ransomed captive or his family did not have sufficient money, he became de facto the slave of his purchaser, who could legally alienate him.29 26   Dig. 49.15.2. All the same, Trajan returned the weapons captured by the Daci in 100 (Dio Cassius 68.9.3). During the war against the Germanic tribes in ad 9, many Romans were captured. Some were then redeemed by their families. That act, writes Dio, was permitted at the time provided that the persons so ransomed remained outside Italy (Dio Cassius, 56.22.4, following Kolendo, “Les Romains prisonniers de guerre des Barbares”). 27  See the analysis in P. Dockès, La libération médiévale, pp. 10–14. 28  CJ 8.50.15. 29  Roman law recognized the restoration of such a person’s free status by postliminium as soon as his ransom price was reimbursed. That status was then attributed to him retroactively from the moment he was captured. And yet, for any other scenario, there seems to have been a reluctance to impose different laws or Roman jurisprudence. The works that clearly explain the legal problematic in this regard are Maffi, Ricerche sur postliminium, part 4; Cursi, La struttura del ‘Postliminium’ nella repubblica et nel principato, chaps. 9– 10; and Mentxaka, “Sobre la existencia de un ius pignoris del redentor.” Maffi and Cursi show the legal evolution of the matter and critique earlier studies. Historians speak almost exclusively of the captive who had the money to ransom himself from his purchaser. That was in fact the only case that attracted the attention of Roman jurists. Mentxaka explains precisely that if the person did not have the means, he became de facto a slave.

596

Rotman

Redemption by individuals thus existed, but not the possibility of ransom by public or imperial authorities. The fact that Roman law of the classical and late Roman era never envisioned the ransoming of captives by the state clearly indicates that exchanges of prisoners of war were not customary. The traditional Roman point of view still had currency in the fifth and sixth centuries. But an evolution was occurring in the status of the captive. A constitution of Honorius (395–423) dating from 409 allowed the captive purchased by a third party to ransom himself from the latter.30 If he could not pay the ransom, the constitution fixed a period of five years of labor, after which he regained his free status.31 The sack of Rome by Alaric probably created the political circumstances surrounding the law.32 To apply that law, Honorius named not only the curials but also the Christians “nearby” (Christiani proximorum locorum); these may have been devout people in general or the local clergy. In addition, the redemption of captives as an act of Christian charity or even as a duty is mentioned during the same period by Ambrose of Milan, who establishes an opposition between Christians and Barbarian enemies.33 Ambrose speaks of private individuals ransoming Christians and also argues that the church should melt down its sacred gold vessels to ransom captives.34 The landed property that the church owned was not at issue because it was inalienable.

30   CTh 5.7.2 (CJ 8.50.20). See Amirante, Prigionia di guerra, pp. 253ff. 31  Note that the process was not considered manumissio, and the liberated captive regained his original status. 32  But Olympiodorus, frag. 7, indicates that, during the siege of Rome by Alaric, many slaves fled (“especially those of Barbarian origin”). 33  Ambrose, De officiis 15.70–71. For cases where the Vandals’ captives were redeemed by Western bishops, see J. Gaudemet, L’Église dans l’Empire romain (IV e–Ve siècles), p. 310 n. 6. 34  “As for these men, therefore, I preferred to return them to you free rather than keep the gold. That host of prisoners, that procession, gleams brighter than the beauty of goblets. It is for this purpose that the Redeemer’s gold ought to be put to use, that is, to redeem men in peril,” Nevertheless, “the Eucharistic goblet must not leave the church in that form, for fear that the sacred chalice service will be put to impious uses. That is why, within the church, sacred vessels that had not been consecrated were first sought out; they were then shattered and finally melted down, divided up into small pieces and distributed to the indigent (egentibus); they also served as ransom for prisoners” (Ambrose, De officiis 28.136–143, M. Testard translation) [my translation from the French—Trans.].

Medieval Slavery In A New Geopolitical Space

597

In fact, the state had also sometimes returned Roman prisoners in the fourth century, even prior to the rule of Ambrose. Julian the Apostate made a special request to the Alamans to return the Roman soldiers they had captured.35 The act of ransoming Christians taken into captivity was twice officially recognized by Justinian: Novella 120 authorizes the church to alienate otherwise inalienable lands to redeem captives,36 and Novella 131 names the bishop responsible for an act of ransom mentioned in the will of a person from his diocese.37 The practice is also attested in the will of Flavios Theodoros, an employee in the offices of the dux of Thebais.38 In that document of 567, the author stipulates that upon his death his parents’ house will be sold to ransom captives (eis anarrēsin aikhmalōtōn). Sixth-century historiography also contains an echo of that new provision. According to Procopius, in 540 the bishop of Sergiopolis ransomed the twelve thousand inhabitants of Sura captured by Khosrow.39 The Talmud of Babylonia reflects a parallel development in Jewish law, demonstrating that in the eastern provinces of the empire, especially Palestine, Syria, and Mesopotamia, the problem of captives was very real in the sixth century.40 To redeem captives, interpreters of the Talmud say, even a synagogue 35  Amm., 18,2; Zos. 3.4. Eunapius (frag. 19) attributes the initiative to Vadomar, king of the Alamans, who asked that his son, a hostage of the Romans, be returned. 36   Nov. Just. 120.9, originally promulgated for the church of Jerusalem. 37   Nov. Just. 131.11. 38   P. Cair. Masp., 3.67312. On the abduction of free persons in Egypt in the sixth century, see papyrus SB 3.6097. For other sources regarding the redemption of captives as an act of charity, see Amiraute, Prigionia di guerra, chaps. 9–10. 39  Proc., BP 2.5. Nevertheless, Procopius indicates that the initiative came from the Persian king and that the bishop Candidus rejected it on the pretext that he did not have enough money. Khosrow, in love with a captive Christian girl, whom he married, had the captives returned before receiving the total sum of two centenaria. Theophanes relates that Emperor Maurice refused all offers from the king of the Avars to ransom Byzantine prisoners (Theoph. A. M. 6092, pp. 279–280). As a result, the Byzantines became enraged against Maurice, who, according to Theophanes, had to repent because, not having ransomed the prisoners, “he sinned against God” (Theoph. A. M. 6092–6093, pp. 279–285). Note, however, that unlike Procopius, who lived in the sixth century, Theophanes was writing at a time when the redemption of prisoners of war was customary. For the ransom of the captives in Gaul during the same period, see W. Klingshirn, “Charity and Power: Caesarius of Arles and the Ransoming of Captives in Sub-Roman Gaul,” which identifies the bishop’s personal motives behind the act of redemption. 40   The Talmud of Babylonia (Baba Batra 1.3; Neusner, ed., 22A:8) makes the ransoming of captives (Pidyon Sbevuym) a “mitzvah raba” (great precept) for which public money, even that intended for the construction of a synagogue, could be used.

598

Rotman

can be sold, provided it has not yet been used. The similarity in this regard to Ambrose and especially to the Novellae of Justinian is noteworthy.41 Let me point out that the Talmud of Babylonia originated in Pumbeditha and Sura, where the ransoming of Christians is attested by Procopius. The parallel development, in Christianity and Judaism, of the custom of ransoming coreligionists also points to the emergence of a new sense of community, which would later be found in Islam as well. All these examples attest to the individual’s identification with a religious community that preserved the free status of its members. As for Byzantium, it came to identify itself as a Christian empire, defined as a community of Christians. This process also had consequences for marital status: in 536, a new Novella abolished as grounds for dissolving a marriage the difference in status that arises when one of the spouses is taken captive and enslaved.42 That legislation was part of Novella 22, whose subject was Christian marriage. The notion of the status of captive and that of marital status thus influenced each other.43 Once again, this was a case of a Byzantine taken into captivity by the enemy. The origin of the distinction between a captive taken by the Byzantines and a Byzantine taken by the enemy can be seen here. The former was defined as a slave, whereas the latter preserved his status as a free person because his religion coincided with that of his state. Christian residents had to be saved from the infidels. No longer was the world of the empire pitted against that of the Barbarians, but rather the world of the Christian empire was pitted against the infidels. As Ambrose said, the new “Barbarians” were the non-Christians. And it was the development of that distinction that changed the civil status of the Christian captive and brought into being prisoner-of-war exchanges. Justinian’s marriage legislation created a contradiction between the Roman captive’s marital and civil status. In decreeing that a marriage remained in force after a person was taken into captivity, the Roman legal system of the sixth century recognized the marriage of a person who had the civil status of a slave, as the Novella explicitly indicates. It is true that it applied only to persons 41  In The Manumission of Slaves in Early Christianity (chap. 4), J. A. Harrill shows that, between the first and third centuries, public cash boxes in synagogues and churches raised funds for the collective emancipation of private slaves. 42   Nov. Just. 22.7 (cf. Nov. Just. 117.11, which establishes an indeterminate waiting period for a spouse in captivity). In application of that method, the Novella voided the right of the husband to the dowry if he wanted to dissolve the marriage. Hence the dissolution of marriage was possible only for the person who gave up nuptial property rights. 43  Note that the law prohibiting the enslavement of a free person by the state (Nov. Just. 22.8) was part of the same Novella.

Medieval Slavery In A New Geopolitical Space

599

remaining outside the empire, but that contradiction necessarily had repercussions within the empire as well. The following example illustrates the paradox that Justinian’s legislative innovation introduced. A Byzantine captive was sold as a slave in enemy territories and reintroduced as chattel within the empire. According to the traditional system, at the time of his captivity he would have been deprived of his status as husband and as free person. But according to the new law of Justinian, there was a contradiction between his marital rights, preserved as those of a free person, and that of the law of purchase, according to which, having been purchased legally by another Byzantine, he was considered property and not a free person so long as he had not reimbursed his purchaser. In other words, though the change in marital status did not raise a problem while the captive remained abroad, once he returned to the empire as a slave, if he was not ransomed by his family or by the church, he found himself with two contradictory statuses: married, which belonged to the status of a free person, but de facto also a slave. That paradox was resolved by legislation in the eighth century that stipulated that a formerly free Byzantine taken into captivity and reintroduced into the empire as a slave regained his status as a free person; that legislation did not take the ius postliminii into account. But if that paradox truly existed until the eighth century, why had the legislation of Justinian not addressed the problem stemming from the innovation? War Captives and the Political Situation For a proper understanding of the origins of that evolution, it is important to examine the relationship between captivity and slavery before the ArabByzantine wars. Consider the late Roman Empire and its relations with the Persians, who had posed a major difficulty for Rome since the second century. The sources are very rich, especially the historiography, but they come solely from the Romano-Byzantine side. During the long period of war in the sixth to seventh centuries, a large portion of the empire’s population found itself in a region invaded by the enemy. The military expeditions pursued by the Persians greatly affected people’s lives. Inhabitants abandoned cities, and people of the Roman Empire were forced to live under foreign occupation much more often than before, especially in the eastern regions. The abduction of civilians as a military objective is rarely mentioned by the Roman sources, however.44 44  G. Greatrex, Rome and Persia at War, 502–532; E. Chrysos, “Some Aspects of Roman-Persian Legal Relations.”

600

Rotman

Samuel Lieu, who studied the fate of the Romans captured by the Persians, shows that the Persians used the conquered population to fulfill their demographic needs.45 Hence they established captives (soldiers or civilians) in colonies in the interior of the Sassanid Empire, far from the borders. Take as an example the case of Antioch, whose residents, captured by the Persian army, were driven to the middle of the Sassanid Empire to create a “new Antioch.”46 Similarly, the inhabitants of Theodosiopolis in Armenia were transferred to the Persian capital,47 and Khosrow threatened to empty Edessa of its inhabitants by taking them to his empire.48 Lieu studies only the fourth century, but the Sassanid authorities’ policy remained the same throughout late antiquity, as shown in Procopius, Agathias, Menander Protector, and the fifth-century historians Eunapius, Olympiodorus, Priscus, and Malchus. Exchanges of prisoners of war were never at issue. The question was not even raised in the peace treaties or in the diplomatic relations between the Romans and the Persians for which we have documentation.49 The texts refer either to massacres of prisoners, to their transfer within the Sassanid Empire for purposes of colonization, or to their imprisonment.50 In no case were the captives who had been taken by the Persians sold as slaves, unlike the captives of the Romans. In fact, the Persians considered their captives not enemies but residents of the recovered Persian provinces. Hence in 503, when the Persians conquered Amida, they allowed some of the residents to live and ultimately liberated them.51 By “Roman prisoners” I mean Romans who were captured by the enemy. But that is not the case for the empire’s historiographers: in Priscus, for example, the aikhmalōtoi romaioi were the prisoners taken by the Romans.52 Priscus mentions cases in which the return of prisoners was stipulated by peace treaties of the fifth century. But only the Romans had to return their prisoners, whom they had captured at the Balkan border.53

45  S. Lieu, “Captives, Refugees, and Exiles: A Study of Cross-Frontier Civilian Movements and Contacts between Rome and Persia from Valerian to Jovian.” 46  Proc., BP 2.14. 47  Sebeos, chap. 33 (112). 48  Proc., BP 2.26. 49  In works by fifth- and sixth-century historians and in Greatrex, Rome and Persia at War, 502–532; and Chrysos, “Some Aspects of Roman-Persian Legal Relations.” 50  Proc., BP 1.13; Theoph. A. M. 6080 (p. 261). 51  Proc., BP 1.7. 52  Priscus, frag. 2. 53  Ibid.

Medieval Slavery In A New Geopolitical Space

601

As for Persians captured by the Romans, they belonged to the Roman authorities. The emperor could make use of them, sell them, or return them to the Sassanid Empire in support of his diplomacy. Hence Tiberius II decided to free the Persian captives who were prisoners in Constantinople and send them to Khosrow.54 Sebeos and Theophanes attribute that act to Heraclius.55 Although the captives taken by the Persians were not reduced to slavery, they were considered slaves by the law of peoples as it was understood in Rome. It is clear that the supposedly universal character of that law turned out to be unilateral on the international scene of the late Roman era. Nevertheless, the norms of Roman law existing in the first to third centuries corresponded to the situation of the late Roman period as well. The case of a person of free status who was captured, sold, and returned to the empire as a slave was not common in sixth-century political reality. The act of redemption motivated by charity returned the captive to the empire; if he had had free status before his capture, he probably regained it at that time. The reduction to slavery of free persons, or rather the transition from the status of a free person to that of a slave, was unilaterally fixed by Roman law. That action thus remained in the hands of the Romano-Byzantine authorities, and that is why Novella 22 of Justinian raised no problems at the time it was promulgated. The paradox was introduced into the legal system at a later time, when the person who was a captive of the enemy retained the rights of a free person within the empire, even though he actually became a slave and returned to the empire as such. From a global viewpoint, however, the paradox emerged for two more general reasons. On the one hand, the legal system persisted in believing in a universal law according to which “the conquered belong to the conquerors”; on the other, changes took place within the empire in sociocultural structures, such as Christian marriage, which remained in force even for the captive. That aspect of Christian marriage no longer tallied with any so-called universalism. But can it really be said that, from the Roman point of view, Roman law of the early imperial period was universal? And why did Roman law claim to recognize a “law of peoples” if such a law was disadvantageous to Roman captives or to the empire’s population? The situation that would have allowed the reduction to slavery of Roman citizens56 was not common during the Roman era as a whole. To understand 54  Menander Protector, frag. 23.8–9. 55  Sebeos, chap. 38 (126); 29 (128); Theoph. A. M. 6114 (p. 308). 56  I use the word “citizen” to designate civilian residents of the empire who had reached legal age and enjoyed free status.

602

Rotman

why, consider Rome’s strategy toward its own enemies in the early imperial era but also in the republican period. Roman policy did not want to acknowledge permanent enemies; it gave them the choice between concluding a peace treaty or being definitively subjugated.57 Such subjugation, especially if it was the result of a war or the repression of a revolt, sometimes led to the enslavement of the conquered population. That strategy explains the “universal” Roman law on prisoners of war, which always gave the advantage to the victor. Clearly, Rome never imagined losing, and the Romans also did not tolerate rivals on the political scene. Within its universal perspective, the Roman law of peoples could serve only Roman interests because Rome never had the intention of sharing its world with a rival. The only real threat of slavery for the Roman population was piracy. That phenomenon, omnipresent in antiquity throughout the Mediterranean, greatly preoccupied the Romans from the first century bc to the first century ad. Roman law considered both enslavement by pirates and any act of abduction, whether of a free person or of a slave, illegal within the empire.58 And among the empire’s first military actions in this regard were Marcus Antonius’s failed attempt to eliminate sea pirates in 74–72 bc and Pompeius’s successful efforts in 70–67 bc.59 It was vis-à-vis the Persians that Roman ideology encountered difficulties, not only because the adversary remained unconquered but also because the wars with the Persians affected the civilian population of the empire. That situation, however, did not raise a problem in terms of the slave status of Roman captives because, once settled in the Sassanid Empire, they did not return. 57  As Buckland says, quoting Otto Karlova: “A war ends either by surrender of the enemy— deditio—or by a treaty of amicitia” (Buckland, The Roman Law of Slavery, p. 306). 58  Ibid., p. 292. The reason was the same in both cases: the abduction by pirates or bandits was considered an act of theft, whereas war (and hence capture during wartime) was a lawful act. For a definition of banditry in Roman law, see B. D. Shaw, “Bandits in the Roman Empire,” pp. 6–7 and 21–22. Shaw provides a social definition of the phenomenon and points to its presence in the Roman world. Piracy became a threat for the Roman state only in the first century bc (pp. 39–40). 59  In “Piracy under the Principate and the Ideology of Imperial Eradication,” D. Braund shows that piracy was not completely eliminated under the principate and that it was imperial ideology that led Roman historians to present it as such (see Res Gestae Divi Augusti 25, where Augustus claims to have personally rid the sea of pirates through a war that was actually a civil war against Sextus Pompeius). As Benjamin Isaac shows, acts of banditry were still present in Palestine, Arabia, and Syria in the first to third centuries (see also Priscus, frag. 4). The Roman authorities aimed to eliminate them (B. Isaac, “Bandits in Judaea and Arabia”). See also P. de Souza, Piracy in the Graeco-Roman World, chap. 6.

Medieval Slavery In A New Geopolitical Space

603

Although the preservation of captives’ marriages did not pose a difficulty in the sixth century, that was no longer the case in the seventh and eighth centuries, when the Arabs replaced the Sassanid Empire on the Byzantine geopolitical scene. Unlike what had happened among the Persians, the enslavement of captives was the general rule among the Arabs, as it was in Byzantium.60 These captives were not inhabitants of the vast regions that the Arabs had conquered in the seventh century but rather were Byzantine soldiers. In the eighth century, the Arabs’ strategy changed, and from that time on the civilian Byzantine population itself became a target of the Arab forces. The paradox introduced by the Novella of Justinian found expression in another universal law, that of commercial exchange. Some of the Byzantine captives reduced to slavery by the Arabs were reintroduced into Byzantium as chattel or as part of the tribute. Since they retained their marital status, they found themselves slaves married to free persons. That was the contradiction the Ekloga attempted to resolve in the eighth century, stipulating that such persons would be free by law but would have to pay their masters their own value—that is, their price as slaves—to ransom themselves. If the captives did not have the means to do so, the new masters would pay their wages until the captives could accumulate that sum.61 The constitution of Honorius was probably the source for that provision, though it is unknown whether it was ever applied to Byzantium.62 In any case, from the eighth century on, the forced reduction to slavery of a Byzantine, both in and outside the empire, was not recognized de jure, even though the laws of commerce were respected. Under no pretext did the law prohibit purchasing such slaves or render their acquisition null and void. The new law of the Ekloga had other consequences. It put an end to the trade in such slaves within the empire, because their sale was now prohibited. In other words, no trader would purchase a slave of Byzantine origin in Arab countries to be sold in Byzantium. Hence the legislative innovation introduced by the Ekloga indirectly modified the slave traffic itself. 60  Attested for the Saracens even in the sixth century; Proc., BP 1.17; Malalas, log. 18.35 (L. Dindorf, ed., p. 447). The Saracens later sold those captured during the repression of the Samaritan revolt in the Persian Empire and in India. 61   Ekloga 8.2: “If someone purchases from the enemies an [originally] free captive and establishes him in his oikos, and if the captive has the means to pay the value agreed upon between them, let him be released as free. But if he does not have the means, let his purchaser keep him as a wage earner until the agreedupon value is reached; the auditors will obviously fix what must be paid as a wage each year to the man thus purchased.” 62  Unlike the constitution of Honorius, which sets a period of five years of labor for the captive to ransom himself, the Ekloga fixes neither a time frame nor a price.

604

Rotman

For Byzantines captured by the Arabs, that legislation paved the way for the practice of exchanging prisoners of war in Byzantium. The Captives of the Arab-Byzantine Wars Maria Campagnolo-Pothitou has studied thirteen exchanges of Arab-Byzantine prisoners of war reported by al-Maqrīzī between 804–805 and 946.63 In addition, Arnold Toynbee provides a list of the exchanges between the two empires between 769 and 969, a list that Athina Kolia-Dermitzaki completes.64 The first exchange for which we possess evidence is that of 769 attested by Theophanes.65 This exchange took place under the reign of Constantine V, the emperor who promulgated the Ekloga with Leo III,66 which confirms the hypothesis that the development of that custom on the Byzantine side was linked to the change in the definition of a free person and of a slave. Further evidence is provided in the Leges militares (Military law), a collection of rules of military discipline that probably dates back to the era of the Ekloga.67 The second version of this collection, edited by E. Korzenszky, contains a paragraph dealing with booty: captives are not considered booty, and the strategos must keep them with him or take them to the emperor for a possible exchange of prisoners of war.68 This collection must therefore date to an era when the exchange of prisoners of war was beginning to be an

63  Campagnolo-Pothitou, “Les échanges de prisonniers entre Byzance et l’Islam.” The author chooses to base her argument on the writings of al-Maqrīzī (1363–1442), explaining that they conserve information lost from al-Tanbīh by al-Masʿūdī (893–956), which has come down to us with several lacunae as a result of the deterioration of the two known manuscripts (Campagnolo-Pothitou, “Les échanges de prisonniers entre Byzance et l’Islam,” pp. 10–11). See also the exchanges of prisoners between the Byzantines and the Arabs, reported by al-Masʿūdī and translated by M. Canard in Vasiliev, Byzance et les Arabes, 2, pt. 2:405–408. 64  A. J. Toynbee, Constantine Porphyrogenitus and His World, pp. 390–393; A. KoliaDermitzaki, “Some Remarks on the Fate of Prisoners of War,” pp. 614–620, and, for the exchanges in Sicily, pp. 603–604. 65  Theoph. A. M. 6261 (p. 444). 66  According to the editor, L. Burgmann, it was published in March 741 rather than in 727 (Ecloga. Das Gesetzbuch III. und Konstantinos’ V, pp. 10–12). N. Van der Wal and J. H. A. Lokin agree (Historiae Iuris Graeco-romani delineatio. Les sources du droit byzantin de 300 à 1453, p, 72). 67  The manuscript tradition links this collection to the Ekloga privata (the legislative manual of Leo III). 68   Leges militares (version B), p. 48.

Medieval Slavery In A New Geopolitical Space

605

established practice.69 The origin of that change lies both in the process by which Byzantine law initially became a Christian law and in the transformation of the geopolitical map of the Mediterranean in the eighth century. In general, the exchanges between Byzantium and the Arabs are reported much more regularly by Arab historiographers than by the Byzantines. But even among the Arabs no mention of exchanges can be found before the eighth century. That is even more surprising in historiographers such as alBalādhurī, who gives a great deal of information on the seventh and eighth centuries. Note, in this regard, the story in al-Masʿūdī of a proposal for exchange coming from Mu‘āwiya in the seventh century.70 The Arabic term used is fidā, which refers to the ransoming of captives and not specifically to exchange. The story ends with the return of the captive to Arab territory, but there is no mention of a corresponding return of Byzantine captives. Another case of redemption of Muslim prisoners captured by the Byzantines is that of 710–720 mentioned by al-Balādhurī.71 It is worth pondering whether these stories do not attest to a custom of public ransoming (that is, one organized by the public authorities) among the Arabs, which might have preceded a need for such a practice among the Byzantines. The ransoming of captives, or fidā, was carried out either through the payment of money or through the exchange of captives. The Koran makes provisions for the liberation and redemption of war captives. These were prisoners of the enemy who found themselves in captivity among the Arabs, The Koran counsels a moderate attitude, recommending that the captive’s life be spared.72

69  The collection contains rules dating back to the sixth century: P. Verri, Le leggi penali militari dell’impero bizanlino nell’alto medioevo, pp. 13ff.; Van der Wal and Lokin, Historiae Iuris Graeco-romani, pp. 73–74; ODB 3:1492, s. v, “Nomos Stratiotikos.” In any case, rule 48 in the Korzenszky edition (Laurentinian Codex 75.6), which is not found in the manuscript edited by W. Ashburner (Leges militares [version A]), must have been written after the promulgation of the Ekloga. 70  Al-Masʿūdī, Murūj ul-dbahab wa-ma‘adin al-jawhar 8.75–76 (The Meadows of Gold, p. 320), quoted in Campagnolo-Pothitou, “Les échanges de prisonniers entre Byzance et l’Islam,” pp. 22–23. 71  Al-Balādhurī, KitābFutūh al-Buldān (History of Muslim Conquests), p. 133 (pp. 112–113 in the new edition by F. Sezgin), cited in Campagnolo-Pothitou, “Les échanges de prisonniers entre Byzance et l’Islam,” p. 23. 72  By being liberated, ransomed, or enslaved (R. Guemara, “La libération et le rachat des captifs. Une lecture musulmane”); Campagnolo-Pothitou, “Les échanges de prisonniers entre Byzance et l’Islam,” pp. 8–10; Koran, “Muhammad,” sura 47:4–10: “When you meet the unbelievers in the battlefield strike off their heads and, when you have laid them low,

606

Rotman

It does not consider Muslim prisoners.73 Did the Koran not anticipate that Arabs might fall into the enemy’s hands? This is not known. In any case, it is clear that for the ransoming of prisoners of war, the starting point of Islamic law was the same as that of Byzantine law. In both cases, the practice was in a state of flux.74 Although the Koran made provisions for the ransoming of its prisoners by the enemy, such an act is not attested for the seventh century. As a result, it is clear that Byzantine prisoners were sold as slaves. As al-Masʿūdī says: “In the time of the Umayyads, there was no major, well-known ransoming of which we may make mention.”75 All the same, al-Masʿūdī states that private acts of ransoming did take place in the time of the Umayyads. Parallel to the situation in the Roman Empire, the private ransoming of Arab prisoners preceded public redemptions. It remains an open question whether public redemptions of Arab prisoners were necessary.76 Campagnolo-Pothitou has attempted a response, arguing that the military, especially the new Arab navy formed by Mu‘āwiya in the late seventh century, needed to supply itself with soldiers. The term fidā, in the sense of “redemption of Arab captives,” is used by the same eighth- to tenth-century historians who describe acts of public redemption in exchange for money or prisoners: once again, the eighth century was the starting point. Nevertheless, truces and peace treaties between the two camps did exist before 769, despite the long duration of the war between the Arabs and the Byzantines. In the seventh century, the Arabs conducted a holy war aimed at the total defeat of the Byzantines; the sieges of Constantinople in the seventh and bind your captives firmly. Then grant them their freedom or take a ransom ( fidā) from them.” 73  Campagnolo-Pothitou, “Les échanges de prisonniers entre Byzance et l’Islam,” pp. 8–10. The paragraph that speaks of the charity (birr) of liberating the slave does not mention the fidā (Koran, “The Cow,” sura 2:177) even though, in the French translation, Denise Mas son renders it as “ransom.” See also M. Khadduri, War and Peace in the Law of Islam, chap. 11, “Spoils of War.” 74  For a comparison, see Yvonne Friedman, Encounter between Enemies: Captivity and Ransom in the Latin Kingdom of Jerusalem. The crusaders were aware of their Arab adversaries’ custom of exchanging prisoners. 75  Al-Masʿūdī, Kitāb al-Tanbīh wa’l-ishrāf p. 188 (B. C. de Vaux translation, Le livre de l’avertissement et de la révision, p. 254.), cited in Campagnolo-Pothitou, “Les échanges de prisonniers entre Byzance et l’Islam,” p, 21. 76  In “La libération et le rachat des captifs,” Guemara bases his argument on the Koran and shows that there were no public exchanges stipulated by Islamic law.

Medieval Slavery In A New Geopolitical Space

607

eighth centuries are proof of that. But throughout the period, peace treaties were signed between the two states, and it was not always Byzantium that initiated them. I have already noted that prisoners of war had no place in these treaties. But all of them do mention slaves as part of the tribute offered by the state in the more disadvantageous situation, the side asking for the treaty.77 The number of slaves to be provided, like the number of horses and the amount of money, was usually fixed as an annual base figure.78 These treaties attest to the significant number of slaves who found themselves in the hands of the Byzantine emperor and of the caliph. By contrast, in all the peace treaties signed after 769, no mention is found of slaves as part of the tribute. This absence argues in favor of the hypothesis that the strategy toward captives and the definition of their civil status had changed. Were the slaves who had been provided as tribute in the previous treaties war captives? That is likely, but information on slaves as part of a tribute comes solely from Byzantine documentation, which does not mention the fate of these slaves (or captives). Hence, to introduce the custom of prisonerof-war exchanges, or even of public redemptions of war captives, Byzantine law had to change. The acknowledgment that captives retained their civil status as free persons was the first step necessary toward making that status correspond to their marital status.79 The political map of the Arab world changed in the eighth century when the Abbasids replaced the Umayyads. In particular, that change translated into the transfer of the capital to the new city of Baghdad. Although the wars on the Byzantine border did not end, they took a different form. After the lifting of the Arab siege of Constantinople in the winter of 717– 718 under the reign of Leo III, the Arabs seem to have given up the prospect of a complete conquest of Byzantium. In the eighth century, they no longer conducted wars but simply made occasional incursions to the interior of Asia Minor to weaken the geographical defenses of the Byzantines. To do so, they besieged cities of strategic importance. In addition, for the Arabs victory entailed sacking the city and capturing its population, which was then sold 77  Scholars sometimes confuse these slaves with ransomed prisoners of war. See N. Stratos’s interpretation of the treaties of 659 and 685 in Byzantium in the Seventh Century, 3:188 and n. 693, 4:48. 78  There are also treaties that give the total figure. Theoph. A. M. 6150 (p. 347); A. M. 6176 (p. 361); A. M. 6178 (p. 363). 79  Slaves belonging to a tribute are not found in the treaties between the Persians and the Byzantines, which tallies well with the fate of captives among the Persians.

608

Rotman

into slavery, as, for example, during the siege and capture of Heraclea.80 That strategy was common to both adversaries. This was no longer the same war as in the seventh and early eighth centuries. The failure of the Arab sieges of Constantinople in 677–678 and then in 717–718 marked a decisive turning point in favor of the Byzantines and consolidated their state. With the advent of the Abassids in the Arab world, the total defeat of the enemy and the holy war were no longer political objectives. The Abassids gave priority to gaining a foothold in the interior against the opposing forces there, to establishing themselves, and to strengthening their empire. That new situation influenced Arab-Byzantine relations as well. The form of war changed, and with it the way the infidel enemy was perceived. Relations were not merely adversarial but also material and intellectual.81 The Byzantine Perspective on Captives The case of deserters was an important consideration in defining the free person and the slave on the geopolitical scene. If deserters returned to the empire, their punishment was enslavement, a practice that the central authorities had legitimated since the eighth century.82 Within the context of Byzantium’s international relations between the fifth and seventh centuries, the question of refugees was elaborated even more fully than that of captives, which shows just how widespread the phenomenon was. Menander Protector provides detailed documentation on the peace treaty of 561–562 between the Roman Empire and the Sassanid Empire.83 Not only did no clause deal with the exchange of prisoners of war; such prisoners were not even mentioned.84 By contrast, the sixth clause concerned refugees. These were actually deserters (automoleō in Greek). The difference between a refugee and a deserter depends on the point of view of the two parties: here I shall use the more neutral term “refugee.” The provision in question gave permission to those who, during the war, had sought refuge with the enemy to return to their homeland. As for those who took refuge in peacetime, they would be returned to those “from whom they had escaped.” 80  According to al-Tabarī in 802 or 803 and al-Masʿūdī in about 806; M. Canard, “La prise d’Héraclée et les relations entre Hârûn al-Rashîd et l’empereur Nicéphore Ier.” 81  M. Canard (“Quelques ‘à-côtés’ de l’histoire des relations”) and H. Kennedy (“ByzantineArab Diplomacy in the Near East from the Islamic Conquests to the Mid-eleventh Century,” p. 135) show that such relations already existed in the seventh century. 82   Ekloga 8.4.2; Leges militares (version A), 22. 83  Menander Protector, frag. 6.1ff. 84  That was also true in the treaties of 574 (ibid., frag. 18.2–4) and 576 (frag. 20.2).

Medieval Slavery In A New Geopolitical Space

609

Roger Blockley explains that this clause expresses the concern of both sides to control the civilian population at a time when there was a dearth of residents in the rural areas. He argues that those who escaped in peacetime were slaves.85 The text is not explicit, however, and these may also have been civilians who found better living conditions in the enemy empire: those “from whom they escaped” could simply designate the empire from which they fled and not specifically “their masters.”86 In any case, the problem of refugees was a burning issue in the border regions. It arose repeatedly in many peace treaties not only between Byzantium and the Sassanid Empire but also in the Balkans, between Byzantium and the Avars, the Huns, and then the Slavs and Bulgars.87 The aim of the treaties governing the problem of refugees was to ensure their freedom when they returned to their homes even if they had done harm to their state.88 Within the context of Byzantine-Arab relations, this was also a religious battle, which influenced the fate of the deserters-refugees. The best-known example is the epic of Digenis Akritas, which dates from the ninth or tenth century. It concerns an emir (the father of Digenis Akritas), a dignitary in the Arab army who directed attacks against the civilian population of Asia Minor. Having deserted to the Byzantines, he embraced the Christian faith and married a young woman from an illustrious Byzantine family.89 This story attests to the role the Christian faith played in establishing “national” identity, especially in Byzantine-Arab relations: consider the moving letter from the emir’s mother, who implores him to return to his country and his religion.90 In this story, Christian faith goes hand in hand with Christian marriage, which is the motive for the emir’s desertion. The story is not unique but reveals Byzantine policy in the tenth century. In De cerimoniis aulae byzantinae, Constantine VII Porphyrogenitus attested to the demographic use, for military ends, of captive Arab prisoners belonging to the state.91 According to this account, captives could be emancipated if 85  R. C. Blockley, The History of Menander the Guardsman: Introductory Essay, Text, p. 257 n. 55. 86  Menander Protector, frag. 6.1. 87  Unlike the fate of Hun refugees: their king demanded them back from the Romans and then crucified them (Priscus, frag. 2). For other accords dealing with the Roman refugees, see Priscus, frag. 9–11, 14–15, and 41. 88  Menander Protector, frag. 6; Theoph. A. M. 6305 (pp. 497–498) for the Bulgar-Byzantine treaty of 812. 89   Digenis Akritas (Grottaferrata version), l.30ff. 90  Ibid., 2.50ff. The same theme is also found in the Arabic literature. 91  Const. Porph., De cerimon. 2.49 (p. 695). Cretan captives who declined to convert were tortured by Nicetas (Theoph. Cont. 5.61).

610

Rotman

they converted, married, and remained in Byzantine territories, especially in the border regions. Nevertheless, they were not delivered from their douleia to the Byzantine emperor, which now took the form of either fiscal or military obligations (which were also of a fiscal nature). The two acts, conversion to Christianity and marriage, not only transformed captives into free Byzantines (they may have already been married in their native country, but that changed nothing) but also made them trustworthy.92 According to the imperial edict mentioned in the Life of Athanasia of Aegina, also dating from the early tenth century, unmarried (widowed or single) Greek women were to marry foreigners who had settled in Byzantium.93 Arab prisoners may have been included among these foreigners.94 Al-Masʿūdī also attests to the presence, in 923, of twelve thousand Christianized Arab horsemen among the Byzantine forces.95 As Marius Canard has noted, parallels existed on the Arab side as well.96 While there is no mention of a conversion to Islam by Andronikos Doukas (906),97 other renegades, such as Leo of Tripoli and Damian, governor of Tarsus, attained high positions in Arab society, often thanks to military careers.98 It is interesting to note that the Arabs told stories of Byzantine and Arab renegades, whereas the Byzantines refrained from speaking of their own renegades.99 Hence, for example, Theophanes tells how Elpidios, the strategos of Sicily, took refuge in Africa, without mentioning his conversion to Islam.100 92  Note that, in the story of Digenis Akritas, the emir falls in love with a Christian woman (whom he had kidnapped) and converts to marry her. Here the romantic character of the narrative makes its appearance. 93   Life of Athanasia of Aegina 2. 94  A. M. Talbot, who comments on this narrative, adopts Kazhdan’s interpretation that the edict was promulgated by Theophilus in 832 (A. M. Talbot, Holy Women of Byzantium: Ten Saints’ Lives in English Translation, p. 143 n. 22). The edict is cited by Dölger (Dölger, Reg. 1, no. 422). 95  Al-Masʿūdī, Murūj al-dhahab wa-ma‘adin al-jawhar (2.60). Other cases, al-ready attested in the ninth century, are cited by Canard (“Quelques ‘à-côtés’ de l’histoire des relations,” p. 109), who does not necessarily see them as captives. 96  Canard, “Quelques ‘à-côtés’ de l’histoire des relations,” pp. 106–108; Canard, “Deux épisodes des relations diplomatiques arabo-byzantines au Xe siècle,” pp. 54ff; Vasiliev, Byzance et les Arabes, 1:196–197; Guemara, “La libération et le rachat des captifs.” 97  Vasiliev, Byzance et les Arabes, 2, pt. 1:181ff. 98   Theoph. Cont. 6.21; Leo of Tripoli is called Rašīq am Wardāmī in the Arabic sources. He was also nicknamed Gulām Zurāfā. Vasiliev believes that this nickname indicates that Leo was originally the slave of someone named Zurafa (Vasiliev, Byzance et les Arabes, 2, pt. 1:163 and n.2). For Damian the governor of Tarsus (Damyana in Arabic), see ibid., 2, pt. l:212ff. 99  Neither Byzantine nor Arab sources specify that Andronikus Doukas converted to Islam. 100  Theoph. A. M. 6274 (pp. 455–456).

Medieval Slavery In A New Geopolitical Space

611

By contrast, the theme treated in Digenis Akritas is also found in Arabic literature, for example, in the story told by Abū al-Faraj, Kitāb al-Aghānī (The Book of Songs)—a collection of songs from the time of Caliph Harun al-Rashid. It is the story of a reciter of the Koran who, after his voluntary exile, arrives in Constantinople, converts to Christianity, and marries a Christian. “The envoy of Caliph ʿUmar b. ʿAbd al-ʿAziz, coming to Constantinople to negotiate an exchange of prisoners and passing through the streets of the capital on a mule, hears a sad voice singing an Arabic verse. Moved by the beauty of the voice and intrigued by the strangeness of such a song in such a place, he enters the house from which the voice is coming. He sees the reciter, who is tirelessly singing the same verse, all the while weeping. He entreats him to return to Islam to be redeemed with the prisoners. ‘But I have a Christian wife and two children,’ he replies. ‘How can I return to Islam?’ and he rejects the proposal.”101 I mentioned three interrelated factors that determined the delimitation between slave and free in Byzantium: first, captivity; second, the shift from the status of free person to that of slave; and third, the institution of marriage (i.e., Christian marriage). These three criteria applied only to slavery by captivity. But they indicate that there is a difference between having the status of slave and being a slave or, in other words, between a slave de jure and a slave de facto. That difference, which later existed between a free person de jure and a free person de facto, can be observed in the first place in the case of captivity, because two different societies and two different legal systems met at that point. But the real catalyst in that case was the change in status of Byzantine marriage, which was related to a person’s Christian faith or Christian identity. Here as well the difference between the Byzantine society of that era and the surrounding world becomes apparent. Nevertheless, the empire was also evolving toward a Christian civilization, and that evolution was accompanied by an internal adaptation.102 I have emphasized the relation between the status of a free person and a Christian identity. Khosrow’s discourse cited by Menander Protector shows how the international and even political perspective of Byzantium had changed as a result of the Christian factor. The Sassanid king explained that the Romans could not conduct a war within the Sassanid Empire because 101  Abū al-Faraj al-Isbahānī, Kitāb al-Aghānī 5.175ff., as cited by Canard, “Quelques ‘à-côtés’ de l’histoire des relations,” pp. 106–107. 102  The roots of this evolution date back, of course, to Constantine I. For the course of action he took to change the institution of marriage, the standard reference is J. Evans Grubbs, Law and Family in Late Antiquity: The Emperor Constantine’s Marriage Legislation.

612

Rotman

they would find there a Christian population that they could not defeat.103 It is clear that Menander was reflecting the point of view of the Byzantines. And in fact when Belisarius captured the Persian city of Sisauranon in 541, he released all residents who had once been Christians and “Romans” (Khristianoi te kai Rōmaioi to anekathen) but sent the “Persians” to Constantinople.104 In this case, the Romans were perceived as Christians, whereas the Persians were non-Christians; in the sixth century, that was a uniquely Roman point of view. Despite the correlation established between the state and religion, the population in the territories under Arab domination was not entirely Muslim (on the contrary, Muslims were a minority in the seventh to ninth centuries), just as not all those subject to the Byzantine emperor were Christian. Hence the Byzantines also took. Christian captives from Islam countries, as attested by the exchange of captives in 855 or 856.105 Thus far, I have been interested in the Byzantine-Arab border, where the religious dichotomy coincided with the political dichotomy. But that equivalence did not correspond to the dichotomy between slave and free. In other words, though Byzantium made efforts to recover its soldiers and its residents captured and enslaved by the Arabs by defining them as free, and though the same thing happened on the Arab side,106 in Byzantium not all free persons were Christian, and not all Christians were free. The same was true of Muslims in the Arab world. In Byzantium, the Christianization of non-Christian slaves was the general rule, the master preferring to convert his slaves to his own religion. The laws dealing with the difference in religion between master and slave shed light on another case of correlation between the state and religion. Byzantine law forbade a non-Christian, whether Jewish, Samaritan, or pagan, to have Christian slaves.107 The ownership of Christians by non-Christians 103  Menander Protector, frag. 16.1. 104  Proc., BP 2.19. Note that Justinian then sent these captives to Italy to fight against the Goths. Hence prisoners had been put to military use at an earlier date. In another instance, Justinian II sent Slavic captives to the Arab border (Nikephoros, Patriarch of Constantinople, Short History 38). Conversion is not mentioned in either case. 105  The fourth exchange on al-Masʿūdī’s list, translated in Vasiliev, Byzance et les Arabes, 1:336–337; Campagnolo-Pothitou’s “Les échanges de prisonniers entre Byzance et l’Islam” (p, 17 n. 94) cites al-Maqrīzī on this matter. 106  Canard provides a list of special delegates sent by the Arab caliphs and the Byzantine emperors on a mission to exchange captives (Canard, “Quelques ‘à-côtés’ de l’histoire des relations”). 107   Nov. Just. 37.7, 144.2 (Nov. Post. Just., coll. 1, nov. 7). That legislative course of action dates back to the fourth century, where only Jewish masters and Christian slaves were at issue (Buckland, The Roman Law of Slavery, pp. 605–607).

Medieval Slavery In A New Geopolitical Space

613

continued to concern lawmakers, indicating that the situation was not rare.108 The legislation tried to foresee all possible situations: for example, a slave who wanted to convert to the Christian faith even though his master was opposed to it, or a master who forced his slaves to become Jewish or Samaritan. The objective of that legislation was to stand in the way of a Christian’s being subjected to a non-Christian. It was self-evident that Christians could have Christian slaves. The legislation was almost identical on the Arab side: laws forbade nonMuslims from having Muslim slaves.109 It is clear, however, that the Muslims themselves had Muslim slaves, but these were of foreign origin. Hence they were not Muslims by birth but were Islamized after being imported to the Arab world.110 The capture of prisoners of war was thus crucial since, apart from commercial importation, it was the only means for acquiring such slaves. But did not the practice of exchanging prisoners of war cause a problem by limiting that source of slaves? On the Byzantine-Arab border, which extended from Caucasus to southern Italy, the religious opposition was therefore clearly defined. It is time to see what was happening on the other active front of Byzantium, that is, in the Balkans. In Byzantine documentation, no case is found of an exchange of prisoners of war between the Byzantines and the Slavs, Bulgars, or Russians. It is true that diplomatic relations between Byzantium and the Balkans in the seventh to ninth centuries are much less well documented than those involving the Arabs, particularly since the Slavonic sources are much less rich than Arab historiography. That is not sufficient for concluding that no exchange took place between Byzantium and the Slavs or Bulgars. In fact, the inscription published by Veselin Beševliev, which concerns the Bulgar-Byzantine peace treaty of 816, mentions an exchange of prisoners (psukhin anti psukhis—“soul for soul”).111 The exchange took place between the Byzantines and the Arabs at a time when the Bulgars were still pagans. 108  These Novellae are repeated in the App. Eklo. 8.1–3, which broadens the prohibition to heretics. See also the eighth canon of Nicaca II (RP, Syntagma, 2:583). 109  According to the fiqh, Arab jurisprudence; Ency. of Islam, 1:24–26, s.v. “‘Abd” (R. Brunschvig). 110  The importation of Muslim slaves from Africa to the Arab world occurred after the era under study. See B. Lewis, Race and Slavery in the Middle East, pp. 11ff.; M. Gordon, Slavery in the Arab World, pp. 23ff. 111  V. Beševliev, Die protobulgarischen Inschriften, no. 41 (pp. 190ff). See Kolia-Dermitzaki, “Some Remarks on the Fate of Prisoners of War,” p. 607 n. 98.  For the dating of the treaty to 816, see W. Treadgold, “The Bulgars’ Treaty with the Byzantines in 816,” which also cites the Life of Nicetas of Medicium 45.

614

Rotman

I have emphasized that one of the conditions for the exchange of prisoners was that both parties treat the prisoners they captured in the same manner. Within the context of Arab-Byzantine relations, that meant that both parties, unlike the Persians, reduced their prisoners to slavery. As for the treatment of prisoners by the Slavs and Bulgars, the little extant documentation comes especially from the Byzantine side. These are saints’ Lives from the tenth century. The Life of Luke the Younger mentions the Bulgar attacks by Simeon.112 Blaise of Amorion is captured when he goes to Rome and is then sold to a “Scythian.”113 His Life also mentions pirates on the Danube.114 Finally, the Life of Fantinos the Younger tells the story of a slave abducted by the Bulgars and imprisoned.115 And it was not only enemies who abducted Byzantine residents. Hence, in the Life of Nikon Metanoeite, the peasants are threatened by bandits who pillage the countryside and kidnap free persons to sell as slaves.116 The saint rescues a kidnapped girl and returns her to her parents, and he makes the bandits swear to never again engage in pillaging.117 Similarly, Gregory the Decapolite runs into Slav bandits on the waterway while traveling to Thessalonica.118 The Byzantines sometimes asked for these captives to be returned through diplomatic means, but no exchange of captives is mentioned after that of 816.119 Each of the three fronts of Byzantium imposed a different policy. Since the Roman era, the Caucasian front had been the scene of several battles against the Eastern adversary (the Persians, then the Arabs). The Caucasian peoples did not themselves conduct an offensive policy against Byzantium. Recall that the alliances between the Khazars and Byzantium were strengthened by imperial marriage connections: in 733, Constantine V married the daughter of

112   Life of Luke the Younger 24, 32–34. 113   Life of Blaise of Amorion 8. 114  Ibid., 9. 115   Life of Fantinos the Younger 61. 116   Life of Nikon Metanoeite 57. 117  Ibid., 70. 118   Life of Gregory the Decapolite 10. 119  Kolia-Dermitzaki (“Some Remarks on the Fate of Prisoners of War,” pp. 606–612) describes the mission of Leōn Choirosphaktēs, sent to the Bulgars by Leo VI to ransom Byzantine prisoners. A great number of letters were sent from Leōn Choirosphaktēs to Simeon (see G. Kolias, Léon Choerosphactès, magistre, proconsul et patrice, pp. 1ff.). Michael McCormick, basing his argument on the Slavonic Lives of the brothers Constantine-Cyril and Methodius, shows that on two occasions the saints return Byzantines captured by the Russians and the Slavs (Origins of the European Economy: Communication and Commerce, A.D. 300–900, pp. 188, 190).

Medieval Slavery In A New Geopolitical Space

615

the Khazar khan.120 In addition, Armenia, though not Chalcedonian, had been Christian since the fourth century. The Balkans were thus notable for their belated paganism. The difference in the religion and laws of the non-Christian Balkan peoples, when compared to the Arabs and Byzantines, lay in the fact that Arabs and Byzantines developed a new practice requiring them to rescue and redeem those of “their own” who had fallen into captivity. The religious criterion determined who these were. Judaism developed a parallel doctrine during the same era.121 The mitzvah (precept) of redeeming captives took on greater scope in the tenth and eleventh centuries as a result of the piracy in the eastern Mediterranean.122 The practice is attested by the documents of the Cairo Genizah.123 It is interesting to see that the development of the practice of redeeming captives as a religious duty among the Jews varied by region. Hence in the eleventh century, the Jewish population of Jerusalem accused the gaon Solomon ben Yehuda of having refused to help in redeeming captives. That gaon was then obliged to leave Jerusalem less than two weeks later to return to Ramla.124 The difference between the two fronts of Byzantium, the northern and the southern, thus lay in religion. That situation continued until the Christianization

120  Theoph. A. M. 6624 (p. 410). T. S. Noonan explains the common interests of the Byzantines and Khazars on the international scene (“Byzantium and the Khazars: A Special Relationship?”; and “The Khazar-Byzantine World of the Crimea in the Early Middle Ages: The Religious Dimension”). Noonan points out, however, that their interests were not always the same. 121  In Sheviya and Pedut: Captivity and Ransom in Mediterranean Jewish Society (pp. 22ff.), E. Bashab explains the development of the custom of putting out a cash box to collect ransom money for captives (kupa le-pidyon shevuym). 122  Y. Friedman, “The ‘Great Precept’ of Ransom: The Jewish Perspective.” 123   Genizah is the name of the room in a synagogue where books and documents written in Hebrew are kept. Because they are written in Hebrew and bear the name of God, they cannot be destroyed. They are therefore hidden away in that storage space. The root of the word Genizah means “to conceal forever.” The Cairo Genizah, discovered in the old synagogue of Al-Fustāt in the nineteenth century, is the most famous and contains more than two hundred thousand pages of books and documents dating from the late ninth century. 124   T.-S. 10 J 272, Mann I, vol. 2, supp. chap. 2, pp. 347–349. The “mitzvah” of redeeming captives in the Talmud of Babylonia became a major religious duty in the twelfth century with the rules of Maimonides (1138–1204) (Bashab, Sheviya and Pedut, p. 23 n. 13). “Gaon” (pl. “geonim”) was an honorific title given to the head of aTalmudic academy in the Babylonia of the post-Talmudic era.

616

Rotman

of the Slavs and Bulgars, that is, until the ninth century. But what happened subsequently? I have shown that the dichotomy between Christianity and Islam corresponded to the situation of Byzantium in its relations with the Arabs. On the Balkan front, until the ninth century the enemies of Byzantium were pagans. There as well the Byzantines conducted a war against the infidels. But they also took advantage of the fact that their Balkan enemies were pagans, and they conducted a policy of Christianization. The Christianization of pagans was easier than that of a people with a monotheistic religion. In addition, the church was much better prepared to convert pagans, an area in which it had a great deal of experience. Byzantine legislation attests to the persistence of the custom of enslaving Bulgars, particularly in the tenth and eleventh centuries. The Novellae in question, because of their legislative nature, have to do with slaves of any origin, but they give details about the Bulgar slaves especially. The Novella of John I Tzimisces, dating to between 972 and 975, fixed a tax that Byzantine soldiers had to pay when they sold the prisoners of war they had enslaved. That tax applied only to sales to a person outside the military.125 That Novella, the only one attributed to John I Tzimisces that has come done to us, has been discussed by Helga Köpstein and by Taxiarchis Kolias.126 The two authors adopt different points of view: Köpstein considers the Novella evidence of the imperial authorities’ struggle against the powerful elite, whereas Kolias concentrates more on its fiscal aspect. It is of interest here for two reasons: first, because it deals with prisoners of war privately owned by the military men who had captured them, whether in the Byzantine infantry or navy, and specifies that these were Bulgar captives;127 and second, because the taking of Bulgarian captives occurred at a time when the Bulgars were Christians. The Novella attests to a well-known fact: the Bulgaro-Byzantine wars continued well after the Christianization of the Bulgars (864) and after the church of Bulgaria came under the religious hegemony of the patriarchate of 125   Nov. Post. Just., coll. 3, nov. 25. 126  H. Köpstein, “Einige Aspekte des byzantinischen und bulgarischen Sklavenhandels im X. Jahrhundert: Zur Novelle des Joannes Tzimiskes über Sklavenhandelszoll”; Kolias, “Kriegsgefangene, Sklavenhandel und die Privilegien des Soldaten.” 127  Hence the rule of the Leges militares (version B), 48, promulgated with the Ekloga— namely, that prisoners of war were not part of the booty and had to be held for an exchange—was not applied in the tenth century to the Bulgar captives. In fact, the taking of captives and their enslavement seem to have been the general rule in the Balkans until the eleventh century.

Medieval Slavery In A New Geopolitical Space

617

Constantinople.128 The correlation between the Eastern enemy’s religion and its politics might lead to the conclusion that once those in the Balkans were no longer religious foes, they would also not be political adversaries. But that reasoning, which may have encouraged Byzantium in its politico-religious strategy, ultimately did not apply to the Balkans of the ninth to tenth centuries, especially where the Bulgars were concerned. In fact, Byzantium did not seek to make its enemies in the Balkans its equals. The proof is that the Byzantine authorities insisted on obedience to the patriarchate of Constantinople. In that respect it found itself in competition with the Latin West. Religious dependency must also have translated into political submission, the cause of the battle conducted by the Bulgar king Simeon and later by Kings Boris II and Samuel.129 On the Byzantine side, however, the view that stipulated the submission of these peoples to the imperial authority combined the political sense of the word douloi, expressed, for example, in the political treatises of Constantine VII Porphyrogenitus, and the sense of “slaves,” the reality being that the Byzantines procured slaves for themselves among these same peoples.130 The military success of the Byzantines, which was greater on the Balkan front than against the Arabs, was thus expressed as the benefit to be drawn from taking captives. The difference between the two fronts of Byzantium was also reflected in the ethnic profile of Byzantine slaves of foreign origin— the war captives who remained in Byzantium without being exchanged were Slavs, Bulgars, and Russians. The term assigned to most of these prisoners was “Scythians.”131 128  An overview of Bulgaro-Byzantine political and religious relations and of the events that preceded the Bulgars’ conversion is provided in R. Browning, Byzantium and Bulgaria: A Comparative Study across the Early Medieval Frontier, pp. 54–55, 144ff.; and in G. CankovaPetkova, “Contributions an sujet de la conversion des Bulgares au christianisme.” 129  For another case of political submission—that of the Serbs—to the Byzantine emperor in the tenth century, see the description of Constantine VII (Const. Porph., De adminis. 32). 130  Const. Porph., De adminis. 8 and 28–30. 131  In her article “Nommer les Russes en grec, 1081–1204,” Patlagean shows that the term “Scythians” had not an ethnic meaning but a geographical, cultural, and political-historical one. That is, as in antiquity and late antiquity, those designated “Scythians” were peoples from the Black Sea hinterland, who were perceived as Barbarians by the empire. Hence, in the tenth century, the Russians were called Rhōs in a political context that considered them a known and non-Barbarian political entity. In other words, the author’s intent determined which term was used. In the hagiographical sources, the term “Scythians” refers to the Barbarians of the northeast (the Balkans, eastern Europe) and adds a cultural sense of “non-Christian.” “The Scythians are not Christians,” said the author of the Spiritually

618

Rotman

Byzantium continually enslaved Slavs and Bulgars even after they had converted to Christianity. In fact, the political enslavement of the Balkans went hand in hand with their conversion. The counterpart to that policy in the private domain was the practice of converting slaves while keeping them in the same legal and social positions. This occurred among the Jews as well. Simha Assaf, who has studied the subject in the Responsa of the geonim in the ninth to eleventh centuries, shows that the conversion of slaves to Judaism was not only recommended but obligatory.132 In the Hebrew sources, the process is called Tevilah leavdut, “ritual immersion for slavery,” that is, the obligation to immerse the person with the aim of making him or her a Jewish slave, parallel to the Tevilah-le-geirut, which, as the sources explain, is the immersion of a person not a slave who wishes to convert to Judaism.133 All the same, in the era concerned, Jews who wanted to “immerse slaves to have them as Jewish slaves,” had to contend with the state’s legislation, which placed limits on the phenomenon. In addition, Jews were not always ready to convert their slaves because, once the slaves had become Jews, they had to be emancipated after seven years, following the rule for the eved ivri (Hebrew slave).134 The evidence in the Responsa and in the documents of the Genizah shows that people preferred to convert their slaves to Judaism, but they sometimes did so in their wills, while also making provisions for the slaves’ emancipation after the owners’ death.135 Beneficial Tales of Paul of Monembasia (9) in the tenth century. During the period of concern here, the name “Scythians” was attributed to the Russians, the Pechenegs, the Serbs, and also the Bulgars (though the Bulgars were often called by their ethnic name). For the Bulgars called “Scythians,” see the story of the abduction of Blaise of Amorion by the “Scythians,” who take him to Bulgaria (Life of Blaise of Amorion 8). See also the use of “Scythians” and “Bulgars” as synonyms in Life of Luke the Younger 24 and 32–34. 132  S. Assaf, “Slaves and Slave Trade among the Jews in the Middle Ages,” pp. 91ff. The Responsa (plural form of responsum) is the name for a Halachic jurisprudential system dealing with ritual, ethical, and civil aspects and using a procedure whereby questions are asked of rabbinical authorities, sometimes by very remote communities, so that they may receive guidance on Halachic problems. The Halacha is the body of Jewish law. 133  The term in Hebrew is Ben-horin (“free”); S. Assaf, The Responsa of the Geonim in the Manuscripts of the Genizah, pp. 88, 91–92, 95. 134  Even non-Jewish slaves could not be used on the Sabbath. But it was preferable to use Jewish slaves in the kitchen so as not to make the wine neseh (according to Jewish law, wine must not be touched by a non-Jew). 135  Even against a slave’s will (Assaf, “Slaves and Slave Trade among the Jews in the Middle Ages,” pp. 94–95); S. D. Goiten, A Mediterranean Society: The Jewish Communities of the Arab World as Portrayed in the Documents of the Cairo Genizah, 1:136–147 (for wills

Medieval Slavery In A New Geopolitical Space

619

The Byzantines’ Loss of Freedom Piracy was a particularly effective means for procuring slaves. It was an old phenomenon, but it took a new form in the Byzantine world, especially in the Mediterranean, and was increasingly present in the literary sources from the ninth century on, especially in historiography and hagiography. These raids disturbed the Byzantine population along the coast, especially after 826, when Arab forces succeeded in occupying Crete. The island served as a base for Arab piracy against the Byzantines until 963–969, when Nicephorus II Phocas recaptured it.136 Nevertheless, the Arab forces that had been expelled from Spain and that occupied Crete appear to have been independent from the other Arab forces (Egyptian, Syrian, and African).137 This can be discerned by their strategy: the aim was not to conquer all Byzantine lands, with Constantinople as the final objective, but to abduct the Byzantine population for its market value. It is obvious that this operation was dictated by a commercial demand for slaves in the Arab territories. Hence in the ninth century, slavery became one of the most important issues on the Arab-Byzantine Mediterranean scene and the major concern of the Byzantines. And not only because of the Arabs: the sacking of Thessalonica in 904 was committed by Egyptian forces, without the knowledge of the Arabs of Crete.138 In fact, there were not only acts of sea piracy but also raids organized and carried out by infantries. In the ninth and tenth centuries, that became a strategy of the Arab forces, which launched raids on the Byzantine cities in Asia Minor and withdrew after capturing and enslaving the local population—in the attack on Heraclea, for example. In the tenth century, raids against the Byzantine population were conducted in the Balkans as well.139 The Bulgaro-Byzantine wars brought about instability in the regions of continental Greece, especially during the reign of Simeon the Bulgarian. All these military operations profoundly marked the Byzantine population and also had economic consequences. In the Life of Luke especially) and 5:148–150. For emancipation encouraged by the geonim, see Assaf, “Slaves and Slave Trade among the Jews in the Middle Ages,” pp. 101 ff. 136  H. Ahrweiler, Byzance et la mer: La marine de guerre, la politique et les institutions maritimes de Byzance aux VIIe–XV e siècles, pp. 93ff. 137  V. Christides, The Conquest of Crete by the Arabs (ca. 824): A Turning Point in the Struggle between Byzantium and Islam, pp. 81ff. 138   The sack of Thessalonica is described in Ioannes Kaminiates, De expugnatione Thessalonicae (esp. 35ff.); Vasiliev, Byzance et les Arabes, 2, pt.1:166ff.; and Christides, The Conquest of Crete by the Arabs, pp. 159–161. But Kazhdan calls into question the authenticity of Kaminiates’ text (“Some Questions Addressed to the Scholars Who Believe in the Authenticity of Kaminiates’ ‘Capture of Thessalonica’”). 139  Vasiliev, Byzance et les Arabes, 2, pt.1:11–12.

620

Rotman

the Younger, for example, the family flees the Arab pirate attacks on the isle of Aegina in Hellas, to Thebes and Peloponnesus. The father finally settles in a village in Phocis, but there he faces the opposition of the villagers. He therefore travels to Constantinople to ask for help from the emperor. An imperial envoy arrives in the village to again divide up the lands.140 The ravages of the Bulgars are also predicted by Luke the Younger. Many are taken captive, then killed or reduced to slavery; some residents take refuge in the cities, while others, especially peasants, flee to the south, to Euboea, the Peloponnesus, and the Gulf of Corinth.141 The same people who fled the Greek isles for the continent found themselves threatened from the north as well. In the Byzantine hagiographical sources, only Arab piracy is mentioned. But raids organized by the Byzantines also occurred. The Byzantine navy raided parts of Syria, Palestine, and even Egypt, as al-Balādhurī relates for the year 710.142 On the reign of Theophilus, Michael the Syrian wrote: “During that time, the Romans came to Antioch by sea, as far as the port; they pillaged merchants, took captives, and left again on their ships.”143 Similarly, in 988 Ibn Hawqal, in his description of the Mediterranean, related: “In our time, the Byzantines were relentless in attempting raids on the coast of Syria and the beaches of Egypt. They drove off the ships of the coastal residents along all the coasts and captured them everywhere. No help, no aid was procured from the Muslims, and no one cared.”144 The eleventh-century documents of the Cairo Genizah also refer to Byzantine piracy: its victims included the Jews of Palestine, who were sold as slaves to Byzantium or redeemed by their Byzantine coreligionists. Consider, for example, the letter from the leader of a Jewish community who collected five thousand dinars to ransom two hundred people.145 The author mentions a large number of Jewish captives returning from Byzantium (Edom), where they were redeemed by the local communities. Upon their return to Palestine, they passed through the city of the author, whose name is not indicated, and the community there provided for them. These two hundred ransomed persons 140   Life of Luke the Younger 2. 141  Ibid., 24, 32–34. 142  Al-Balādhurī, Kitāb Futūh al-Buldān, p. 133 (pp. 112–113 in the F. Sezgin edition); Christides (The Conquest of Crete by the Arabs, pp. 163–165) provides a list of references in the Arabic sources that speak of it. 143  Michael the Syrian, Chronique 12.21 (3:100), J. B. Chabot translation [my translation from the French—Trans.]. 144  Ibn Hawqal, Kitāb Sūrat al-ard (Configuration of the Earth), BGA ed., p. 205, J. H. Kramers and G. Wiet translation, 1:19 [my translation from the French—Trans.]. 145   T.-S. Loan 28, folio 1 (Mann II, 1:354–356, commentary by Jacob Mann, pp. 348–349).

Medieval Slavery In A New Geopolitical Space

621

were only part of a much larger group of captives sold as slaves. The letter is a response to a request from the Jewish community of Al-Fustāt. The objective of the Byzantine assaults, exactly like that of the Arabs, was to abduct civilians, who were then taken to Constantinople, where they were sold or kept with an exchange in view: witness the Sylloge Tacticorum, attributed to Leo VI, and the first-person description by Hārūn ibn Yahyā, who recounts how he arrived at the Byzantine court as a captive abducted in Ashkelon.146 Although this type of tale suggests that war captives went to the court of the caliph or of the emperor, that was not always the case. Hence Emperor Leo VI asserted that if the enemy did not want an exchange of prisoners, they could be disposed of as desired, attesting that the custom of selling them still existed.147 In the exchange of 845, the number of Arab prisoners in the hands of the Byzantines was far greater than the number of Byzantines who were captives of the Arabs. Thus, to make their numbers even, al-Wathiq ordered the Byzantine slaves who had been sold in Baghdad and Rakka to be ransomed, and he personally brought out the Byzantine women from his own harem.148 In Asia Minor, the Byzantines used the same strategy. According to the Tactica of Leo VI, the Arabs of Cilicia were to be attacked by sea when they attacked by land, and vice versa.149 And in 855 the Byzantines seized the city of Anazarbus without warning and took into captivity the Zutt tribe—including the women, children, and livestock—which the Arabs had installed there in 835.150 146   Sylloge Tacticorum 50 (following Kolia-Dermitzaki, “Some Remarks on the Fate of Prisoners of War,” pp. 585–586); Hārūn bn. Yahyā recounts the story of his captivity (between 880 and 890 or 912–913) in a passage cited in Ibn Rusta’s Kitāb al-Aʿlaq al-nafīsa, BGA ed., pp. 199–127. A translation of the passage describing Constantinople is provided in Vasiliev, Byzance et les Arabes, 2, pt. 2:382–394. The best-known description is that of the poet Abū Firās in his Rūmiyyāt (Byzantine poems), written during his imprisonment in Constantinople between 962 and 966. See also L. Simeonova, “In the Depths of Tenthcentury Byzantine Ceremonial: The Treatment of Arab Prisoners-of-War at Imperial Banquets.” Simeonova points out that only illustrious prisoners received such treatment. And see Kolia-Dermitzaki, “Some Remarks on the Fate of Prisoners of War,” pp. 599–606. 147   Leonis Imperatoris Tactica 16.9–11. Athina Kolia-Dermitzaki estimates that between 50 percent and 70 percent of captives were sold (“Some Remarks on the Fate of Prisoners of War,” pp. 604–605). For the use of captives in artisanal production, which exploited their knowledge, see S. Patoura, Οἱ αἰχμάλωτοι ὡς παράγοντες ἐπικοινωνίας καὶ πληροφόρησης (4ος–10ος αἰ..). 148  Vasiliev, Byzance et les Arabes, 1:201, based on al-Tabarī, Ta‌ʾrīkh al-rusul wa’l-mulūk, 3:1353 and app. p. 314. See also Y. Ragib, “Les esclaves publics aux premiers siècles de l’Islam.” 149   Leonis Imperatoris Tactica 18.138. 150  Vasiliev, Byzance et les Arabes, 1:223, based on al-Tabarī, Ta‌ʾrīkh al-rusul wa’l-mulūk, 3:1426.

622

Rotman

Prisoners were also used as spies. The unavoidable reference is the story “Abū Qir wa Abū Sir,” recounted in the Thousand and One Nights, in which Abū Qir wrongly accuses Abū Sir (both natives of Alexandria) of being a spy for the Byzantines. He explains to the Arab king that both were prisoners of the “sultan of Christians” and that Abū Sir, to liberate his wife and his child from the Roman sultan, promised the sultan to return to Arab territory and kill the king (malik).151 For the duration of the Arab occupation of Crete, the Byzantines repeatedly tried to recapture that island, with short-lived successes.152 At the same time, they—like the Arab forces—launched maritime attacks intended to abduct the civilian population, though the Byzantine attacks are less well attested. They conducted a similar strategy in Asia Minor. In 942–943, for example, Byzantine forces arrived in Diyarbakir and captured a large number of residents before devastating the entire region as far as Arzen.153 Al-Muqaddasī (d. late tenth century) describes how the Byzantine ships took Arab prisoners to Palestine, where they were exchanged or ransomed (three for a hundred dinars). He names the following guard stations (ribāt) where such exchanges occurred: Gaza, Mimas, Ashkelon, Mahuz (port of) Ashdod, Mahuz Yubna, Jaffa, and Arsuf (the former Apollonia).154 In the episode of the miracle of the mother and of the dragon of Theodore Tyron, the hero, Theodore the Recruit, is the leader of a military expedition conducted with the objective of abducting Arabs and taking them to the emperor.155 Theodore arrives in Syria, where he captures forty men, eighty women, and a hundred children. Upon their return, the men are thrown into prison and the women enslaved. That narrative, a Byzantine version of King Kong, is very unusual in that it mentions the Byzantine raids. But why are the Byzantine sources silent about their offensive side in such a war? It is quite possible that an ideological motive is concealed behind that silence. 151   Alf Layla wa-Layta, night 874 in the Habicht and Fleischer edition, 11:24ff. In the tenth century, the term “sultan” meant “power, authority” or “holder of power, authority” (Ency. of Islam, 9:849, s.v. “sultān” [J. H. Kramers and C. E. Bosworth]). The Mahdi edition of the most ancient sources does not contain that night. 152  Tibi, “Byzantine-Fatimid Relations,” esp. appendices 3–7 (pp. 103–107), which present other episodes involving Byzantine attacks. 153  Yahyā Ibn Saʿīd Al-Antāq, PO 18, pt. 5:730 (chap. p. 98). 154  Al-Muqaddasī, Ahsan al-taqāsīm fī maʿrifat al-aqālīm (The Best Divisions for Knowledge of the Regions) 194–195 (pp. 209–210 in the A. Miqucl French translation; p. 184 in the B. A. Collins English translation). 155   Theodore Tyron, Miracle of the Dragon 2. For that saint, see N, Oikonomides, “Le dédoublement de saint Théodore et les villes d’Euchaïta et d’Echaneia.”

Medieval Slavery In A New Geopolitical Space

623

Despite the raids by the Byzantine army, those launched by the Arabs, especially during the period of their domination in Crete, continued to disturb the population on the Byzantine coast of the Mediterranean. As a result, a considerable number of Byzantine inhabitants were kidnapped by Arab pirates and sold as slaves in the Arab world. That was the situation in the Greek isles, Sicily, southern Italy, Peloponnesus, the coasts of continental Greece, and the Aegean coast of Asia Minor. The Byzantine strategic response, especially in the Greek isles, Peloponnesus, and Sicily, was sometimes to evacuate the local population. All the hagiographical narratives of that time that are set in these Mediterranean regions attest to this. The raids were not limited to the coastal cities and villages but also threatened areas inland.156 The motivation for these acts was the market value of the persons abducted, who could be sold in the Arab territories. For the pirates, however, that price could also be obtained if the families redeemed the abducted. That custom was very widespread, as detailed, for example, in Methodius’s In Praise of Nicholas of Myra. The author recounts the attack on Lesbos by the Arabs of Crete. These Arabs capture a priest from Mytilene and his three disciples. Nicholas then goes to Crete and liberates them.157 Another source for the same saint’s Life— Three Miracles of Nicholas of Myra—relates how the son of a peasant is captured by the Arabs of Crete; the saint returns him to his father.158 Similarly, in the Spiritually Beneficial Tales of Paul of Monembasia and in the Life of Nilus of Rossano, a bishop, a metropolitan, and a monk leave for the Arab countries in search of captured persons—who are in all cases their spiritual responsibility— in order to redeem them.159

156  A few examples: Life of Luke the Younger 2; Spiritually Beneficial Tales of Paul of Monembasia 8 (for Calabria); Life of Leo Luke of Sicily 4ff., in which the Sicilian peasants flee into the mountains of Calabria because of the Arab attacks; Life of Athanasia of Aegina 1, in which the saint loses her husband during the Arab attacks on the island; Life of Theoctiste of Lesbos 15, in which the heroine leaves her city for her sister’s village and is abducted by the pirates of Crete; Life of Elias the Younger, 3–4ff.; Life of Nikon Metanoeite 20, 23; Life of Theodora of Thessalonica 6, in which Theodora’s family leaves the island of Aegina and moves to Thessalonica because of the attacks by the Arabs, who take captives and kill residents, including Theodora’s brother. 157  Methodius, In Praise of Nicholas of Myra 42–43. 158   Three Miracles of Nicholas of Myra 8ff. These scenes involving piracy by the Arabs of Crete helped Kazhdan to date the narratives (A. Kazhdan, “Hagiographical Notes,” Byzantion 54 [1984]: 176–181). 159   Life of Nilus of Rossano 68–69; Spiritually Beneficial Tales of Paul of Monembasia 8.

624

Rotman

In another example, a widow sells a property to the monastery of Iviron for fifteen nomismata to redeem her son captured by the Arabs.160 The act of redemption is portrayed as being carried out by an individual, which accorded with the pirates’ objective and differed from the exchanges organized by the public authorities. Redemption was not always viewed favorably. Hence Nilus of Rossano opposes the journey of the Calabrian metropolitan to redeem members of his flock because that act requires negotiations with the infidels. But he himself sends a hundred nomismata to the emir of Palermo to ransom three monks from his monastery. The metropolitan, named Blattōn (Vlaton), manages to emancipate captives in Africa because he persuades the Arab governor of Africa that the woman the governor married, herself a captive abducted by the Arabs, is in fact Blattōn’s sister.161 In any case, individual redemptions and public exchanges did not always succeed in rescuing the abducted persons, who, as Ibn Butlān attests, were sold at Arab markets.162 Another result of the frequent abductions was a new type of sainthood. Theoctiste of Lesbos, for example, whose Life dates from the early tenth century, is captured by the Arabs of Crete but manages to escape during a stop in Paros.163 Her sainthood is essentially attributable to her abduction, as is that of Joseph the Hymnographer, whose Life (late ninth century) describes his captivity and imprisonment in Crete, where he becomes the spiritual leader of the Christian captives.164 In these cases, sainthood is linked either to escape or to capture by the Arabs. The two themes, in fact—suffering as destiny and the misfortunes caused by the Muslim infidels—merge. The story of Joseph the Hymnographer is a characteristic example, but the phenomenon reaches its apogee with the story of Elias the Younger, whose vocation manifests itself through his abductions by the infidels, his loss of freedom, and his sale as a slave, which lead him to serve as a missionary in the Muslim countries.165 Sainthood adapted itself to the extremely difficult situation of that Byzantine

160   Iviron I 16 (of 1010). 161   Life of Nilus of Rossano 70–72. See G. da Costa-Louillet’s note “Saints de Sicile et d’Italie méridionale,” p. 152 n. 2. We find the same motif in the story told by Procopius, in which Khosrow’s marriage to a Christian captive from the city of Sura makes it possible to redeem other captives (Proc., BP 2.5). Note that in all the sources specifying the price and the number of persons, it is always three persons ransomed for a hundred nomismata, which corresponds to the evidence in the documents of the Genizah. 162  A. Mez, The Renaissance of Islam, pp. 160–163. 163   Life of Theoctiste of Lesbos 15. 164   Life of Joseph the Hymnographer 6ff. 165   Life of Elias the Younger 4ff.

Medieval Slavery In A New Geopolitical Space

625

population and developed a new model. It proved to be advantageous to set oneself up as a unique victim. The Byzantine Jews were a special case: they could be redeemed by their coreligionists living in Arab territories, as attested by the documents of the Cairo Genizah. The community of Alexandria occupied a strategic position propitious for the purchase of Byzantine Jews abducted by the Arab pirates and sold in the ports of Damietta and Alexandria. But that community did not always have the financial means necessary, and they corresponded with the rich community of Al-Fustāt to obtain ransom money. A letter dated 1028 attests to the abduction of ten Jews from Attaleia (Anatalia in the text).166 Another relates the case of an abducted Byzantine Jewish woman.167 Still another speaks of Byzantine Jewish captives, including a woman, a child of ten, and a doctor and his wife, clearly attesting that the Jews abducted in Byzantium were not necessarily merchants.168 This same letter states that the rich Ephraïm bn. Shemarya of Al-Fustāt did everything in his power to see that Jews not remain slaves of the Muslims. A document published by Jacob Mann attests to the abduction of five young Jews of Assarbilo (which David Jacoby has identified as Strobilos) who were taken to Alexandria.169 Furthermore, a letter from the community of Alexandria dated 1026 and addressed to the Palestinian Jewish community of Al-Fustāt concerns the redemption of the Byzantine Jews.170 The most lively description is found in a letter to Yehuda bn. Sa‘adya, probably composed by the captives themselves.171 Attacked at sea by Arab pirates, they were taken to Egypt and purchased by Muslims (Ishmaʿelim—“Ismaelites”) and Christians (ʿarelim—“the uncircumcised”). But the leader of the local Jewish community intervened with the authorities, who ordered the purchasers to return their new slaves to the pirates within five days. And since the community did not have the means to redeem or to provide for them, they turned to Yehuda bn. Sa‘adya le Naguid, the leader of the Jewish community of Al-Fustāt. As for the Arab pirates (shodedim—“plunderers”), who were anxious to dispose of their captives so that they could conduct another raid in Byzantium (Iavan), they planned to sell the captives to Muslims and Christians or to kill them if they were not redeemed by the deadline. 166   T.-S. 13 J 1420 (Mann I, no. 12, 1:87). 167   T.-S. 13 J 2025 (Mann I, no. 13, 1:88). 168  MS. Adler, 2804, fol. 7 verso (Mann I, no. 15, 1:89–90). 169   T.-S. 13 J 34 (Mann I, 2:344–346, supp. chap. 2); D. Jacoby, “What Do We Learn about Byzantine Asia Minor from the Documents of the Cairo Genizah?” p. 90. 170   T.-S. 24.29 (Mann II, 1:367–370). 171   T.-S. 10 J 278 (Mann I, 2:363–365, supp. chap. 2).

626

Rotman

Another letter, written by two brothers in Jerusalem in 1053 and sent to their sister in Toledo, tells the story of two Spanish Jews abducted by the Byzantines on their way to Palestine. They were sold in Ramla and probably redeemed by the local community.172 That practice was also followed in Spain. Hence three letters from Yehuda Halevi (1075–1141) speak of his attempt to pay the ransom price of a Jewish woman who had been abducted and sold in Toledo.173 Like Byzantine hagiography, the Hebrew literature of that time also depicted the captive. The story that comes to mind is that of the four Jewish captives in the seventh part of Sefer Ha-Qabbalah (The book of tradition) by Abraham Ibn Daud (mid-twelfth century). The four great scholars (hahamim gedolim), who leave by boat from Bari, are abducted by Arab forces. They are then sold to different countries (Egypt, Ifriqiya, and Spain). Three of them become the founders of major rabbinical schools: in Al-Fustāt, Qairawan, and Cordova.174 Despite the change in the status of Byzantine captives introduced by the Ekloga in the mid-eighth century, the Roman and Byzantine legal systems did not foresee the situation of the ninth and tenth centuries, where so many civilians found themselves slaves in enemy territory. One of the social consequences of that political reality can be found in the legislation of Leo VI in the late ninth and early tenth centuries, which included 113 Novellae, a number comparable only to that of the Novellae of Justinian. Leo VI’s legislation was very concerned with the problems of the time. For example, Novella 40 resolves the difficulty of the property belonging to a Byzantine in captivity.175 Previously, such a captive had been prohibited from preparing a will.176 Leo VI modified that law to allow captives to draw up a will valid in the empire provided that they not bequeath to the enemy. That case suggests that relations existed between captives and their families. In the same way, that legislation affirmed (in Novella 36) the son’s right to inherit if 172   T.-S. 13 J 94 (S. Assaf, Texts and Studies in Jewish History, pp. 108–113). 173  S. D. Goitein, “Autographs of Yehuda Halevi,” pp. 397ff. 174  Abraham Ibn Daud, The Book of Traditions (Sefer Ha-Qabbalah), ed. G. D. Cohen. See Cohen’s “The Story of the Four Captives,” in which he argues that Ibn Daud used known literary elements to create a fictive history. 175   Nov. Leo. 40. 176  From the third century on, a slave who was the heir of a person who died in captivity could inherit if the testator had made his will prior to his captivity (Dig. 38.1. 12, “servus heres scriptus ab eo, qui in hostium potestate decesserit, liber et heres erit”). There, in a word, is the explanation for why a Roman captured by the enemy was considered a slave: being within the enemy’s potestas, he had no potestas of his own. Note the version of the law in the Basilica: “The captive’s last will is considered as if he were not a captive” (Basilic. 35.1.13). Unfortunately, the scholia of that law have not come down to us.

Medieval Slavery In A New Geopolitical Space

627

his father died a captive. But though the captive now had the right to prepare a will, the question remains: what notaries could he have used to do so?177 The delimitation between slave and nonslave in the Roman Empire was defined and imposed by the Roman authorities on the basis of its legal categories. These categories aspired to be universal, but that universality turned out to favor the Roman conqueror. Hence the delimitation between slave and free was also in the empire’s hands. That was no longer the case in the Byzantine world after the seventh century. From that time on, this delimitation did not depend solely on the Byzantines, because it was no longer their empire alone that set the rules of the game on the geopolitical scene. Thus far, I have examined the delimitation between slave and nonslave from the point of view of the reduction to slavery. As for the reverse procedure, emancipation, prisoner-of-war exchanges became the means to restore freedom to persons who had become slaves at the hands of the enemy. Although at first glance redemption resembles the exchange of prisoners, it deserves special attention. In such cases, a ransom price had to be agreed upon, even though, from the eighth century on, the person was free de jure. Moreover, ransoming was an altogether private economic transaction. The public authorities, despite the exchanges of captives they conducted, proved to be unable to secure de facto the freedom of their subjects, and it was up to the subjects to take care of such matters. Nevertheless, using their judicial authority, the public powers modified the delimitation between slave and nonslave within the empire: when such captives entered Byzantine territories, they were de facto free. That development in the definitions and delimitations of legal status went hand in hand with an evolution in Byzantine law: it became a Christian law. I mentioned the similarity between Byzantine law and Islamic law with respect to these definitions. But unlike Islamic law, which was a religious law from the start, Byzantine law had its roots in a nonreligious system and outlook and conserved its nonreligious structures and rationales. The evolution of Byzantine law concerning the delimitation between slave and free person, though it began in late antiquity, can also be understood within the context of the influences and adaptations of the emergent medieval world in which 177   Nov. Leo. 36. The permission granted in the third century to inherit from a person who died in captivity was limited to wills prepared before the person was taken into captivity. Note that the Ekloga does not deal with all the consequences resulting from the change it introduced. Unlike in the Novellae of Justinian, the status of slaves occupied an important place in the Novellae of Leo VI. Of the 113 Novellae, more than a quarter had to do with changes in the status of the slave as individual.

628

Rotman

Byzantium found itself in the seventh century. For the first time, its primordial adversary came to be its religious foe as well. Another innovation appears in Byzantine treaties of the tenth century, which include a paragraph dealing with slaves who flee to the adversary. The problem of fugitive slaves had always existed. For example, many slaves had taken advantage of Alaric’s siege of Rome by fleeing; Olympiodorus points out that these slaves were “especially of Barbarian origin.”178 In the fourth century, Procopius spoke of fugitive Ethiopian slaves who deserted the Ethiopian army. They created an independent force and staged a coup d’état in the Himyarite kingdom (Yemen), having chosen as their leader a certain Abramos, the slave of a Roman. They constituted a political threat that the Ethiopian king could not rid himself of.179 The first military expedition he sent to oppose Abramos defected and joined forces with the slave leader. The second returned without success. The episode ended with Abramos agreeing to a tribute to the Ethiopian king. A special provision in certain peace treaties resolved the question of refugees by decreeing that they had the right to return to their native countries. It is clear that, in a politically and economically unsettled situation, residents would try to find a better life where they could.180 For the state, that meant not only a loss of residents, especially in the border regions, but also a loss of the money these residents paid in taxes. In the fifth and sixth centuries, some Byzantines sought to settle outside the ravaged region and thus escape their fiscal obligations. Later, however—when the neighboring state was Muslim—refuge was no longer such a simple solution. The Byzantines became dhimmīs in Muslim territory, and as such had to pay special taxes. Conversion to Islam, moreover, often proved to be too radical a change, though it could be useful to abducted Byzantines, to refugees no longer able to maintain their social position, or to conquered Byzantines. Hence Nikon Metanoeite traveled to Crete to bring Christians who had converted to Islam during the Arab occupation back to Christianity.181 In other words, in the new medieval world, where the political adversary of Byzantium 178  Olympiodorus, frag. 7. 179  Proc., BP 1.20. For Justinian’s policy in that region in 530, see I. Shahid, Byzantium and the Arabs in the Sixth Century, 1, pt. 1:144ff.; I. Shahid, “Byzantium and Kinda.” The Himyarite kingdom had strategic importance not only for Ethiopia but also for the commercial pathways between the Arabian Gulf and the Persian Gulf on the route to India (Cosmas Indicopleustes, Christian Topography 2.50, also mentions the war conducted by the Ethiopian king to capture this kingdom; ibid., 2.56). See also “Homeritae” in RE 8, pt. 2.2182–2188. 180  Proc., BP 1.15. 181   Life of Nikon Metanoeite 20.

Medieval Slavery In A New Geopolitical Space

629

was also its religious adversary, the phenomenon of refugees was not widespread. On the contrary, the state now had a new role, that of protecting the religion of its residents. (During the iconoclastic era, however, that role was undermined: the Byzantine authorities persecuted those who venerated icons, and territories under Arab control were no longer considered an adequate refuge.)182 Hence refugees are no longer mentioned in the Byzantine-Arab treaties but are still mentioned in those between Byzantium and the Bulgars so long as the Bulgars remained pagans—in the treaty of 813, for example.183 I have emphasized this theme to show that refugees of free status had ceased to be an international issue. That was not true in the case of slaves. The two Russo-Byzantine treaties of 911 and 944, reported in the Laurentian Chronicle (the so-called Chronicle of Nestor), contain an article that committed at least one party to return fugitive slaves to the other.184 In the treaty of 911, it was the Byzantines who agreed to return fugitive slaves to Russian traders, whereas in the treaty of 944 that obligation applied to both sides. The Arab-Byzantine treaty of Aleppo in 969, described by Ibn al-‘Adim, committed only the Arabs to return fugitive slaves, whether Christian or Muslim, to the Byzantines.185 The importance that tenth-century treaties granted refugee slaves indicates that there must have been a significant number of them at the time. Theophanes, for example, relates the case of an Arab slave who stole money and then fled to the Byzantines.186 There is also evidence of that phenomenon in the Life of Luke the Younger, dating from the tenth century: the imperial authority has the saint arrested by soldiers who are looking for fugitive slaves.187 In a Novella dating from between 949 and 959, Constantine VII Porphyrogenitus again banned granting refuge to fugitive slaves within the 182  Byzantium sometimes returned its Arab prisoners in exchange for Christian relics and icons. The best-known example is the “holy image or holy face, the Mandylion.” In 942– 943, the Byzantines proposed to exchange their Arab prisoners to recover that image, which was located in the church of Edessa; PO 18, pt. 5:730 (chap. 93); Theoph. A. M. 6118 (p. 327); E. Patlagean, “L’entrée de la Sainte Face d’Edessc à Constantinople en 944”; A. Cameron, “The Mandylion and Byzantine Iconoclasm.” 183  Theoph. A. M., 6305 (pp. 497–498). 184   Laurentian Chronicle, in The Russian Primary Chronicle, pp. 65ff.; I. Sorlin, “Les traités de Byzance avec la Russie au Xe siècle,” pp. 329–336. 185  Ibn al-‘Adīm, Zubdat al-Halab fī ta’rih Halab, ed. S. Dahhân (Damascus, 1951), 1:163–169; French translation by M. Canard, Histoire de la dynastie des Hamdanides de Jazîra et de Syrie, pp. 831–835; Latin version in Ex Camaleddini annalibus halemensibus (in Excerpta ex Historiis Arabum 3). 186  According to Theophanes, the Arabs used the episode as a pretext to attack the Byzantines; Theoph. A. M. 6208 (p. 387). 187   Life of Luke the Younger 4–7.

630

Rotman

empire and established the rate of compensation for anyone who returned a slave to his master.188 That measure encouraged slaves to seek refuge in an enemy state where they could enjoy a different life and a new status. The twelfth article of the Arab-Byzantine treaty of 969 addresses that problem precisely: “If a Muslim or Christian slave, man or woman, flees to a country other than the territories designated, so long as the slave remains in those territories, the Muslims must not hide the slave but must denounce him or her. The slave’s owner will pay a price of thirty Greek dinars for a man, twenty dinars for a woman, and fifteen for a young boy or young girl; if the owner does not have the means to purchase the slave, the emir will charge the owner a duty of three dinars and will remit the slave to the owner.” The thirteenth article specifies that “if the fugitive slave is baptized, the Muslims will not have the right to keep the slave: the emir will charge the slave’s owner a duty and will return the slave to the owner.”189 The Russo-Byzantine treaty of 911 contains an important diplomatic innovation. There is a paragraph stipulating that a person who finds a captive from the other camp reduced to slavery in a third country must ransom him or her.190 The redeemed captive would then be transferred to his or her native country and would repay the person. Through that clause, the two sides decided to form a common defensive front against the abduction of their subjects. This was not a religious front, however, because at the time the Russians were not officially Christianized, as this treaty attests.191 That clause proves that the redemption of persons abducted from their country and enslaved had already become the rule. Recall the case related in the Life of Nahum: Byzantine diplomats in Venice found disciples of Methodius who had been abducted in Moravia and sold as slaves. Having redeemed them, the diplomats took the disciples with them to Constantinople.192 I believe that this diplomatic innovation developed as a continuation of the Byzantine legislative evolution in the status of Byzantines captured by the enemy. For them to be recognized as free by another state, they first had to be 188   Nov. Post. Just., coll. 3, nov. 13. 189  Translation in Canard, Histoire de la dynastie des Hamdanides. Note that the prices are slightly different in the version of the Ex Camaleddini annalibus halemensibus (in Excerpta ex Historiis Arabum 3): thirty-six dinars for a male slave, twenty dinars for a female, and sixteen for a child. 190   Laurentian Chronicle, p. 67. 191  Ibid., p. 65 (as in the treaty of 944, ibid., p. 77). The same treaty specifies that some of the Russians were Christians and others were not, which is important for the question of oaths. For these treaties and for the Russian ones of the time, see H. Ahrweiler, “Les relations entre les Byzantins et les Russes au IXe siècle.” 192   Life of Nahum, pp. 287–288.

Medieval Slavery In A New Geopolitical Space

631

considered so in their own state. Hence this article contains the same rule as the law that originally changed the status of Byzantine captives—Ekloga 8.2. It concerns the case in which the redeemed captive does not have the means to reimburse his purchaser. The treaty of 911, like the Ekloga, decreed that he had to pay his price to the person who had acquired him by working for him until he accumulated the sum paid. The appearance of these new subjects in tenth-century peace treaties attests to the importance of mobility—from the status of slave to that of free person and vice versa. The articles also show that this was an international problem or even that the specific international context of Byzantium made that mobility possible, unlike what had occurred in the geopolitical space of the Roman Empire. It is also clear that the problem preoccupied the Byzantines, since it was they who were risking their freedom; in other words, they too were becoming victims of the political situation. The response of Byzantium was not simply to change its internal definitions of the slave and the free person through legislation. It went further, attempting to establish through peace treaties a new international law that would allow slaves and free Byzantines to be recognized as such outside the empire. The custom of exchanging captives was the first stage in this course of action.193 Byzantium was thus obliged to recognize the same statuses among its adversaries. It is also clear that, in the fifth and sixth centuries, the difference in religion was the source of the custom of redeeming captured Christian Byzantines, and that in the tenth century that practice became widespread. That change on the political scene raised a problem for Byzantium; how to procure slaves? I mentioned the Novella of John I Tzimisces (dating from between 972 and 975) that points to the continuous wars and raids in the Balkans as a permanent source of slaves.194 This supply source was combined with the no less important one provided by trade.

193  The Russo-Byzantine and Arab-Byzantine treaties not only speak of slaves but also deal with the case of thieves and murderers who flee abroad, stipulating that each party had to return them. 194   Nov. Post. Just., coll. 3, nov. 25. The Novella of Alexius Comnenus of 1095 attests to the practice of private abductions by the Byzantines to procure slaves, especially Bulgars (Nov. Post. Just., coll. 4, nov. 35).

632

Rotman

The Slave Trade: The New Commercial Map of the Medieval World

The Commercial and Political Context for the Traffic in Slaves Since trade was an important means for procuring slaves, the place of the slave market must be situated within the general context of trade in the Mediterranean basin. My starting point is the sixth century, when the Roman Empire succeeded in extending its borders to the maximum degree. By the eighth century, this was no longer the same empire. The major changes in the geopolitical map had consequences both for slavery and for trade in Byzantium and the surrounding world. In the fifth to seventh centuries, the empire possessed vast resources for providing all sorts of commodities. Importation was one means, especially in the case of luxury goods. That was also true for slaves. Take as an example the importation occurring via the Red Sea, as described by Cosmas Indicopleutes, who mentions slaves imported to Egypt from Ethiopia.195 They are attested as well by sixth-century Egyptian papyri.196 Although castration was prohibited in the empire, it was practiced illegally, and eunuchs were luxury slaves with a higher value from the start.197 They were imported from Eastern regions such as Mesopotamia, Caucasus, and even India, which was a source of other precious commodities.198 This commercial outlook changed in the seventh century for both international and domestic trade. In the early seventh century, Heraclius, exarch of Carthage, made a wager with his vice general Gregoras: which was the quickest way to go from Carthage to Constantinople, by land (attempted by Niketas, son of Gregoras) or by sea (attempted by Heraclius the Younger)?199 By the end of

195  Cosmas Indicopleustes, a sixth-century Egyptian trader, describes his personal travels in his Christian Topography. These travels led him to the Arabian Gulf and to the Persian Gulf, as far as the mouth of the sea (2.29–30). He describes the eastern traffic between Egypt, India, China, and Ethiopia (2.43–64) and mentions the slave trade (2:64). 196   P. Str. 1404, edited with commentary in F. Preisigike, “Ein Sklavenkauf des 6. Jahrhunderts. P. gr. Str. Inv. Nr. 1404.” 197  According to two appraisals, in 530 and 531, CJ 6.43.3.1; 7.7.1.5. 198  Later laws of the fourth to fifth centuries concerning castration targeted the dux of Mesopotamia (CJ 4.42.1) and the Eastern prefect (CJ 4.42.1; cf. PLRE 2:1179; CJ 12.5,4; PLRE, 2:930). See Proc., BP 1.15.31, for the Persian-Armenian origin of Narses; Proc. BG 4.3; BP 2.15, for the importation of eunuchs from the Abasgi and Lazi Caucasian peoples; and Dig. 39.4.16.7 for the third-century tax on luxury goods, including on the spadones Indici, “the Indian eunuchs.” 199  Theoph. A. M. 6101 (p. 297).

Medieval Slavery In A New Geopolitical Space

633

the century, such a venture was no longer possible. The new geopolitical map also imposed a new economic and commercial map.200 On this new medieval map, the Jews occupied an important position in international trade, especially the slave trade. Ibn Khordādhbih’s description of the “al-Rādhāniyya” (Kitāb al-Masālik wa’l-mamālik, or Book of routes and kingdoms) is central to the entire debate by historians regarding the slave trade in the Middle Ages.201 The Radhaniyya, Jewish merchants of slaves and other goods, have been much written about in modern scholarship. They were not the only Jewish merchants specializing in the slave trade,202 but Ibn Khordādhbih’s description has the most geographical details and is also the oldest (the first version of his book was completed in about 846).203 In modern scholarship, Évelyne Patlagean’s “Byzance et les marchés du grand commerce vers 830–vers 1030. Entre Pirenne et Polanyi” (Byzantium and the large trade markets from ca. 830 to ca. 1030: From Pirenne to Polanyi) places Byzantium at the center for the first time, granting it the role of a commercial power active in the global trade of its time (ninth to eleventh century). I shall extend that chronological framework back to the seventh century, when the Mediterranean map underwent its great geopolitical shift. Despite the number of works devoted to the Radhaniyya and to Jewish merchants generally, until now there has been no map of the global slave traffic for that time.204 That explains why Byzantium has been excluded from the historical debate centering on the questions raised by medieval commerce. In other words, had a map existed, the occultation of Byzantium, an integral part of it, would not have been possible. In the belief that such a map may prove enlightening as to the particular position of Byzantium, I have constructed one, based 200  Michael McCormick, however, supports the idea of a much earlier decline of Roman Mediterranean commerce, which in his view was linked to a general decline in the empire’s economy (Origins of the European Economy, part 1, “The End of the World”), For a different point of view, see P. Horden and N. Purcell, The Corrupting Sea: A Study of Mediterranean History, pp. 153–172. Horden and Purcell show a continuity in commercial communication. I shall not join the debate here on the decline of Mediterranean traffic in the seventh to eighth centuries. I would simply like to show how the map of international traffic changed as a function of geopolitical shifts and what role Byzantium played in that change. 201  Arabic names are cited here in the standard Latin transliteration. 202  A. Gieysztor, “Les Juifs et leurs activités économiques en Europe orientale”; E. Ashtor, “Gli Ebrei nel commercio mediterraneo nell’alto medioevo (sec. X–XI).” 203  Ibn Khordādhbih, Kitāb-al-Masālik wa’l-mamālik, p. 153 (129) ff. 204  The map provided in McCormick (Origins of the European Economy, p. 762) is limited to Europe between 700 and 900.

634

Rotman

on the most notable geographical accounts and the corresponding studies. But where to begin with such a geographical map? The episode of the conquest of Constantinople by Heraclius is an invitation to examine the position of Byzantium in relation to the Mediterranean. There has been a tendency to consider that basin the navel of the world. Since the time Rome made the Mediterranean its own mare domesticum, that image has been perpetuated, even in modern scholarship. Whether in Henri Pirenne’s thesis or in Fernand Braudel’s conception, the Mediterranean has always remained at the center of any economic and commercial discussion.205 According to this view, the Mediterranean world is a unit apart. But though that was true for the Roman Empire and for the Italian cities in the late Middle Ages and Renaissance, it was less true between these two eras. As for Byzantium, it has been placed outside the debate because Constantinople is considered to be geographically on the margins. I cannot speak of the place of the Mediterranean on the commercial medieval map without referring to Michael McCormick’s writings.206 His Origins of the European Economy: Communications and Commerce, A.D. 300–900 (2001) not only demonstrates a new commercial dynamic in the eighth to eleventh centuries but also situates it within the new geopolitical-economic medieval context. The Mediterranean functions as the commercial intermediary between the European side and the Arab side, as the arena in which and around which the movement of people and commodities occurred. The slave trade had a central role in that dynamic, which always depended on the wealth of the Arab world. As for Byzantium, though it participated in that economic activity and played its part in this new world of medieval communication, it does not appear to have had a very determining role there. Yet Byzantium lived its political, economic, and commercial life between three seas. I shall therefore return the city of Constantinople to its place at the center of the international map; after all, Constantinople was founded to occupy that place. Along the edges of the map, the surrounding countries played an important role in the world market: to the west, the Spain of the Umayyads, a major consumer of imported slaves; to the north, the Baltic Sea, where Arab and Byzantine coins attest to the commercial link, between eastern Europe and the Arab world and Byzantium; to the south, the Red Sea, which opened the maritime route to India and China; and to the east, the Caspian Sea and Khorāsān, important regions for international trade because of the Khazars 205  Horden and Purcell, The Corrupting Sea, chap. 2. 206  McCormick, Origins of the European Economy; M. McCormick, “New Light on the ‘Dark Ages’: How the Slave Trade Fuelled the Carolingian Economy.”

Medieval Slavery In A New Geopolitical Space

635

and the Samanids. It immediately becomes clear that Byzantium’s centrality is much more than a geographical position. Nevertheless, a geographical map, however detailed it may be, is not sufficient for following the chronology. It also cannot situate the sources, which provide key information on trade in the space concerned but do not speak of commercial routes. That is why I have provided a table with significant information concerning the commercial map of Byzantium (table 1). This information is organized both chronologically and geographically. The regions are defined from west to east: western Europe, Spain, and the western Mediterranean, Italy-Sicily-the Adriatic Sea and the central Mediterranean (as far as Ifriqiya and Tripolitania), the eastern Mediterranean (as far as Syria, Palestine, and Egypt), Byzantium (its regions in Asia Minor and the Balkans), the Balkans (outside the Byzantine Empire), Caucasus-Russia-the Black Sea,207 and Iran-Iraq. Let me first mention a few general elements that played a role in shaping international trade in the Byzantine world. The Balkan region, and especially its Slavic and Bulgar populations, constituted the most important source of slaves for Byzantium. In the Balkans, war turned out to be an effective means for procuring slaves. The same was true of international trade: the Slavs and the Bulgars were the principal source of slaves throughout the regions mentioned. Africa was also an important source of slaves, even in late antiquity.208 Nevertheless, African slaves were not imported to the Arab world in numbers sufficient to satisfy demand. The Arabs in fact made a distinction between black and white slaves and granted a greater value to whites. In the Arabic language of Spain, black slaves continued to be called ʿabīd (slaves), whereas white slaves were designated by the term saqaliba.209 This word, which attests to the Slavic origin of these slaves, corresponds to the word sklavos, which from the twelfth century on was the Greek synonym for the word “slave.”210 In Ibn Khordādhbih’s description (from ninth-century Iraq), saqaliba designates simply “the Slavs,” whereas for Ibn Fadlān this word refers to the Bulgars. The Slavs and Bulgars therefore appear as the principal source of slaves on the commercial map. 207  When using the term “Russia,” I refer to the regions extending from the Baltic Sea to the Black Sea and the Ural Mountains. 208  B. Lewis, Race and Color in Islam, pp. 28–38; Gordon, Slavery in the Arab World, pp. 49ff. and chap. 5. 209   Ency. of Islam, 8:872–881, s.v. “Al-Saqāliba” (P. Guichard and M. Meouak). 210   H. Köpstein, “Zum Bedeutungswandel von ΣKΛΑΒΟΣ/SCLAVUS”; H. Kahane and R. Kahane, “Notes on the Linguistic History of ‘sclavus.’ ”

636

Rotman

Medieval Slavery In A New Geopolitical Space

637

638 Table 1

Rotman Trade in the eighth to eleventh centuries

Western Europe

Central Europe

Central Mediterranean: Italy-Adriatic Sea

Eastern Mediterranean: Syria-Palestine-Egypt

688: presence of Arabs in Cyprus

776: letter from Pope Adrian I to Charlemagne

822–852: al-Ghzāl on the Spain-Baltic routes 825: privileges granted by Louis the Pious to a Jewish merchant of Sargossa 846: Agobard of Lyons writes against the Jewish slave traders

824: Arab conquest 812/814: the Franks leave the Adriatic Sea of Crete 814–820: ByzantineVenetian edict against trade with the Arabs 833–844: Arab raids in Sicily and Peloponnesus; Byzantine raids in 830–860: Arab raids Syria in southern Italy 831: Arab conquest of Palermo 872/875: the Arabs in Dalmatia, a ByzantineVenetian operation

876: Venetian edict against the slave trade 878: Arab conquest of Syracuse

639

Medieval Slavery In A New Geopolitical Space

Byzantium

Balkans

Caucasus-Black Sea-Russia

Iraq-Iran

685–695: Justinian II (first reign): sales of Slavic slaves in Asia Minor (seals of the kommerkiarioi) Rhodian Sea Law (7th–8th centuries) 728: last known sigillographical mention of apothēkai

712: first kommerkiarioi, attested in Thessalonica

729: Itil the Khazar capital (first evidence) 740: documents attesting Jews among the Khazars

762: Abbassids settle in Baghdad

801: Irene lowers taxes 809: tax on slaves entering by the Dodecanese (Nicephorus I)

831–833: new kommerkiarioi

Abydikoi (officials in charge of import taxes) of Thessa­ lonica attested

860: the Serbs pay their tribute to the Bulgars in slaves 864: Christianization of Boris of Bulgaria

Leo VI (886–912)

893: transfer of Bulgar commerce from Constantinople to Thessalonica; Vladimir the Bulgar returns to paganism

841: the thema of Kherson mid-ninth century: —Life of George of Amastris (tax in Trebizond). —Russian merchants in Constantinople 860/1: Christian mission in Russia; a Jewish Khazar dynasty 867–886: coins of Basil I in Russia 895: coins of Leo VI and Samanid coins in Kiev

846: first version of Kitāb al-Masālik wa’l-mamālik, by Ibn Khordādhbih

874: Samanids settle in Bukhara

640 Table 1

Rotman Trade in the Eighth to Eleventh Centuries (cont.)

Western Europe

The Responsa of the geonim: Jewish merchants between France and Spain (tenth century)

968: Liutprand of Cremona writes of castration in Verdun of slaves destined for Spain

Central Europe

Central Mediterranean: Italy-Adriatic Sea

903/906: Raffelstetten (Linz) Customs 945: Venetian edict Regulations 906: Moravian and against the slave Russian slave trade trade in Byzantium 960: prohibition on Venetian ships transporting Jewish slave merchants 960: Abraham Iakobi, Jewish slave merchant in Prague 968: Liutprand of 961: Hungarian Jewish merchants Cremona: Byzantine customs in Corfu in Russia 971: edict against Venetian weapons trade with the Arabs

992: Byzantine privileges in Venice at the customs of Abydos

1082: Chrysobull of Alexius Comnenus to the Venetians 1090: taxes on the Jewish merchants of Salerno

Eastern Mediterranean: Syria-Palestine-Egypt

904: sack of Thessalonica Muhammad Ibn ‘Umar, Faqīh 911/912: Byzantine-Russian expedition in Crete

964/5: conquest of Crete and Cyprus (Nicephorus II) 969/970: treaty of Aleppo

641

Medieval Slavery In A New Geopolitical Space

Byzantium

Balkans

911–912: Book of the Prefect

907/911: Byzantine906: Moravian and Russian slave trade in Russian commercial treaties Byzantium 912: Russians participate in the Byzantine attack on Crete River slave traffic by the Russians

922: Ibn Fadlān, envoy to the Bulgars of Volga: route from Baghdad to the Caspian Sea

944: ByzantineRussian commercial and military treaty 957: Olga’s journey to Constantinople

988: Kitāb Sūrat alard, by Ibn Hawqal 999: the destruction of the kingdom of the Samanids

Constantine VII (913–959) (De administrando imperio)

Nicephorus II Phocas (963–969) John I Tzimisces (969–976)

961: Hungarian Jewish merchants in Russia

Caucasus-Black Sea-Russia

968: ByzantineRussian victory over the Bulgars 969: the Russians destroy Itil 971: Byzantine victory over the Russians and Bulgars 988: Christianization of Vladimir 989: Kherson under Russian control

Iraq-Iran

Ibn Butlān (mideleventh century: description of the slave markets in Cairo (Greek and Armenian slaves)

642

Rotman

As for the importation of African slaves, it merely responded to the demand of the Arab market. Hence the slave trade seems to have moved north to south and, in the Arab world, also from south (Africa) to north (Arab territories). The Arab world therefore extended the Roman Empire to the east, importing African slaves. The Slavs and Balkans in the Ural Mountains did not mint coins before the tenth century: that explains the demand in these regions for Arab and Byzantine coins, which have been found as far north as the shore of the Baltic Sea and as far east as Oka Lake on the Volga. These finds provide a more complete portrait of commercial exchanges.211 The Radhaniyya Modern specialists in the Radhaniyya have searched for their origin and have tried to discover whether Ibn Khordādhbih’s description of them is authentic.212 Historians who have examined them within the more general context of Jewish trade at the time have studied the specificity of these traders. The scholarship has illustrated that Jewish communities were found everywhere and were thus able to serve as a network for these merchants. In fact, their special merchandise, slaves, was not their only one. Eliyahou Ashtor speaks of precious fabrics and pearls—luxury goods.213 Ibn Khordādhbih also

211  M. Espéronnier, “Les échanges commerciaux entre le monde musulman et les pays slaves d’après les sources musulmanes médiévales,” pp. 17–18 (and n. 2). Byzantine, Arab, and Samanid coins, for the most part from the ninth and tenth centuries—though there are also coins from the eighth and eleventh centuries and even Byzantine coins from the sixth and seventh—were found at the following archaeological sites: Gorodische near Novgorod (E. Nosov, “Rjurikovo, Gorodišche et Novgorod,” pp. 148 and 152), Timerevo on the Volga (V. Sedyh, “Timerevo—Un centre proto-urbain sur la grande voie de la Volga,” pp. 175–178, with Western coins), Gnezdovo (T. Puškina, “Les trouvailles monétaires de Gnezdovo: Un marqueur des relations commerciales”), and Kiev (G. Ivakin, “Kiev aux VIIIe–Xe siècles,” pp. 231–232). For the total number of tenth- to eleventh-century dirhams in the finds of Russia, see T. S. Noonan, “The Impact of the Islamic Trade upon Urbanization in the Rus’ Lands: The Tenth and Early Eleventh Centuries”; and T. S. Noonan, “Khazaria as an Intermediary between Islam and Eastern Europe in the Second Half of the Ninth Century: The Numismatic Perspective.” 212  S. Assaf, “Slaves and Slave Trade among the Jews in the Middle Ages”; M. Gil, “The Radhanite Merchants and the Land of Radhan”; E. Ashtor, “Aperçus sur les Radhanites”; Ashtor, “Gli Ebrei nel commercio mediterraneo”; C. Cahen, “Ya-t-il eu des Radhanites?”; Gieysztor, “Les juifs et leurs activités économiques en Europe orientale”; and C. Verlinden, “Les Radaniya: Intermédiaries commerciaux entre les mondes germano-slave et grécoarabe.” 213  Ashtor, “Gli Ebrei nel commercio mediterraneo,” pp. 458–460.

Medieval Slavery In A New Geopolitical Space

643

mentions sabers. This is important because, as in the case of slaves, these goods traveled between the Muslim and Christian worlds. Before studying the Jews’ routes more closely, let me point out that the Jews were restricted in the slaves they could purchase. Unable to buy Christian slaves in the Byzantine Empire or Muslim slaves in the Arab world, they legally procured only imported slaves. Recall on this matter the conversion to Judaism of slaves purchased by the Jews, as attested by the documents of the Genizah. In fact, Jews were prohibited from converting slaves to Judaism in Byzantium and in Islamic regions. Legally, therefore, the Jews could “immerse to have as Jewish slaves” only foreign slaves before they arrived at their destination. In other words, slave merchants had to convert them before selling them.214 It is therefore clear that demand from a Jewish market also played a role in the development of the specialization in slave trafficking. To quote from Ibn Khordādhbih’s description of the Radhaniyya: Itinerary/itineraries of the Jewish merchants (known as) Rādhāniyya, who speak Arabic, Persian, Rūmī (= Greek), Frankish (= French), Andalusian (= Romance), and Slavonic, and travel from the East to the West and viceversa, by land and by sea. From the West they import eunuchs (khadam), young slaves of both sexes (djawārī and ghilmān), silk brocade (dībadj), beaver pelts (djulud al-khazz), [pelts of] the sable (sammūr) and [other] furs, as well as swords. [First itinerary] They embark in the land of the Franks (Firandja) on the Western Sea (= the Mediterranean) and disembark at al-Faramā (= ancient Pelusium), then they transport their merchandise by land (‘ala ’l-zahr) as far as al-Kulzum (ancient Clysma), a distance of 25 parasangs; they then traverse the Eastern Sea (= the Red Sea) from al-Kulzum to al-Djār (the port of Medina) and to Djudda. From here, they continue their journey to Sind, to India, and to China. From China, they bring back musk (misk), wood of aloes (ʿūd), camphor (kāfūr), cinnamon (dār sīnī) and other [products] which are imported from these countries. Thus they return to al-Kulzum, then transport their [consignment] to al-Faramā and embark on the Mediterranean. Sometimes, they made a detour through Constantinople (Al Kustantīnyya) with their merchandise, which they sold to the Byzantines. Sometimes furthermore, they went to sell them [in the land of the] king of Firandja.

214  According to the Responsa of the geonim.

644

Rotman

[Second itinerary] When they chose to do so, on leaving Firandja, they transported their merchandise by sea, on the Mediterranean, disembarking at Antioch (Antākiya), whence they made their way, in three overland stages to al-Djābiya; they then sailed on the Euphrates (al-Furāt) to Baghdād, then on the Tigris (Didjla) to al-Ubulla, and from here they gained access to Oman (‘Umān), Sind, India and China, all these countries being contiguous with one another. [Third itinerary] Their overland route: those among them who set out from al-Andalus or from Firandja, cross over [the strait] to Lower Sūs (al-Sūs al Aksā) and arrived in Tangier, whence they make their way to Ifrīkiya (approximately equivalent to present-day Tunisia, but more specifically Kayrawān), To Misr (Egypt, but more specifically al-Fustāt), they subsequently pass through al-Ramla (in Palestine), Damascus (Dimashk), Kūfa, Baghdād, Basra (these last three in ‘Irāk), al-Ahwāz (in Khūzistān), Fārs and Kirmān, arriving in Sind, India and China. [Fourth itinerary] Sometimes they take a route to the rear of Rome (Rūmiya), through the land of the Slavs (al-Sakāliba), reaching Khamlīdj (or Khamlīkh), capital of the Khazars, then [they sail] on the sea of Djurdjān (= the Caspian), [arriving] at Balkh (in northern Afghanistan) and in Transoxiana (Mā warāʾ al-Nahr) before attaining the camp (wurt) of the Toghuzghuz and moving into China.215 There were therefore sea, river, but also land routes (see map). Several questions have been raised about these Radhaniyya: their origin, their name, their routes, their merchandise, their raison d’être, and even whether they really existed. I will not investigate whether these routes corresponded to the reality. But I shall take on the task of comparing these routes to other information. Ibn Khordādhbih’s description is situated within an Arab geographical perspective, especially since he was writing in Iraq. Nevertheless, it reveals international commercial traffic, with the slaves originating in central and eastern Europe (the Balkans), whereas the markets were located in the southeast. Yet these traders did not travel directly from the northwest to the southeast but made long and arduous detours in conveying the slaves.216

215   Ency. of Isam,” 8:363–364, s.v. “Al-Rādhāniyya” (C. Pellat). 216  I leave aside the question of which was the most convenient conveyance for slaves, boats or caravans: Gieysztor, “Les Juifs et leurs activités économiques en Europe orientale,” p. 506; J. Ferluga, “Mercati e mercanti fra mar Nero e Adriatico: Il commercio nei Balcani dal VII all’XI secolo,” pp. 452–455. In any case, the Radhaniyya used both modes

Medieval Slavery In A New Geopolitical Space

645

The Position of the Byzantine Empire The routes retraced by Ibn Khordādhbih indicate that Constantinople was among the destinations of the Radhaniyya traders, but only as a spice market. The other routes, by contrast, including those for slaves, did not pass through the Byzantine Empire, even though a slave market existed in Constantinople.217 To the east (Balkans-Caucasus-Caspian Sea) and to the west (a maritime route from the Mediterranean coast of present-day France to Antioch, which in the ninth century was under Arab domination), they circumvented Byzantium. Yet a single glance at the map shows that Byzantium was exactly between the sources of slaves and the slave markets. Why did these routes not pass through that empire? Unlike the other ninth-century states, the Byzantine Empire had enjoyed political continuity since the Roman era. And that included a central economic authority located in Constantinople from the fourth century on. The Byzantine market radically changed in the seventh century, after the loss of the wealthiest provinces. Any consideration of the question of Byzantine routes must take into account the new geopolitical reality to which Byzantium had to adapt. That adaptation took place between the late seventh and the mid-ninth centuries, an era prior to Ibn Khordādhbih’s description. The Byzantine authority’s control over the circulation of goods in the seventh to eighth centuries was manifested above all in the kommerkiarioi, the officials responsible for the collection of taxes on goods, as attested by their seals, which date back to the reign of Justinian II. That control was also exerted through the imposition of taxes (kommerkion) and special bonds or fees (apothēkai) on the circulation of goods. This subject has been studied by Hélène Antoniadis-Bibicou, Michael F. Hendy, and Nicolas Oikonomides.218 They all agree that through their system of bonds and customs duties, the Byzantine authorities controlled not only the circulation of goods within the empire, since the kommerkion was actually a tax on circulation and not on sale, of transport. Note as well that the waterways were not accessible all year long and that it is therefore probable that the traders traveled by land during the winter. 217   Patria Const. 2.64. For the Byzantine spice market, see Patlagean, “Byzance et les marchés du grand commerce,” pp. 596–597. J. Koder (“Maritime Trade and the Food Supply for Constantinople in the Middle Ages”) does not mention spices as being among the goods that the Byzantine ships brought to Constantinople. 218  H. Antoniadis-Bibicou, Recherches sur les douanes à Byzance. L’“octava,” le “kommerkion” et les commerciaires, pp. 157ff.; M. Hendy, Studies in the Byzantine Monetary Economy, c. 300–1450, pp. 630–634; N. Oikonomides, “Silk Trade and Production in Byzantium from the Sixth to the Ninth Century: The Seals of Kommerkiarioi”; N. Oikonomides, “Le kommerkion d’Abydos, Thessalonique et le commerce bulgare au IXe siècle.”

646

Rotman

but also, by virtue of the location of the Byzantine ports, international circulation between the Black Sea, the Aegean Sea, and the Adriatic Sea. There was another aspect to that control: legal regulation of the private transfer of goods. That regulation of commercial maritime transport has come down to us in the Rhodian Sea Law, a collection of rules dating from the seventh to eighth century.219 This source clearly attests to the circulation of slaves by sea. It regulates, among other things, the captain’s responsibility for the value of the slaves entrusted to him to be transported from one trader to another. That responsibility applies even if the slaves escape, a risk that distinguishes slaves from any other merchandise.220 The seals discovered in seven provinces of Asia Minor, dating from between 694 and 695, attest to a major sale of Slavic slaves (andrapoda; sklabōn in the Bithynian seals) conducted by the kommerkiarios George, apo khupaton under the reign of Justinian II.221 Oikonomides interprets this evidence within the political context: the slave sale was a punishment inflicted on the families of Slavic soldiers established in Asia Minor who deserted to join the Arab camp.222 According to him, the seals attest that the kommerkiarioi, being close to power, also enjoyed privileges, such as a monopoly on that type of sale. Hendy’s interpretation is entirely different: according to him, these seals attest to the role played by the kommerkiarioi in supplying soldiers, who were Slavic war captives established in Asia Minor.223 In any case, these seals are not the only evidence of the empire’s control over the slave trade. I would like to insist here on the importance of the role of Byzantine customs duties. In 801 Empress Irene lowered taxes on imported goods to appease public opinion, which was unfavorable following her coup d’état of 797 against her son Constantine VI, whom she arranged to have blinded.224 The tax cut led 219   Leges navales; Lex rhodia, ed. W. Ashburner. Its linguistic kinship to the Ekloga has been widely noted. For dating, see Van der Wal and Lokin, Historiae Iuris Graeco-romani delineatio, p. 73; D. Letsios, “Sea Trade as Illustrated in the ‘Rhodian Sea Law’ with Special Reference to the Reception of Its Norms in the Arabic Ecloga.” See also Koder, “Maritime Trade and the Food Supply,” p. 115 and n. 37. 220   Leges navales 15. 221  Published in Oikonomides, “Silk Trade and Production in Byzantium,” pp. 51–53 and nn. 109–116. 222  Ibid., pp. 51–53 (appendix 2: “A Giant Sale of Slaves in 694/95”). 223  Hendy, Studies in the Byzantine Monetary Economy, pp. 630–634. A new seal has been published by S. Bendall (“Slaves or Soldiers?”). According to Bendall, that discovery, dating to 693–694 (close to the time of the Byzantine defeat in 692–693), tends to confirm Oikonomides’s interpretation. 224  Theoph. A. M. 6293 (p. 475).

Medieval Slavery In A New Geopolitical Space

647

to lower prices in Constantinople.225 Nicephorus I restored the taxes, adding a special tax on the circulation of slaves in 809.226 In accordance with that special tax, slaves who did not pass through customs at Abydos would be taxed at the rate of two nomismata apiece (10 percent of the average price of a slave). Theophanes explains that this measure particularly targeted slaves who passed through the Dodecanese. Either these slaves, intended for the markets of Constantinople, arrived through the Dodecanese so that the traders could avoid paying the tax (that hypothesis does not explain why slaves are the only goods mentioned, however), or these slaves were not intended for these markets. I interpret this to mean that the Radhaniyya took the detour to avoid Byzantine customs. The same author, Ibn Khordādhbih, writes that the Russian merchants heading from the Black Sea to the Mediterranean were obliged to pay a tithe to the Byzantine authorities.227 The tithe on the contents of a slave ship was no negligible sum, and the Radhaniyya therefore preferred to stop at Antioch, the port that in the ninth century was closest to the southern border of Byzantium, rather than pass through Constantinople, the largest seaport in that part of the world, which was also a major consumer of slaves.228 Another mention of the influence of Byzantine taxes on the Arab market is found in a text written in about 840 by Abū ‘Ubaid al-Qāsim ibn Sallām, quoted by Hamilton Gibb. This text reports fiscal regulations, supposedly those of Caliph ‘Umar I (634–644), in accordance with which Muslim traders paid 2.5 percent on their merchandise, whereas the dhimmī traders paid 5 percent, and foreigners (identified a few lines later as rūm) paid 10 percent “because they were taking the same percentage from the Muslim merchants when these entered their territory.”229 But these regulations probably reflect a reality subsequent to ‘Umar I’s time. In fact, a treaty of ʿUmar II (717–720), also translated with commentary by Gibb, specifically bans taxes on trade.230

225  C. Mango and R. Scott, The Chronicle of Theophanes Confessor: Byzantine and Near Eastern History A.D. 284–813 (pp. 653–654 n. 4); and N. Oikonomides, “Le kommerkion d’Abydos, Thessalonique et le commerce bulgare au IXe siècle” (p. 242). Both works mention the letter from Theodore Studites to Irene (Theodori Studitae Epistulae 7). This letter provides many details on these measures. Theodore Studites speaks of the taxes on The Straits that navigators from the east, west, and north had until then been “throttled” into paying. Oikonomides interprets these taxes as kommerkia. 226  Theoph. A. M. 6302 (p. 486). 227  Ibn Khordādhbih, Kitāb al-Masālik wa’l-mamālik, p. 155 (130). 228  The tenth-century slave market of Constantinople is described in the Patria Const. 2.64. 229  H. Gibb,” Arab-Byzantine Relations under the Umayyad Caliphate,” p. 230. 230  Gibb, “The Fiscal Rescript of ‘Umar II,” p. 7 (xiv).

648

Rotman

The customs system in seventh- and eighth-century Byzantium allowed the empire not only to control the commercial routes, a control that changed form in the tenth century, but also to preserve something of a monopoly on the slave trade. Patlagean shows how that centralization changed during the reign of Theophilus, or more precisely, in 831–832, when a new type of seal appeared: “The bearer was called imperial or general kommerkiarios of a thema,231 or more often of a city, and above all of a port. In other words, the concessionaire of the public monopoly on silk imports became a collector of customs duties, who had nothing more to do with silk, since he now had general jurisdiction over foreign trade.”232 Oikonomides examines that change within the context of the kommerkion in tenth-century Thessalonica, arguing it was a decentralization strategy; it would culminate in the edict of 992 targeting the Venetians.233 But what were the consequences of the Byzantine economic strategy on the international slave trade? At first glance, the map I constructed reveals that the principal source of slaves at the time, eastern Europe, was, from the vantage point of the Arab world, the hinterland of Byzantium. That region of the Balkans, where the political situation was unsettled from late antiquity on, acquired an economically strategic position with respect to slavery in the era under study. There were internal Balkan routes for the Byzantine market.234 For the international market, Byzantium’s customs system functioned as an obstruction intended to preserve its commercial hegemony. This system also had consequences in the other three regions through which the slave merchants might pass: Caucasus, Italy, and western Europe.

231  The organization of the empire into themata, military and administrative units whose origin dates back to the seventh century (and perhaps even to the reign of Emperor Maurice, between 582 and 602), to a certain extent replaced the old system of Roman provinces. 232  Patlagean, “Byzance et les marchés du grand commerce,” pp. 620–621, which follows Oikonomides’s analysis (“Silk Trade and Production in Byzantium”) of the apothēkai concessions. 233  Oikonomides, “Le kommerkion d’Abydos”; Oikonomides, “Le marchand byzantin.” 234  J. Ferluga, “Der byzantinische Handel auf der Balkanhalbinsel vom VII. bis zum Anfang des XIII. Jahrhunderts”; Ferluga, “Mercati et mercanti fra mar Nero e Adriatico”; A. Gieysztor, “Les marchés et les marchandises entre le Danube et la Volga aux VIIIe–XIe siècles,” pp. 513–514. See the route reconstructed from the numismatic evidence in McCormick, Origins of the European Economy, pp. 375–379, and the routes between central Europe and Constantinople in the ninth century (pp. 548ff.). The journey between Thessalonica and Belgrade took eight days (Const. Porph., De adminis. 42).

Medieval Slavery In A New Geopolitical Space

649

The eastern boundary that separated the Arab world from the Slavic territories extended from the Black Sea to the Caspian Sea. Three rivers flowed north to south: the Dnieper, the Don, and the Volga. Trade among the Russians, Bulgars on the Volga, and Arabs, attested in the eighth century, intensified in the ninth to tenth centuries.235 For these two centuries, the Khazars controlled the access of tradespeople on all sides, thanks to the Khazars’ position between the eastern bank of the Black Sea and the western bank of the Caspian Sea. Their capital city of Itil at the mouth of the Volga, with its large fairs, attracted merchants from every region.236 The merchants who passed through Itil (that is, on the Volga-Caspian Sea route) had to pay a tithe to the Khazar princes.237 In the ninth century, Byzantine control over the Black Sea took the form of a tax to be paid at the port of Trebizond, as attested in the Life of George of Amastris by Ignatius the Deacon.238 That control intensified when the thema of Kherson was founded in 841,239 and traffic between the Russians and the Arabs was routed to the east as a result. In the Khazar region of Caucasus, the penetration of the Jews began in about the mid-seventh century, according to Constantin Zuckerman.240 It intensified following the establishment of a dynasty of Jewish kings in the royal Khazar

235  Espéronnier, “Les échanges commerciaux entre le monde musulman et les pays slaves.” The most important information dates from the ninth to tenth century. See T. S. Noonan, “Volga Bulgharia’s Tenth-century Trade with Samanid Central Asia.” 236  Ibn Fadlān, Voyage chez les Bulgares de la Volga, pp. 66–71. M. Espéronnier, “Villes et commerce: La Khazarie et la Bulgarie de la Volga, d’après les textes arabes et persans des IXe et Xe siècles,” pp. 413–416. Espéronnier also points out the commercial role of the other two Khazar cities, Balandjar and Samandar. The author notes that the sources speak of foreign (Arab and Russian) merchants who passed through these cities, rather than of Khazar merchants. 237  Ibn Khordādhbih, Kitāb al-Masālik wa’l-mamālik, p. 155 (130); Ihn Hawqal, Kitāb Sūrat al-ard, BGA ed., p. 392; Ibn Fadlān, Voyage chez les bulgares de la Volga, p. 71. For an overview, see T. S. Noonan, “Les Khazars et le commerce oriental.” For the relations between the Khazars and the Arabs based on the numismatic and historiographical evidence, see Noonan, “Khazaria as an Intermediary between Islam and Eastern Europe.” 238  In “Byzance et les marchés du grand commerce” (pp. 603–604), Patlagesn explains the commercial position of Trebizond; Life of George of Amastris 33. 239  In “Short Notes: Two Notes on the Early History of the Thema of Cherson” (pp. 210–215), C. Zuckerman dates it to 841 rather than 837 (Dölger, Reg. 1, no. 431). 240  After the fall of the Sassanid Empire; C. Zuckerman, “On the Date of the Khazars’ Conversion to Judaism and the Chronology of the Kings of the Rus Oleg and Igor,” p. 241.

650

Rotman

court in 861.241 The founding of that dynasty expressed the Khazars’ aspiration to remain independent of the two religious blocs surrounding them.242 Farther to the east, the Samanids settled in Bukhara in 874. Their first coins, discovered in Kiev in recent times, date back to the reign of Ismā‘īl bn. Ahmad (892–907), thus indicating relations between the Russians and the Muslims of Khorasan.243 In the ninth century, Byzantium reacted to that unfavorable situation in the Black Sea with Christian diplomacy, exploiting the fact that the “Russians” of Tauric Chersonese were still pagan. The first Christian mission that the Byzantines sent to them dates from 860–861, when Photios was the patriarch of Constantinople.244 Although the ships of Arab merchants did not pass through the waters of the Byzantine Empire near the Balkan coasts and Asia Minor, the ports of Italy were open to them. Byzantine slave traffic is very well attested in Italy, especially for the tenth century. As regards the eighth to ninth centuries, the letter from Pope Adrian I to Charlemagne in 776 had to do with the competition between the Byzantines (“Greeks”) and Arabs (“Saracens”) for the slaves sold in Italy (the mancipia; their origin is not indicated).245 These Byzantines were maritime merchants. The letter of Adrian I consolidated the hegemony of the Byzantine traders by banning the sale of slaves to the Arab ships. Recall that at that date Rome was part of the Byzantine Empire. It is therefore possible that this letter belongs to the context of Byzantine political policies, especially since no religious motive is advanced in it.246 241  Ibid., pp. 241–250; P. B. Golden, “Khazaria and Judaism.” 242  It also set them up as a third empire (D. Shapira, “Two Names of the First Khazar Jewish Beg”). 243  G. Ivakin, “Kiev aux VIIIe–Xe siècles,” p. 232; Noonan, “Khazaria as an Intermediary between Islam and Eastern Europe,” pp. 202–204. Noonan points to the economic growth of the Bulgars on the Volga resulting from the trade with the Samanids (“Volga Bulgharia’s Tenth-century Trade”) that accompanied their Islamization (see Shapira, “Two Names of the First Khazar Jewish Beg,” p. 235). For the Khazar economy, see Noonan, “What Does Historical Numismatics Suggest about the History of Khazaria in the Ninth Century?”. 244  Before Kiev was conquered by the Russians (Ahrweiler, “Les relations entre les Byzantins et les Russes”). Byzantium imposed the same policy in Crimea during the same period, with the agreement of the Khazars (Noonan, “The Khazar-Byzantine World of the Crimea in the Early Middle Ages”). 245  MGH, Epistolae, vol. 8; Codex Carolinus, no. 59 (pp. 584–585). For the presence of Arab traders in Carolingian Europe, see M. McCormick, “Voyageurs, monnaies et esclaves,” especially the map of the diffusion of Arab and Byzantine coins (p. 44). 246  For the importance of the Byzantium-West routes in general, see M. McCormick, “Byzantium on the Move: Imagining a Communications History.”

Medieval Slavery In A New Geopolitical Space

651

In Italy, the Byzantine authorities tried to maintain control over the northsouth commercial routes, even when the Franks made inroads between the late eighth and early ninth centuries.247 To assert their hegemony, the Byzantines made use of Venice. Hence in 814–820 Leo V and the dux of Venice tried to prohibit Venetian merchants from engaging in trade with the Arabs, and in 876 Dux Urso Participacio I banned the purchase of slaves from pirates and the transport of slave merchants.248 Byzantine hegemony, which remained strong in Italy, did not prevent Adrian I from turning to Charlemagne at a time prior to the Franks’ incursions into Italy. In fact, the aforementioned letter was part of a correspondence that aimed to restore a political relationship between the pope and the Frankish king.249 In any case, the Franks had other maritime routes of access: the Mediterranean ports under their domination; the commercial routes that passed through the Frankish territories of western Europe, especially after the victory of Charles Martel and the Frankish conquest of the northern Iberian peninsula that followed; and, in the ninth century, routes in central Europe as well. It was precisely from these ports that the Radhaniyya embarked.250 The Radhaniyya were not the only Jewish merchants who specialized in the slave trade. In two articles, Alexandre Gieysztor sets out the routes of Jewish merchants generally in central Europe, Germany, Moravia, Hungary, and Poland.251 Their presence is attested in the Raffelstetten Customs Regulations of 903–906, a document that specifies the total duties on the merchandise

247  The Franks arrived at the Adriatic Sea but were obliged to leave in 811–812 (Const. Porph. De adminis., 28; Dölger, Reg. 1, no. 385). Nevertheless, they maintained indirect contact through their Lombardian route, and, like the Byzantines, they made use of Venice (McCormick, Origins of the European Economy, pp. 759–767). 248  Tafel-Thomas 1:3 (no. 3), p. 5 (no. 7). 249  E. Patlagean, “Variations impériales sur le thème romain,” pp. 38ff. 250  Ibn Khordādhbih mentions Firanja as the Radhaniyya’s starting point. Gil (“The Radhanite Merchants and the Land of Radhan,” pp. 310ff.) believes that Firanja was Lombardy. McCormick (Origins of the European Economy, p. 691) thinks the port was Venice (“New Light on the ‘Dark Ages’”). Because Ibn Khordādhbih also mentions Firanja as the starting point for the land route, and because his description specifically includes the “king of Firanja,” I believe rather that it designates the “country of the Franks.” In fact, Jewish merchants bringing slaves to Arles on the Rhone are also mentioned by Agobard of Lyons (Agobardi Lugdunensis Opera omnia, no. 6). 251  Gieysztor, “Les juifs et leurs activités économiques en Europe orientale”; Gieysztor, “Les marchés et les marchandises entre le Danube et la Volga.”

652

Rotman

required by customs.252 Charles Verlinden has attempted to reconstitute the Radhaniyya’s tenth-century European routes.253 All this evidence is for the tenth century and clearly identifies the routes connecting northeastern and southwestern Europe. An important commercial stop on that route was Verdun, which Liutprand of Cremona mentions in 986 as a center for castrating slaves on the way to Spain.254 During the same period, al-Muqaddasī mentions Slavic slaves (saqaliba) who were castrated in Spain, then sent to Egypt.255 The tenth-century Responsa by the geonim speak of Jewish merchants who transported castrated slaves.256 The routes that passed through what is now France led to the Spain of the Umayyads, which provided large markets for saqaliba. But other routes took the slaves to Venice, as is apparent in the Life of Nahum, which describes the disciples of Methodius abducted in Moravia and sold as slaves in Venice.257 The importance of Jewish merchants in the traffic between present-day France and Spain is attested in the mid-ninth century by the commercial privileges regarding the slave trade that Louis the Pious granted a Jewish merchant of Saragossa and two Jews from Lyons (825) and by the writings of Agobard of Lyons in 846 against the Jewish traffic in Christian slaves in France.258 According to this information, Lyons and Arles served as stops on the route

252  MGH, Capitularia Regum Francorum, ed. A. Boretius and V. Krause, vol. 2, no. 253 (pp. 249–252); H. Dopsch, “Raffelstettener Zollordung,” in Lexikon des Mittelalters 7:397. Raffelstetten is near present-day Linz, which attests to the importance of the route on the Danube between eastern and western Europe (for this route, see also McCormick, Origins of the European Economy, pp. 553–557). 253  C. Verlinden, “La traite des esclaves. Un grand commerce international au Xe siècle.” 254  “Antapodosis” 6.6, in Liudprandi Cremonensis Opera omnia. He describes the castration of slaves by traders who made a very large profit on them (“Verdunenses mercatores ob immensum lucrum facere et in Hispaniam ducere solent”). 255  Al-Muqaddasī, Description de l’Occident musulman au IV e–Xe siècle, p. 56. 256  Assaf, “Slaves and Slave Trade among the Jews in the Middle Ages,” pp. 100–102 and nn. 62–63. I should note that the Jewish prohibition on castration (ibid.; Assaf, The Responsa of the Geonim in the Manuscripts of the Genizah, p. 192) explains why Jewish merchants needed intermediaries to perform it. Nevertheless, it is not known who performed castration in Verdun and why it was done there. 257   Life of Nahum, pp. 287–288. See the route connecting central Europe to the Adriatic described by McCormick, Origins of the European Economy, pp. 369ff. (“The Amber Trail”). 258  A letter sent to Louis the Pious; Agobardi Lugdunensis Opera omnia, no. 11 (pp. 189–195), “De insolentia Iudeorum (ad Ludovicum).” For the Jewish merchants who received privileges in France, see Ashtor, “Aperçus sur les Radhanites,” pp. 250–251.

Medieval Slavery In A New Geopolitical Space

653

taking the Jewish slave merchants to Spain, and more precisely to Cordova. These were always land and river routes.259 The efforts in the ninth century to prevent Arab slave merchants from reaching the ports of western Europe were reinforced by the promulgation of Byzantine and Venetian decrees. The intention of these decrees was primarily political: to limit the grant of aid of all kinds, which the Arab forces had until that time received in wartime. But the decrees also had consequences for trade because they worked to the great advantage of Jewish merchants. Recall that, according to Ibn Khordādhbih, the Radhaniyya also traded in weapons. As early as 593, Gregory I had protested the sale of Christian slaves to Jewish merchants.260 In a letter to Libertinus, praetor of Sicily, Gregory I specifically stated his opposition to a Jewish slave merchant named Nasas who was purchasing Christians. But his letter did not target non-Christian slaves, which explains why eastern Europe (especially the Slavic regions) was before its Christianization the principal source of slaves in Europe.261 Christians also needed these Jewish merchants as intermediaries, which allowed them to take advantage of the high demand from the Arab markets in Spain. That commercial front, however, was outside the Byzantine influence. I return, therefore, to the eastern Mediterranean, to consider from the Byzantine point of view the map of international slave trade in the first third of the ninth century. The starting point for the north-south slave traffic was between the Dnieper River and the Adriatic Sea. Byzantium, which extended from these regions to the Mediterranean, in fact blocked maritime traffic between the Balkans and the Arab world from Spain to Iraq, at least where the slave trade was concerned. Byzantium was the principal consumer market for Slavic slaves outside the Arab world. The Byzantines were therefore in competition with the Arabs for that type of merchandise. 259  As for the sea routes linking Muslim Spain to northern Europe, al-Ghzāl (Yahyā bn. Hakam al-Bakrī), the emir’s envoy from Cordova to Constantinople between 822 and 852, describes the sea routes of Arab merchants who circumvented western Europe in order to disembark in the ports of the Baltic Sea. In “Les échanges commerciaux entre le monde musulman et les pays slaves” (p. 20 n. 18), Espéronnier cites the account given by R. Dozy, Recherches sur l’histoire et la littérature d’Espagne pendant le Moyen Âge (Paris and Leiden, 1881), 2:267–278. That description, along with the circumstances presented (al-Ghazāl’s embassy to the Normans) is probably imaginary (Ency. of Islam 2:1038, s. v. “Al-Ghazāl, Yahyā b. Hakam al-Bakrī,” [A. Huic Miranda]). 260   Gregorii I Papae Registrum epistolarum, Libri I–VII, ed. P. Ewald and L. M. Hartmann, MGH, Epistolate I, 3.37. 261  See the writings of Agobard of Lyons on this subject: Agobardi Lugdunensis Opera omnia, no. 6 (“De baptismo maniciporum Iudaeorum”).

654

Rotman

The Arab world, then, imported Slavic slaves through routes that circumvented the Byzantine Empire; to the east through Caucasus and the Caspian Sea; to the west by the routes leading from Raffelstetten to Spain via France.262 The eastern and western parts of the Arab world (Iraq and Spain) thereby procured white slaves. I have cited the customs system of Byzantium to explain the detours taken by the slave merchants. But they also faced taxes in Raffelstetten and Itil. How, therefore, to justify these long detours? I think this is quite simply a matter of dates. The Byzantine ports were already blocking slave traffic in the early ninth century, whereas taxes in Raffelstetten and Itil are attested for the tenth century.263 In other words, Khazar kings and Bavarian princes were practicing in the tenth century what Byzantium had already been practicing for at least a hundred years.264 Iraq and Spain can be considered two different regions from a commercial standpoint, but a third region under Arab domination must also be taken into account: the eastern and central Mediterranean, that is, Syria, Palestine, Egypt, and Ifriqiya. According to Ibn Khordādhbih’s description, the Radhaniyya passed through these four regions. But to arrive there, they made a long detour, which necessarily raised the price of slaves. As a result, there was not an adequate commercial supply to meet the demand for slaves in these regions (the price of an ordinary slave in tenth-century Egypt was 33 dinars, versus 20 nomismata in Byzantium). It was at that moment, the ninth to tenth century, that Arab piracy came into play. If it was profitable, that alternative for supplying slaves, risky in nature, must have been responding to a sufficiently large market demand in the region, given that the price I mentioned was that for victims of piracy. In this regard, let me point out that there are few mentions extant of Byzantine women being ransomed. The small number of references tallies with the price of female slaves in Arab countries, which was greater than that for male slaves: and in the case of female singers, the price could go even higher.265 262  Merchants of Bavarian (Bawari) and Slavic (Sclavi) slaves were exempted from the tax of one-twelfth the price of each slave at Raffelstetten. 263  Ibn Hawqal and Ibn Fadlān wrote in the second half of the tenth century. Carolingian toll stations are attested for the ninth century (McCormick, Origins of the European Economy, pp. 640–652), but without any mention of the special tax on slaves. 264  It is clear that the importation of slaves was subject to the kommerkia before Irene’s tax exemptions and Nicephorus I’s special tax. 265  The document Ms. Adler 2804 verso (Mann I, no. 14, 1:88–39), conserved in the Cairo Genizah, attests to the redemption of a Byzantine Jewish woman. For the prices of slaves in the Arab countries, see Y. Ragib, “Les marchés aux esclaves en terre d’Islam,” pp. 757ff.;

Medieval Slavery In A New Geopolitical Space

655

In the ninth and tenth centuries, Arab piracy threatened all the coastal zones of the eastern Mediterranean, from Egypt to southern Italy and the Adriatic Sea. But not all the pirates were Arabs. Byzantines participated as well, landing on the coasts of Egypt, Syria, and Palestine. Nevertheless, it seems that Byzantine piracy did not reach the same scope as Arab piracy, which covered the entire Aegean Sea, the Ionian Sea, Sicily, and southern Italy. In addition, throughout the ninth century Arab forces from Ifriqiya endeavored to conquer Sicily and southern Italy: Palermo was conquered in 831, Syracuse in 878, and in 872–875 the Arabs reached Dalmatia. The booty from the attacks by Arab pirates arrived at several destinations, usually Egypt and Ifriqiya. Hence Byzantines who were kidnapped by Arab pirates were put up for sale on Arab markets. That phenomenon is attested by the Byzantine hagiographical narratives; the documents of the Genizah; and Arab writers such as Ibn Butlān, who describes the advantages of slaves from every background. The same author mentions Armenian slaves as well.266 I have already spoken of Arab piracy with respect to the slavery of war captives. Here I shall confine myself to pointing out once again that this Arab threat became an everyday reality on the Byzantine coast, and even in the hinterland, beginning in 827–828, when Arab forces from Spain conquered Crete. Crete, located between Cyprus and Sicily and between Egypt and the Aegean Sea, of which it marks the boundary, is by virtue of its position the most strategic site in the eastern Mediterranean.267 The Rhodian Sea Law, though not in the nature of an imperial promulgation, was very important for the commercial traffic in the Byzantine seas around the late seventh to eighth centuries. In fact, that source says nothing about the routes and little about the merchandise, coinciding in that respect with the legal sources of the same period. The Rhodian Sea Law focuses on the transfer and maintenance of goods in the boats, and on the responsibility for these goods. Nevertheless, it is possible to conclude that the Byzantine maritime market was very open at the time, despite the pirate attacks, which are

A. Cheikh-Moussa, “Figures de l’esclave chanteuse à l’époque abbasside,” pp. 45–60; E. Ashtor, Histoire des prix et des salaires dans l’Orient médiéval, pp. 58–59. 266  Mez, The Renaissance of Islam, pp. 160–163. For the presence of Armenians in the Mediterranean, see Life of Lazarus of Mount Galesios 9, quoted in Patlagean, “Byzance et les marchés du grand commerce,” p. 598. 267  For a comparison with the position of Cyprus, see R. J. H. Jenkins, “Cyprus between Byzantium and Islam, A.D. 688–965”; and A. Cameron, “Cyprus at the time of the Arab Conquests.”

656

Rotman

mentioned in the text.268 Although the boats were Byzantine, the merchants were not necessarily so. The commercial routes passed through the islands where the ships made stops,269 which shows the importance of the islands for maritime trade, especially at a time when pirates were on the attack. The expatriate Arabs of Spain who conquered Crete in 827 inherited not only a base for conducting raids, which Crete had already been for the Byzantines, but also an island that dominated maritime trade in the eastern Mediterranean.270 That became obvious in the first half of the tenth century, when Muhammad ibn ‘Umar composed the Faqīh, a text whose kinship to the Byzantine Rhodian Sea Law suggests that the latter was its source.271 The Faqīh was written by a Muslim. Yet the common points between these two sources clearly show that religion had little importance for commercial law. Nevertheless, unlike the Byzantine collection, which has no precise geographical scope (its title is only an allusion to the ancient sea law of Rhodes),272 the Faqīh was composed in Crete by an Arab Cretan. That source therefore reflects the commercial importance of Crete, whose benefit for the local Arab authorities is clear. Like the Rhodian Sea Law, the Faqīh mentions slaves as merchandise rather than as part of the boat crew, but without indicating their provenance or their destination.273 Like Byzantium, Arab Crete did not only block the southern Mediterranean; it also deprived the empire of the use of the Aegean Sea as a domestic waterway.274 That led to the evacuation of the Aegean isles and even of the coast of Peloponnesus in the tenth century. It was only in 961 that Nicephorus II 268   Leges navales 15. 269  Ibid. 270  V. Christides. The Conquest of Crete by the Arabs (ca. 834): A Turning Point in the Struggles between Byzantium and Islam. 271  V. Christides, “Raid and Trade in the Eastern Mediterranean: A Treatise by Muhammad b. ‘Umar, the Faqih from Occupied Moslem Crete, and the Rhodian Sea Law, Two Parallel Texts.” See the exhaustive study of Hassan S. Khalilieh, Islamic Maritime Law. An Introduction. 272  For the Roman Lex Rhodia, see Dig. 14.2 (“De lege Rhodia de iactu”). For the evolution of Roman sea law and its Greek origins, see J. Rougé, Recherches sur l’organisation du commerce maritime en Méditerranée sous l’Empire romain, pp. 397–413. For maritime loans in Roman law, see J. Andreau, Banque et affaires dans le monde romain, IVe siècle av. J. C.– IIIe siècle apr. J.-C., pp. 108–112. 273  Christides, “Raid and Trade in the Eastern Mediterranean,” pp. 77–78. Hassan S. Khalilieh, Islamic Maritime Law. An Introduction, pp. 101–102. For the crew and the dimensions of the boats, sec McCormick, Origins of the European Economy, pp. 402–430. 274  E. Malamut, Les iles de l’Empire byzantin, VIIIe–XIIe siècles, 2:535ff.

Medieval Slavery In A New Geopolitical Space

657

Phocas succeeded in recovering that Mediterranean base. With that conquest, the Aegean Sea once again became Byzantine. In 965, the same emperor managed to recapture Cyprus and Cilicia; Antioch followed in 969.

The New Politico-Commercial Strategy of Byzantium in the Tenth Century The conquest of Crete in 961 was part of the great military plan of Nicephorus II Phocas, who reconquered the southern part of Asia Minor for the Byzantine Empire, as well as Cyprus and northern Syria. In addition, the city of Antioch, the Syrian port of such importance for the Radhaniyya, also became Byzantine once again. The maritime dominance of Byzantium in the eastern Mediterranean was thus restored in the tenth century. As Ibn Hawqal notes, Byzantine ships easily traveled to Palestine and Egypt.275 Let me add, to the examples provided by Patlagean,276 the cases found in hagiographical narratives of Byzantines captured by Arab pirates and sold to Arab countries, especially Egypt and Ifriqiya. The Byzantines who pursued them to ransom the captives embarked from Arab ports (Arab piracy continued well after 961), without fear of being captured themselves. For example, in the Spiritually Beneficial Tales of Paul of Monembasia, the monk who travels to Ifriqiya to redeem three brothers belonging to his monastery visits the country’s governor with no risk to his freedom.277 The Byzantine ships that stopped in Arab countries established commercial ties between the Byzantine and Arab ports, as the two journeys reported in the Life of Nicholas of Sion (tenth century) attest. In the first, Nicholas leaves Lycia on the boat of a man named Menas and arrives in Askhelon after five days.278 The ship is coming from Palestine but belongs to a Greek. The second journey, which was supposed to be a pilgrimage to Jerusalem, takes him to Egypt.279 Then there is the correspondence that the Jewish communities of Egypt maintained with their coreligionists in the Byzantine Empire, especially in the tenth century, as attested in the documents of the Genizah.280 275  Ibn Hawqal, Kitāb Sūrat al-ard, BGA ed., p. 205, trans. J. H. Kramers and G. Wiet (Configurations de la terre), 1:19. 276  Patlagean, “Byzance et les marchés du grand commerce,” pp. 611–612. Patlagean shows how that maritime commerce allowed Byzantium to become a strong presence in the eastern Mediterranean. 277   Spiritually Beneficial Tales of Paul of Monembasia 8. 278   Life of Nicholas of Sion 8–9. 279  Ibid., 29–32. 280   T.-S. 16.251 (Mann I, no. 18, 1:92–93), a letter from the Jewish community of Alexandria to the community of Mastaura (in Lydia), asking for money to ransom its members abducted by Arab pirates and sold to Alexandria. The Ketubah, composed in Mastaura in 1022, is

658

Rotman

And it was precisely in that century that the decline of Jewish merchants occurred. The reason may lie in Arab piracy or in the increase of Byzantine commercial channels in the Mediterranean. Ashtor believes that the commercial development of Italian cities began to take over the role previously performed by Jewish merchants.281 But he too overlooks the Byzantine factor at play on the map, a factor that was particularly decisive for Italy. Let me now turn to examine the trade policy of the Byzantine Empire from the early tenth century onward in its other mare domesticum: the Black Sea. Trade between the Russians and the Byzantines in the North Sea is attested in the mid-ninth century by Ibn Khordādhbih who, in his description of the Radhaniyya, devotes a passage to the Russian merchants.282 The three principal north-south commercial routes were riverways: the Dnieper, the Don, and the Volga.283 The first two routes ended at the North Sea, while the third continued on to Itil, the Khazar capital on the mouth of the Volga, which Ibn Hawqal calls “the river of the Russians,” and to the Caspian Sea. As a result, at the time Ibn Khordādhbih was writing (the first version of the Kitāb al-Masālik wa’l-mamālik dates from 846), Russian merchants encountered customhouses on both the Byzantine and the Khazar sides.284 This Russian traffic accounts for the Arab and Byzantine coins found in Kiev (on the Dnieper) and in Gnezdovo (on the Volga), which date from the time of Basil I (867–886), Leo VI (886–912), and the reign of the Samanid king Ismā‘īl bn. Ahmad (892–907). This proves that the Russians needed coins: they began to mint them themselves only in about the mid-tenth century.285 Samanid dinars have been found in the area corresponding to the fourth route of the Radhaniyya who, according to Ibn Khordādhbih’s description, took Slavic slaves behind the Byzantine empire to Transoxiana. The treaties of 907 and 911 between the Byzantines and the Russians inaugurated a new policy in that region. It went hand in hand with the growth of also conserved in the Genizah (T.-S. 16.374, Mann I, no. 19, 1:94–96). See M. T. Reinach, “Un contrat de mariage du temps de Basile le Bulgaroctone”; D. Jacoby, “What Do We Learn about Byzantine Asia Minor?” pp. 84ff. 281  Ashtor, “Gli Ebrei nel commercio mediterraneo.” 282  Ibn Khordādhbih, Kitāb al-Masālik-wa’l-mamālik, p. 154 (130). 283  The first two routes have been studied in J. Shepard, “Constantinople—Gateway to the North: The Russians.” 284  Ibn Khordādhbih, Kitāb al-Masālik wa’l-mamālik, p. 154 (130); in 922, Ibn Fadlan (p. 71) attests to the tithe paid for each slave ship to the king of the Khazars. 285  In the ninth century, before Kiev was conquered by the Russians, it was the Russians of Tauric Chersonese who faced an obstruction in the Black Sea (Ahrweiler, “Les relations entre les Byzantins et les Russes”).

Medieval Slavery In A New Geopolitical Space

659

Russian commercial might in the tenth century. These treaties, as well as that of 944, granted commercial privileges to the Russians, who obtained right of access to Byzantine markets, especially Constantinople.286 No tax is mentioned, and the Byzantines agreed to lodge Russian merchants and to aid the Russian ships against pirate attacks.287 There was also a military alliance, as a result of which the Russians participated in Byzantine expeditions. Russian ships also took part in the Byzantine attack on Arab Crete described by Constantine VII Porphyrogenitus.288 That military alliance, though not permanent, was even more important in the region of the Black Sea.289 In 944 a Russo-Byzantine victory was achieved against the Bulgars. And in 969 the Russians destroyed the kingdom of the Khazars.290 Princess Olga’s conversion to Christianity (about the mid-tenth century) was followed by an official journey to Constantinople, which she undertook in 957.291 It was her son Sviatoslav who in 968 made possible the Byzantine victory over the Bulgars.292 This does not mean that the relations between the Byzantines and the Russians always remained calm.293 In 988 Prince Vladimir allowed Basil II to crush Bardas Phocas’s rebellion. The Christianization of the Russians has traditionally been attributed to the same prince, who wished to marry Basil II’s sister.294 In the second half of the tenth century, the Russians became the

286  All three treaties are described in the Laurentian Chronicle, pp. 65–75, translated into French with commentary by I. Sorlin in “Les traités de Byzance avec la Russie au Xe siècle,” pp. 329–336, 447–452. 287  For foreign merchants and their accommodations in the capital, see Jacoby, “The Byzantine Outsider in Trade,” pp. 132–134. 288  Const. Porph., De cerimon. 2.44. 289  In 944, the Russians again attacked Byzantium, a war that led to the treaty of 944 (Sorlin, “Les traités de Byzance avec la Russie,” pp. 452–455). 290  Ibn Hawqal, Kitāb Sūrat al-ard, BGA ed., p. 392; and, on the same date, the destruction by the Russians of “al-Bulghar.” Was this the principal city of the Volga and Kama Bulgars? 291  Ioannes Scylitzes, Synopsis historiarum, p. 240; C. Zuckerman, “Le voyage d’Olga et la première ambassade espagnole à Constantinople.” 292  For Russo-Byzantine relations in the second half of the tenth century, I have followed A. Poppe, “The Political Background to the Baptism of Rus’: Byzantine-Russian Relations between 986–989.” 293  Consider Sviatoslav’s attacks of 970, which the Byzantines succeeded in repelling. The Russian attack and the Byzantine victory are also described in Life of Basil the Younger (in Moscou gr. 249), fol. 136 verso. 294  Yahyā Ihn Saʿīd Al-Antāk, PO 23, pt. 3:422–424; Poppe, “The Political Background to the Baptism of Rus,’” pp. 205–210.

660

Rotman

second power in the Black Sea and, in 989, they obtained control of Kherson.295 Basil, for his part, proceeded to annex Armenia, and, with the end of the Samanid kingdom in 999, the Russo-Byzantine domination of Caucasus was complete.296 With respect to the slave trade, the Russians figure as suppliers for the Byzantines in the Russo-Byzantine treaties of the first half of the tenth century.297 Slaves are in fact the only merchandise mentioned.298 In De administrando imperio, Constantine VII Porphyrogenitus describes the river traffic in slaves by the Russians.299 The Russian victory over the Khazars also had consequences in that region. According to Ibn Hawqal’s description, the Khazars took Russian and Bulgar captives in their raids conducted on the Black Sea. They then sold them as slaves, probably on Arab markets,300 a description in agreement with that given by Constantine VII concerning the Pechenegs’ raids on the Russians.301 The Russians’ victory freed them from that threat and opened markets to the south for which they no longer needed a customs middleman. But during the first half of the tenth century, Russian dominance in the Black Sea came about at the expense of another neighboring kingdom, that of the Bulgars. I have shown that Byzantium used the treaties it had concluded with the Russians in its international policy during the tenth century primarily against 295  Yahyā Ibn Sa‘īd Al-Antāk, PO 23:432–433; Leonis Diaconi Historiae 10.10. The taking of Kherson by the Russians has traditionally been explained by Basil II’s refusal to give his sister Anna in marriage to Vladimir as promised. Poppe (“The Political Background to the Baptism of Rus,’” pp. 198ff.) argues that “the ‘Cherson problem’ remains the key question in the interpretation of Byzantine-Russian relations around the time of the conversion of Rus’ to Christianity.” He presents a different interpretation of the events, claiming that the Russians’ expedition against Kherson was not directed against Byzantium but rather supported Basil II in his policy against the partisans of Bardas Phocas (pp. 211ff.). At the time, the Russian control of Kherson, which began in 989, was presented as part of Byzantine policy. 296  For the rise and fall of the Samanids, see Ency. of Islam, 8:1025–1031, s.v. “Sāmānids” (C. E. Bosworth). 297   Laurentian Chronicle, pp. 68, 75; Sorlin, “Les traités de Byzance avec la Russie,” pp. 458–459. 298  In the treaty of 944, payment is to be made in silk; Sorlin finds an equivalence between silk and gold in the treaty (“Les traités de Byzance avec la Russie,” pp. 458–459). 299  Const. Porph., De adminis, 9. See I. Sorlin, “Voies commerciales, villes et peuplement de la Rôsia au Xe siècle d’après le De administrando imperio de Constantin Porphyrogénète.” 300  Ibn Hawqal, Kitāb Sūrat al-ard, BGA ed., pp. 392, 394. 301  Const. Porph., De adminis. 4.

Medieval Slavery In A New Geopolitical Space

661

the Bulgars, who were causing disturbances in the empire and who became particularly threatening during the reign of Simeon (893–927). Historians have traditionally seen that war as a reaction against the transfer of Bulgar merchants in 893 from Constantinople to Thessalonica.302 The-ophanes Continuator explains that this measure had been taken by Stylianos Zaoutzes, Leo VI’s minister, in favor of two merchants from Hellas.303 In the tenth century, the Byzantine authorities found adequate allies in the Black Sea: the Russians. The largest market on that sea was opened to the Russians for political and military reasons, as specified in the treaties. The expulsion of the Bulgar merchants from the Constantinople market followed the same logic.304 The shift incited major Bulgar opposition, leading to the wars waged by Simeon, after which the Byzantines were forced to once more accept the Bulgar merchants in Constantinople. The expulsion of the Bulgars, which can be explained by the Byzantine policy in the Black Sea, also gives a glimpse of a commercial and political plan: Byzantium organized the Black Sea and the Aegean Sea by sharing its commercial hegemony in the Black Sea with the Russians and by setting up the Bulgars in the Aegean Sea. Indeed, according to Theophanes Continuator, the motive for their actions was the profit to be made by the Byzantine merchants of Hellas. In fact, Byzantium applied the same policy in the AdriaticIonian Sea. It is clear that Byzantium did everything in its power to assert its commercial hegemony in Italy in the eighth and ninth centuries.305 That proved to be especially true for the slave trade, as indicated by edicts issued from Byzantium and Venice—and from both at the same time—which tried to limit that traffic. Nevertheless, the prohibition, renewed four times in documents (in 876, 945,

302   Theoph. Cant. 6.9 (p. 357); 6.3 (p. 354), following Patlagean, “Byzance et les marchés du grand commerce,” p. 622. 303  Note that, during the same year, Vladimir, the eldest son of Boris Michael, returned to paganism. It was also Leo VI, the signatory of the treaties with the Russians in 907 and 911, who, in about 911–912, promulgated the Book of the Prefect. 304  See also Eparch. bib. 9.6, where the presence of Bulgars (“and other foreigners”) may have led to financial problems for a shop in Constantinople. 305  I have not mentioned, however, the importance of the Dalmatian Islands for navigation in the Adriatic (J. Ferluga, “Les îles dalmates dans l’Empire byzantin”; J. Ferluga, “Navigation et commerce dans l’Adriatique aux VIIe et VIIIe siècle”) or the Byzantine-Venetian response to the arrival of the Arabs in that sea in 872–875, namely, Basil I’s creation of the thema of Dalmatia.

662

Rotman

960, and 902), shows that these attempts failed until the mid-tenth century.306 But why prohibit the Venetian slave traffic? Among the documents published by Gottlieb Lukas Friedrich Tafel and Georg Martin Thomas are prohibitions targeting trade with Arab merchants and Arab countries in general, especially in wartime.307 These acts are consistent with Byzantine policy in the eighth to tenth centuries, which obstructed the slave traffic with Arab countries through the use of customhouses. In Italy, Venetian edicts produced the same result by banning Venetians from slave trade with the Arabs. The Arab markets were the destinations of Italian merchants, especially Venetians.308 But not only were these merchants in competition with the Byzantines, they sometimes sold Byzantines as well.309 The edicts thus had two objectives: to put an end to the competition in the slave traffic destined for the Byzantine markets, and to stop the slave trade in abducted Byzantines. Byzantium had exactly the same preoccupations during the same period on the eastern front. Byzantium controlled the commercial traffic on the Adriatic Sea through its customhouse in Corfu, as attested by Liutprand of Cremona.310 Possibly as a result, the Venetian placitum of 960 banned all slave trade between central and eastern Europe conducted by Venetian ships on the Adriatic Sea. The slave traffic did not necessarily pass through that sea, however. The Radhaniyya, for example, did not come that way, and the Life of Elias Spēlaiōtēs tells us of slave merchants in the villages of Calabria in the second half of the tenth century.311 306  Tafel-Thomas, 1:5 (no. 7); p. 16 (no. 12); p. 17 (no. 13); Pacta Veneta:I trattai con Bisanzio 992–1198, no. 1. 307  Tafel-Thomas, 1:3 (no. 3); pp. 25–30 (no.14). 308  For the new position of Venice at that time, see McCormick, Origins of the European Economy, pp. 523ff. 309   Life of Nahum, pp. 287–288. For the slave merchants passing through Moravia, see Gieysztor, “Les juifs et leurs activités économiques”; Ashtor, “Gli Ebrei nel commercio mediterraneo,” pp. 508–509 (also for the ninth century); and Ashtor, “Aperçus sur les Radhanites,” pp. 260–263. Blaise of Amorion, whose Life dates from the first half of the tenth century, is captured when he returns to Rome and is then sold in Bulgaria (Life of Blaise of Amorion 8; for his route, see Malamut, Sur la route des saints byzantins, pp. 258– 260) (it is still not known who abducted him). The decree of 876 (Tafel-Thomas, 1:5, no. 7) explicitly targeted profit-hungry Venetian merchants who purchased slaves abducted by pirates and bandits in order to market them. 310  Patlagean, “Byzance et les marchés du grand commerce,” p. 622 and n. 121; “Relatio de Legatione Constantinopolitana” 65, Liudprandi Cremonensis Opera omnia 65. 311   Life of Elias Spēlaiōtēs I 18 (Life of Elias Spēlaiōtēs II 12). The saint criticizes the merchant for his activities, but the merchant, who does not repudiate them, finally meets his death.

Medieval Slavery In A New Geopolitical Space

663

Therefore, the Venetian placitum of 960 also banned the sale of slaves, even to Jewish merchants. And the same document mentions as well the obligation to ransom captives.312 Nevertheless, that series of edicts did not succeed in eliminating the Venetian slave traffic between 876 and 960. At that time, Byzantium changed its strategy. Instead of trying to block Venetian merchants with edicts and customs duties, Byzantium now made use of the Venetians by the same means: it used edicts, hence its legal authority, and customs, its fiscal authority. The chrysobull, or imperial edict, of Basil II (992), published and studied by Agostini Pertusi, and more recently by Marco Pozza and Giorgio Ravegnani in Pacta Veneta, reveals that new policy of the Byzantine emperors, which a century later would produce the chrysobull of Alexius Comnenus (1082).313 Basil II granted considerable fiscal privileges to the Venetian merchants passing through customs at Abydos: they now paid seventeen solidi per boat instead of thirty. That gave them a great advantage over all the other non-Byzantine merchants (the text mentions the Amalfitans, the Jews, and the Lombardians of Bari).314 As Pertusi explains, this was not the unilateral concession of a privilege but a bilateral accord that considered the Venetians not “a vassal power of the eastern emperor” but “a sovereign state.”315 He shows that this attitude developed during the tenth century as a result of the Venetian victories along the eastern coast of the Adriatic Sea.316 All the same, the chrysobull explains that the Byzantine emperor could now use the Venetians in his servitium (which must be translated as douleia). In the late tenth century, that implicitly meant that aid would be granted to the emperor in his wars against the Arabs in Sicily. To assure himself of the Venetians’ aid against the Normans in 1082, Alexius Comnenus granted them an exemption from customs duties in all the chief ports of Syria, Asia Minor, the islands, Greece, Epirus, Macedonia, Thrace, and Constantinople. In addition, a commercial sector was granted them in the

312  Tafel-Thomas, 1:17 (no. 13). 313   Pacta Veneta: I trattati con Bisanzio 992–1198, no.1 (of 992, pp. 21–25 n. 2; of 1082, pp. 35–45); A. Pertusi, “La crisoballa del 992.” 314  See the different interpretation of David Jacoby (“The Byzantine Outsider in Trade,” pp. 133–134), who argues that the Byzantines wanted to keep their monopoly over the silk market (for the chrysobull of 1082 and the continuation of that imperial policy, see ibid., pp. 134ff.). See also L. Ralph-Johannes, Handel und Politik. Zwischen dem byzantinischen Reich und den italienischen Kommunen Venedig, Pisa und Genua in der Epoche der Komnenen und der Angeloi (1081–1204), pp. 1–8. 315  In Pertusi’s words, “Venezia e Bisanzio nel secolo XI,” p. 131. 316  Ibid., pp. 127ff.

664

Rotman

capital, consistent with the same policy and in the same manner as for the Russians in the tenth century.317 It is therefore clear how the global slave trade and the definitions of the Byzantine institution of slavery evolved as a function of the changes in the medieval geopolitical arena: political, geographical, economic, but also cultural. But what became of the slaves within the empire?

317   Pacta Veneta: I trattati con Bisanzio 992–1198, no. 2 (of 1082, pp. 35–45). Ralph-Johannes, Handel und Politik, pp. 8–16.

Slavery in Late Medieval Europe William D. Phillips In the later Middle Ages, from the twelfth century through the fifteenth, Europe’s internal economy continued to develop and strengthen. Slavery was still present, although agriculture throughout most of Europe was the concern of serfs or free peasants. Domestic and artisan slavery persisted, particularly in Italy and the Iberian Peninsula. In this chapter we shall examine late medieval European slavery, both as an important segment of slavery’s long history and as a prelude to the development of American slavery. Such an examination is especially important, because slavery in the New World would owe its origins to several late medieval traditions. First, the Crusades greatly expanded the activities of the Italian merchants, the most active European slave traders of the period, and, as an additional result of the Crusades, Europeans vastly expanded their knowledge of sugar, whose history in the Americas is impossible to disentangle from that of slavery. The Spanish and Portuguese reconquests kept slavery alive and flourishing in the Iberian Peninsula. The revival of Roman law and its influence on European legal development was also crucial, because the rules regulating slavery occupied such an important place in the Roman legal system. As a consequence of that revival, Spain and Portugal both had fully developed laws regulating servile status even before they began to expand overseas. We also shall consider the consequences of the Black Death in creating a heightened demand for slaves.

The Crusades and Sugar

As we have seen, slavery was constantly present in the Mediterranean lands from ancient to modern times. In the society of Christian Europe, slavery persisted around the shores of the inland sea even as it was declining in continental Europe. When western Europeans embarked on the Crusades, the Middle Ages’ energetic though abortive movement of territorial expansion, the participants of all ranks and callings knew slavery as an unquestioned aspect of their society, although the numbers of slaves were small everywhere in Europe. Source: Phillips, William D., Jr., “Slavery in Late Medieval Europe,” in William D. Phillips, Jr., Slavery from Roman Times to the Early Transatlantic Trade, Minneapolis: University of Minnesota Press, 1985, 88–113. © 1985 William D. Phillips, Jr., used by permission of the author.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_023

666

Phillips

Slavery In Late Medieval Europe

667

In the Muslim lands the crusaders found slaves and, more important, the opportunity for economic expansion. In 1095 Pope Urban II, in a sermon preached in the southern French city of Clermont, called for Western knights to march to Jerusalem and free Christianity’s most holy city from Muslim control. The Crusades, as the Western campaigns came to be called, owed their origins to a complex set of conditions present in several cultural regions. It is not sufficient to account for their origin solely on the basis of the appeal of papal propaganda for the knights and nobles of western Europe. One must go back several decades before 1095 and examine Near Eastern events in order to understand the background of the Crusades. In the middle of the eleventh century a group of nomads from central Asia, the Seljuk Turks, seized influence over the caliphate of Baghdad, the nominal center and directing force of Islam. Called in first as mercenaries, the Seljuks made themselves the most powerful group at the center of the Islamic world, able to dictate to the caliph. Thereafter a group of Seljuks began to move into Byzantine territory in Asia Minor. Once they had gained the land, the Seljuk conquerors sought to reestablish Byzantine agricultural practices. They “recolonized their devastated lands with Greek, Armenian and Syrian farmers, often kidnapping and enslaving entire villages and towns in Byzantine and Armenian lands.”1 Beyond the economic losses they suffered, the Byzantines clearly recognized the strategic threat of the Turks in their Anatolian possessions. Therefore the Byzantine emperor Alexius Comnenus called on the pope for a contingent of Western knights to cooperate with his own troops in driving the Turks from Anatolia. Because the pope never recorded his reasons for calling the First Crusade, historians have speculated and suggested a number of possibilities, most of which have little to do with the Byzantine request for a limited number of knights to help regain Anatolia. The most startling departure from what the Byzantines wanted came when Urban II called for the crusaders to go on to Jerusalem and described their goal as a land of milk and honey. Why the change? There was one strong connection between aid to the Byzantines and military action beyond Anatolia: the pilgrimages by Western Christians to Jerusalem had been quite popular, and the Seljuk irruption threatened to cut the routes used by pilgrims. Perhaps the pope did not distinguish carefully among the several Muslim groups, only considering it improper and unfortunate for Muslims to control the most holy places of Christianity. Another papal 1  Speros Vryonis, Jr., Byzantium: Its Internal History and Relations with the Muslim World: Collected Studies (London, 1971), 9:234.

668

Phillips

desire may have been to use the crusaders as part of a campaign to heal the division between Rome and the Byzantine church, a split that had existed for centuries but had only been formalized since 1054. Internal changes in Europe were probably additional motivations for the pope’s actions. For some time the papal administration had cooperated with secular authorities in an attempt to reduce domestic warfare, establishing the Truce of God and the Peace of God to limit the periods of war and shield noncombatants. What better way to save Europe from unwanted and destructive combat than to direct the energies of the knights, whose whole life was devoted to war, toward acceptable goals outside their homelands? For the knights, as for the nobles who led them, the compensations were more than adequate. They could gain religious rewards for participation; the pope allowed a moratorium on their debts while they campaigned; and they could hope to gain lands for themselves. The latter was especially important in a Europe that was in the first phases of its great medieval population rise, with available land for seigneurial holdings increasingly difficult to secure. In addition to the secular warriors and the church, one other European group had much to gain from the Crusades. This was made up of the merchants and seamen of the Italian towns, principally Venice and Pisa, who saw great economic opportunities in the carrying trade and possibilities of commercial privileges in the eastern Mediterranean. Their activities, too often neglected by historians of the Crusades, had a great impact on slavery.2 The response to the Byzantine appeal as relayed by the pope was overwhelming. By 1100 the crusading armies had established a line of enclaves along the shores of the eastern Mediterranean. The more familiar part of the story covers the kingdom of Jerusalem and the other principalities set up by the crusaders, but more significant for our study were the concessions the Italian merchants secured. For their support in conquering numerous towns, including Acre, Tripoli, and Beirut, Italians were given a portion of the urban land and the countryside. In Antioch in 1098 the Genoese, for example, received the “church of St. John and the square in front of it, a fonduk [a combination office building-residence-warehouse], a well, and permanent exemption from all taxes.”3 From their new possessions in the crusader states, the Italians engaged

2  For the Crusades, see Kenneth M. Setton, ed., A History of the Crusades, 2d ed., 5 vols. (Madison, 1969–); Stephen Runciman, A History of the Crusades, 3 vols. (Cambridge, Eng., 1951–54); and Hans Eberhard Mayer, The Crusades (New York and Oxford, 1972). 3  Charles Verlinden, The Beginnings of Modern Colonization: Eleven Essays with an Introduction, trans. Yvonne Freccero (Ithaca and London, 1970), pp. 79–80.

Slavery In Late Medieval Europe

669

in agriculture, producing market commodities such as sugar and cotton, and, above all, they engaged in commerce. The crusader states lasted fewer than two hundred years before the Muslims were able to reassert control over the territory. For those two centuries, although the crusaders imposed their own code of laws, they did little more than preserve the preexisting tenurial arrangements and labor dues of the Muslim peasants. As we know, agricultural slavery was atypical in the Muslim world. In Palestine before the Crusades, the peasants were subject to wealthy and powerful landlords: the state, rich nobles and townsmen, and religious foundations (waqf ). The cultivators leased the lands they worked for a payment of one-third or one-half of the crops to the owners. Because the crusaders were few in number and did not import European peasants to work the land, they were content to displace the Muslim landowners and continue the traditional relations with the peasants. The Europeans had little direct contact with the peasants living on their lands, dealing with the communities through the rayises (village headmen). The lease payments were generally collected in kind, and, since there was no demesne land, corvée labor was required only for the upkeep of roads and bridges. One of the most recent historians of the Crusades quotes a contemporary Muslim as reporting that Christians treated the peasants more favorably than the previous Muslim masters had done.4 Even though most Muslim laborers in the crusader states were free, slaves were still present. H. E. Mayer described Castle Safed, which controlled Jacob’s Ford, the most important crossing of the Jordan River, on the road from Acre to Damascus. After it was rebuilt in 1240, “the castle held a garrison of 50 Templar knights, 30 serving brothers, 50 Turcopole cavalrymen, 300 archers, and another 820 men and 400 slaves….”5 Muslims enslaved as a result of their capture in war and eastern European slaves brought in by the Italians were sold in urban slave markets, such as the one in Acre, and were mainly employed as domestics. No Christian could be enslaved, but Muslims who accepted Christianity were not guaranteed enfranchisement unless they had fled into Christian territory. Local converted Muslims generally remained in their servile condition. In fact, abuses were constantly reported. Landowners prevented Muslims from accepting baptism; some even prevented them from attending Christian sermons. Italian merchants fraudulently brought in numbers of eastern European Christians for sale as slaves, passing them off as Muslims.6 4  Meron Benvenisti, The Crusaders in the Holy Land (New York, 1972), pp. 20, 217–18. 5  Mayer, Crusades, p. 151. 6  Ibid., p. 178; J. S. C. Riley-Smith, The Feudal Nobility and the Kingdom of Jerusalem, 1117–1277 (Hamden, Conn., 1973), pp. 62–63, 76, 88, 258.

670

Phillips

Since most of the rural labor force was free, slavery did not play a crucial role in the internal economy of the crusader states. Nevertheless, the Crusades, taken as a whole, did have an important impact on the European experience with slavery in several indirect, but important, ways. By providing commercial bases, the crusader states allowed the Italian traders to expand their trading contacts throughout the eastern Mediterranean, including their hold on the slave trade. This expansion came at a propitious time for them, just when Europe’s economy was beginning to grow. The Crusades were in part a consequence of this economic quickening in Europe, and in turn they stimulated it. Taking advantage of their position in the eastern Mediterranean, the Italians extended their commercial connections north into the Black Sea and south to Egypt, unconcernedly crossing religious frontiers in the process. One of the most profitable portions of the Italian trade was slaving, and it was one that outlasted the Crusades by two centuries and more. In addition, Europeans learned a great deal from their exposure to the sophisticated ambience of the civilizations of the East. It was after the Crusades that rich Europeans, both noble and bourgeois, embraced a more elaborate style of life, with better clothes, a more varied diet, increased consumption, and richer houses. The example of the East stimulated these demands in the West: Italians in the East catered to the demands, providing spices, silks, sugar, and slaves. In urban houses servants were required, and the Italian merchants could easily supply slaves to fill the need. The fine Italian households of the late Middle Ages and the Renaissance were staffed with slaves, an aspect often forgotten when scholars wax eloquent about “Renaissance individualism.” One discovery that the Europeans made during the Crusades was sugar production. Sugar cane, as we saw in chapter 4, had been grown by the Muslims in the Middle East and the Mediterranean, and it continued to be grown by Europeans in the crusader states. It is not true, as it is sometimes said, that the first direct contact that Westerners had with sugar was in Syria and Palestine at the time of the First Crusade. Sicily and parts of Spain already produced sugar, and we have seen that Venice imported Egyptian sugar in the late tenth century. Some Europeans, and perhaps some crusaders, had already tasted sugar. Nevertheless, sugar cane plantations were new to the Frankish invaders and attracted the attention of the chroniclers. Fulcher of Chartres reported the hardships that limited food supplies caused for the army of the First Crusade and went on to say that in those cultivated fields through which we passed during our march there were certain ripe plants which the common folk called “honeycane” and which were very much like reeds. The name is compounded

Slavery In Late Medieval Europe

671

from “cane” and “honey,” whence the expression “wood honey” I think, because the latter is skillfully made from these canes. In our hunger we chewed them all day because of the taste of honey. However, this helped but little.7 William of Tyre, in his description of the irrigation projects around Tyre, also mentioned sugar. All the country round about derives immense benefits from these waters. Not only do they supply gardens and delightful orchards planted with fruit trees, but they irrigate the sugar cane also. From this latter crop sugar (zachara) is made, a most precious product, very necessary for the use and health of mankind, which is carried by merchants to the most remote countries of the world.8 Western sugar manufacturing and trade in sugar to the West dramatically increased as a result of the Crusades. After they reached Jerusalem, the crusaders maintained the cultivation of sugar cane in the states they established. This afforded them immediate revenues and, more important in the long run, created additional demand in the West, as returning crusaders and pilgrims took home samples of cane sugar and thus helped to spread the taste for it. Sugar plantations were controlled by the king of Jerusalem, by corporate groups such as the military orders and the Italian cities, and by individuals. As one example, the Knights of St. John, or the Hospitalers, received lands in the crusader states for their maintenance. Scholars have identified 858 pieces of property in the kingdom of Jerusalem; of these, the Hospitalers held 171 at various times. Like other wealthy landholders in the region, the Hospitalers generally allowed the native communities to continue their normal life without interference; they normally asked only that a rent be paid by each community. Exceptions to this rule were farms devoted to specialized crops: truck gardens, vineyards, olive groves, and sugar cane fields, proof of sugar’s importance in the economy. The king of Jerusalem also received indirect income from sugar. Among his regalian rights was the ability to impose duties on manufacturing: dyeing, tanning, and the production of sugar.9 7  Fulcher of Chartres, A History of the Expedition to Jerusalem, 1095–1125, trans. Frances Rita Ryan, ed. Harold S. Fink (Knoxville, 1969), p. 130. 8  William of Tyre, A History of Deeds Done beyond the Sea, trans. and ed. Emily Atwater Babcock and A. C. Krey, 2 vols. (New York, 1943), 2:6. 9  Jonathan Riley-Smith, The Knights of St. John in Jerusalem and Cyprus, c. 1050–1310, vol. 1. of A History of the Order of the Hospital of St. John of Jerusalem, ed. Lionel Butler (London

672

Phillips

Sugar production, dependent as it was upon water for irrigation and for motive power to turn the mills that crushed the cane, was concentrated in those places where water was readily available. The main centers for sugar production were coastal: Sidon, Acre, and Tyre (the latter the most important), and inland in Galilee and the Jordan Valley. In these places sugar production predated the Crusades, and soon the Westerners expanded production. The king of Jerusalem possessed sugar plantations and mills near Acre, and in 1160 King Baldwin III farmed out the royal mills to Renaud Fauconnier and gave him the right to construct new mills.10 Disputes over water rights emerged among the Western landowners in the crusader states. One example of this was the long-term acrimony that developed between the Templars and the Hospitalers over the watercourse called the Flum d’Acre. In the 1230s the Templars built a dam on the river to increase the water flow to their mills at Doc and Recordane. Rising waters threatened the Hospitalers’ lands upstream, and the two orders agreed to a compromise about the level that the river would be allowed to reach. By 1262 the dispute reemerged. The Templars were accused of obstructing the river with additional construction, while the Hospitalers—because of their intensive cultivation of sugar cane—had diverted so much water to irrigate their fields that the level of the river had fallen far enough to render the Templars’ mill at Doc useless. Another dispute was between the monastery of La Quarantaine and the widow of the lord of Sidon, Eustace Garnier, who had controlled the water rights in the area and had restricted the monks’ use of irrigation water to one day in each two-week period. Because of pressure exerted by the viscount of Sidon, Garnier’s widow allowed the monks weekly use of the water.11 Italian cities, too, received land grants in the crusader states. The example of Tyre will serve to illustrate the general pattern. Even before the Crusades, merchants from Amalfi, Venice, and Pisa had traded in Tyre for silks and and New York, 1967), pp. 425–27, 434; Wilhelm von Heyd, Histoire du commerce du Levant au Moyen-Age, trans. Furey Reynaud, 2 vols. (Leipzig, 1885–86; reprint, Amsterdam, 1967), 2:684; Mayer, Crusades, p. 163. Wallace K. Aykroyd (The Story of Sugar [New York, 1967]) generally has accurate information, but he is mistaken when he says that in Syria and Palestine “Western Europeans, represented by the Crusaders, saw and tasted sugar for the first time.” (p. 13.) Many other Westerners had known sugar at home for over a century, but the Crusades did popularize sugar and created a demand for it in the West. 10  Jean Richard, The Latin Kingdom of Jerusalem, trans. Janet Shirley, 2 vols. (Amsterdam, 1979), p. 73: Joshua Prawer, The Crusaders’ Kingdom: European Colonialism in the Middle Ages (New York, 1972), pp. 363–64. 11  Riley-Smith, Knights of St. John, p. 446; Richard, Latin Kingdom, p. 125. See also Benvenisti, Crusaders in the Holy Land, p. 254, on archeological ruins of sugar factories.

Slavery In Late Medieval Europe

673

sugar. In the division following the European conquest, Venice gained control of Tyre. In 1240 the Venetian agent in Tyre reported on the twenty-four estates owned in the city’s vicinity, devoted to market crops and including truck gardens, vineyards, orchards, and cane fields. The labor in these estates was Syrian and was employed under sharecropping arrangements. The reputation of Tyre as a sugar-producing region was widely known. It was from there that Frederick II secured experts to help revitalize the Sicilian industry in the midthirteenth century.12 Although their existence was precarious toward the last, the Christian states continued to produce sugar for their own use and for export. When the Muslims drove the last of the Westerners from the mainland in 1291, the Christian refugees moved to the Mediterranean islands and took sugar production with them.13 Sugar thereafter underwent a westward progression, as its intensive cultivation and processing spread to the islands and mainland areas held by the Christians. In most of these regions, sugar had been grown by Muslims since the eighth or ninth centuries, but only in the later Middle Ages did it come to be exploited more fully. Parenthetically, it must be noted that the fall of the crusader states in the East did not mark the end of commerce in sugar between Muslims and Christians. Even though sugar from the Christian-held Mediterranean islands soon began to be exported to Europe, Damascus and Tripoli in Syria and Alexandria and other Egyptian ports continued to provide Western merchants with sugar and other goods.14 Among the first islands to be exploited for sugar production was Cyprus. Sugar cane had probably been grown on the island since it was introduced by the Muslims in the seventh century, but it was only after the crusader states had fallen that it became important in the island’s economy. Famagusta, the principal city, became a vital trading center when it inherited Acre’s role as the nexus of Christian trade in the region. Its prosperity depended on shipbuilding and exports, of which the most valuable were sugar, salt, wine, and cloth. 12  Noël Deerr, The History of Sugar, 2 vols. (London, 1949–50), 1:76–77; Richard, Latin Kingdom, p. 351; Joshua Prawer, “Étude de quelques problèmes agraires et sociaux d’une seigneurie croisée au XIIIe siècle,” Byzantion 22 (1952): 5–61; Verlinden, Modern Colonization, p. 20; James Westfall Thompson, Economic and Social History of the Middle Ages (300–1300), 2 vols. (New York, 1928; reprint, 1959), pp. 362, 424; Edmund O. von Lippmann, Geschichte des Zuckers, seiner Darstellung und Verwendung (Leipzig, 1890), p. 181. 13  Saladin’s armies destroyed sugar mills in Tiberias and Acre, something Saladin himself regretted. Richard, Latin Kingdom, p. 176. 14  Heyd, Commerce du Levant, 2:686; Thompson, Economic and Social History, p. 423.

674

Phillips

The first recorded export of sugar from Cyprus dates from 1301 when a Genoese merchant in Pisa purchased ten chests of the commodity. Sugar was produced in several parts of the island, and numerous producers were involved: the Lusignan dynasty itself, the Knights of St. John, Venetians such as the Corner (or Cornaro) family, and a Catalan family—the Ferrers. The most important plantations were those of the Corner family at Piskopi, in the southern part of the island along the river Kouris. Although some slaves were used in Cypriot sugar production, especially on the royal estates, the bulk of the labor was nonslave and included Greek and Syrian immigrants given special political status by the Venetians. The Syrians were often skilled sugar workers, brought in especially for their expertise. In 1494 an Italian visitor reported that some four hundred workers were engaged in sugar production at Piskopi. Sugar cane in Cyprus was grown in irrigated fields, and there, too, disputes over water rights involved rival producers.15 Other islands in the eastern Mediterranean also produced sugar. Crete was in Byzantine hands until the early thirteenth century, when the Catalan count of Monserrat conquered it in the aftermath of the Fourth Crusade. Shortly afterward the Venetians bought Crete and in the fifteenth century fostered sugar production there.16 Sicily was probably the most important Mediterranean center for sugar in the late Middle Ages. Sugar cane had been introduced shortly after the Muslims conquered the island in 878, and by the end of the ninth century Sicilian sugar was being sold in North Africa. Palermo was in the main region of sugar production in Sicily, and ibn-Hauqal wrote that “the banks of the streams around Palermo, from their sources to their mouths, are bordered by low-lying fields, upon which the Persian reed is grown.”17 In the eleventh century the Normans took Sicily, and their conquest was motivated in part by sugar. The Christians of Sicily had provided samples of sugar, among other products, to the Normans in order to demonstrate the island’s wealth and to encourage them to expel the Muslims. The sugar industry persisted in Norman 15  David Jacoby, “Citoyens, sujets et protégés en Chypre,” Byzantinische Forschungen 5 (1977): 174–777; Freddy Thiriet, La Romanie vénitienne au Moyen Age: Le développement et l’exploitation du domaine colonial vénitien (XIIe–XVe siècles) (Paris, 1959), p. 333; Gino Luzzatto, Studi di storia economica veneziana (Padua, 1954), pp. 117–23; Gino Luzzatto, Storia economica di Venezia dall’XI al XVI sècolo (Venice, 1961), pp. 54, 64, 196; Frederic C. Lane, Venice: A Maritime Republic (Baltimore and London, 1973), pp. 144–45; Mayer, Crusades, pp. 231–32, 237, 239; Verlinden, Modern Colonization, pp. 19–20, 97n. I wish to thank David Jacoby for his suggestions on this topic. 16  Deerr, History of Sugar, 1:79, 83; Verlinden, Modern Colonization, pp. 96–97; Thiriet, Romanie vénitienne, pp. 417–18. 17  Ibn-Hauqal, quoted by Deerr, History of Sugar, 1:76.

Slavery In Late Medieval Europe

675

times. Both the king and the cathedral of Montreale in Palermo owned sugar mills. One indication of the novelty that sugar cane presented to the Norman knights when they first saw it under cultivation is provided in a letter that Hugh Falcandus wrote to the treasurer of the diocese of Palermo. Falcandus described the great extension of wonderful reeds which are called by the inhabitants honey canes, from the sweetness of the juice they contain. The juice of these canes, when it has been boiled with care and to the right extent, converts itself into a kind of honey; if on the other hand it is subjected to a more complete and perfect boiling, it becomes condensed into the substance of sugar.18 No doubt because of the destruction accompanying the Norman conquest of Sicily and of the inexperience of the new estate and mill managers, the Sicilian sugar industry underwent a decline during the Norman period. The reign of the emperor Frederick II (king of Sicily from 1197 to 1250) marked a temporary revival. Consistently sponsoring enlightened reforms for the economy of the kingdom, Frederick was particularly interested in sugar production. He had the fields hedged to prevent damage to the cane, and once, during an infestation of caterpillars, he mobilized the population and assigned a daily quota for the hand collection of the pests. In 1239 he had two experts in sugar refining sent from Tyre to Sicily to examine the industry there and recommend improvements.19 After Frederick’s death, Sicily underwent over a century and a half of disordered political conditions, which rendered Frederick’s reforms useless. Sugar’s decline was exacerbated by the expulsion of the Muslims from the island. Sugar received favor from Sicily’s rulers early in the fifteenth century, and stability returned under the Aragonese king Alfonso V (1416–58). During his reign the Sicilian sugar industry reached a peak, aided by studies of proper irrigation of the fields undertaken by the University of Palermo.20 It has often been asserted that the famous three-roller sugar mill—so characteristic of the New World plantations—came into use in Sicily during

18  Quoted by Deerr, History of Sugar, 1:77. 19  Verlinden, Modern Colonization, p. 20; Deerr, History of Sugar, 1:76–77; Richard, Latin Kingdom, p. 351. 20  Carmelo Trasselli, “Producción y comercio del azúcar en Sicilia del siglo XIII al XIX,” Revista Bimestre Cubana 72 (1957): 138–41. The article was also published as “Produzione e commercio delle zucchero in Sicilia dal XII al XIX sècolo,” Economia e Storia 2 (1955).

676

Phillips

Alfonso V’s reign. This is almost assuredly not the case, and the new mill did not appear until after the discovery of America.21 Despite the lack of the three-roller mill, the support of the island’s rulers did help sugar reach new heights in Sicily in the fifteenth century. An indication of its importance there was the designation of one of the entries into the city of Syracuse as the Porta degli Zuccheri, or the gate of the sugars (or sugar workers). Also, in 1446, one of Syracuse’s streets was named Cannamella, honey cane.22 Because of its potential profitability, sugar production spread to the Italian mainland, although it never became important there. In the thirteenth century a small amount of sugar cane was grown in Lombardy. Some sugar was produced in the kingdom of Naples, and around 1550 Ferdinand de’ Medici introduced it unsuccessfully near Florence.23 Most labor for the Mediterranean sugar industry—on both the Christian and Muslim sides of the sea—was free. Although here and there some slaves may have been used, they were unusual. The close identification of sugar cane and slave labor came later, in the lands of the Atlantic. At this point we must defer the continuation of sugar’s history in order to examine the parallel paths of slavery and the slave trade in the later Middle Ages.

Slavery in Italy

From ancient times to the seventeenth century, Italy knew slavery. While slavery declined in Europe north of the Alps and serfdom rose, the trade in and the use of slaves persisted in Italy. One reason was the durability of urban life in 21  The mistaken statement is found in Lippmann, Geschichte des Zuckers, pp. 217–18, and Deerr, History of Sugar, 1:77, 2:536. Verlinden follows Lippmann and Deerr (Modern Colonization, p. 20). The definitive statement on the mill is by Moacyr Soares Pereira, A origem dos cilindros na moagem da cana: Investigacão em Palermo (Rio de Janeiro, 1955). Of later writers, José Pérez Vidal supports Pereira totally in La cultura de la caña de azúcar en el Levante español (Madrid, 1973), pp. 63–65, while J. H. Galloway leans toward Pereira’s case in “The Mediterranean Sugar Industry,” Geographical Review 67/2 (1977): pp. 186–87. 22  Trasselli reported that only free workers were employed in the Sicilian sugar industry, “Producción y comercio,” p. 135. Verlinden’s citation of the rental of a Turkish slave to work for a Sicilian sugar plantation owner in 1462 probably represents only an exception. Charles Verlinden, L’esclavage dans l’Europe médiévale, vol. 2; Italie—Colonies italiennes du Levant—Levant latin—Empire byzantin (Bruges, 1977), 2:231. 23  Gerald A. J. Hodgett, A Social and Economic History of Medieval Europe (New York, 1974), p. 196; Deerr, History of Sugar, 1:79.

Slavery In Late Medieval Europe

677

the Italian peninsula, and urban life in preindustrial societies often included domestic slavery. More important was Italy’s geographic location and the propensity of many Italians for long-distance trade. Located in the center of the Mediterranean, Italy had links with the entire basin of the inland sea and with what Fernand Braudel has called “the greater Mediterranean,”24 those lands beyond the shores of the sea but linked to it. Through the passes of the Alps, Italians developed trading ties with the northern parts of western and central Europe. Italian traders journeyed to the Byzantine cities, to the markets of the eastern Mediterranean from Anatolia to Egypt, to the North African ports (where trans-Saharan products could be obtained), and to the cities of the Iberian Peninsula (which gave them an opening on the Atlantic). They even ventured beyond Constantinople to reach the markets of the Black Sea. At one time or another, all these areas served the Italian traders as markets for the sale or purchase of slaves. Early on the Venetians exported slaves to the eastern Mediterranean; around the late eleventh or early twelfth centuries they were joined by merchants of other Italian cities. Our task now is to examine slavery in Italy proper from the eleventh century to the end of the fifteenth. The Italian use of slaves was widespread, and most were recruited by the slave trade. In the late eleventh century, when the Normans conquered southern Italy, they often enslaved members of the defeated population, but this was an atypical situation.25 Wars in the Italian peninsula only rarely ended with the enslavement of the defeated. The Norman aberration was likely due to the position of the Normans as invaders from outside who did not identify with those they vanquished. Italians mainly imported their slaves from outside. Already by the twelfth century, Venice had a slave population that was recruited from a wide area, though it was a small percentage of the total Venetian population.26 By the mid-twelfth century Genoa possessed Muslim slaves, and there is documentary evidence that one of the wealthy Genoese merchants of the time, Guglielmo Burone, was a slaveowner.27 In 1300 the will of a Genoese resident in Famagusta

24  Fernand Braudel, The Mediterranean and the Mediterranean World in the Age of Philip II, trans. Siân Reynolds, 2 vols. (New York, 1972), 1:168–230. 25  Dennis Mack Smith, Medieval Sicily, 800–1713, vol. 2 of A History of Sicily (London, 1968), p. 15. 26  Christopher Brooke, Europe in the Central Middle Ages, 962–1154 (New York, 1964), pp. 116–17. 27  Eugene H. Byrne, “Genoese Trade with Syria in the Twelfth Century,” American Historical Review 25 (1920): 206.

678

Phillips

on Cyprus indicated that he owned five female slaves and one male slave.28 Iris Origo wrote that by the end of the fourteenth century there was hardly a well-to-do household in Tuscany without at least one slave: brides brought them as part of their dowries, doctors accepted them in lieu of fees—and it was not unusual to find them even in the service of a priest.29 In the fourteenth and fifteenth centuries, slaves in Italy came from a number of areas and ethnic groups. There were Muslims, Tatars, Circassians, Russians, Bulgarians, Armenians, Albanians, a few from the islands of the Mediterranean, and fewer still from black Africa and the Canaries. Most were women, and most were used as domestics and concubines. They and their children could attain freedom in a variety of ways.30 One story illustrates several aspects of the common pattern. In 1427 Cosimo de’ Medici of Florence had an illegitimate son by a Circassian slave girl. The child later was legitimized and ended his life as the archpriest of Prato.31 In the Italian states probably the greatest number of slaves were used as domestic servants, and a percentage of these acted as concubines for their masters. Small numbers of others were used in artisan and craft workshops. There is little evidence of slave use in agriculture. Small market farmers may have employed a few slaves as laborers, but plantation agriculture was not important in the peninsula. Italians did make use of a limited number of slaves in agricultural pursuits on the islands they controlled in the Mediterranean, but the practice was not translated often to the mainland. There were other slaves in Italy, additional victims of the slave trade, acquired elsewhere and present in Italy only until they could be transshipped for sale elsewhere.32

28  Verlinden, Modern Colonization, p. 91. 29  Iris Origo, The Merchant of Prato: Francesco de Marco Datini, 1335–1410 (New York, 1957), pp. 90–91. 30  G. Pistarino, “Fra liberti e schiave a Genova nel Quattrocento,” Anuario de Estudios Medievales 1 (1964): 353–55. 31  Joseph and Frances Gies, Merchants and Moneymen: The Commercial Revolution, 1000– 1500 (New York, 1972), pp. 244, 257. 32   Michel Balard, “Remarques sur les esclaves à Gênes dans la seconde moitié du XIIIe siècle,” Mélanges d’Archéologie et d’Histoire 80 (1968): 627–80, especially 651–53; Giovanna Balbi, “La schiavitù a Genova tra i sècoli XII e XIII,” in Mélanges offerts à René Crozet, ed. Pierre Gallais and Yves-Jean Riou, 2 vols. (Poitiers, 1966), 2: 1025–29, especially 2:1029; Domenico Gioffrè, Il mercato degli schiavi a Genova nel sècolo XV (Genoa, 1971),

Slavery In Late Medieval Europe

679

Because most slaves were young women living in the household, it is not surprising that sexual liaisons developed between them and their masters. Scholars have uncovered a plethora of Italian documentary evidence concerning concubinage and the complications it produced. Sexual use and abuse of slaves were features of most slave systems, but for medieval Italy we have our first relatively complete records of such activities. Iris Origo has pointed out the psychological turmoil that the presence of young, defenseless female slaves created in the slaveholding families. Not without justification, free women came to hate slaves, and often the slaves reciprocated.33 Giovanna Balbi found that in twelfth-century Genoa domestic concubinage between masters and female slaves [was] certainly widespread and frequent…. The affectionate relations between slaves and free citizens was not limited only to the family ambience, but also [included] those outside the family: in that case they led not only to procreation but sometimes to matrimony as well….34 Several paths were open to the master who impregnated his own slave. He could arrange a marriage for the slave; he could raise the child as a slave; he could sell the mother while she was still pregnant; or he could free the mother and marry her. The last option has not left many traces in the records. Origo quoted two letters written by an agent of a merchant who owned a female slave. The slave you sent is sick, or rather full of boils, so that we find none who would have her. We will sell or barter her as best we can, and send you the account. Furthermore, I hear she is with child, two months gone or more, and therefore she will not be worth selling.35 The agent wrote again some time later about the same slave. “No man will have her. She says she is with child by you, and assuredly seems to be. The pother she makes is so great, she might be the Queen of France.”36 A letter describing pp. 89–90; Lane, Venice, p. 51; Jacques Heers, Esclaves et domestiques au Moyen-Age dans le monde méditerranéen (Paris, 1981), pp. 135–41. 33  Iris Origo, “The Domestic Enemy: The Eastern Slaves in Tuscany in the Fourteenth and Fifteenth Centuries,” Speculum 30 (1955): 321–66, especially 340–44. 34  Balbi, “Schiavitù a Genova,” 2:1028. 35  Origo, Merchant of Prato, p. 93. 36  Ibid.

680

Phillips

a different situation tells the same story of the helplessness of the slave and the indifference of the master. We spoke to the chaplain to whom your slave belonged and he says you may throw her in the sea, with what she has in her belly, for it is no creature of his. And we deem he speaks the truth, for had she been pregnant by him, he would not have sent her…. Methinks you had better send the creature to the hospital.37 Not all slaveowners were this harsh and uncaring. Many treated their slave mistresses and their illegitimate children with great kindness and bequeathed them generous amounts in their wills.38 In cases in which an outsider rendered a slave pregnant, the law in the Italian cities was clear: the master, not the slave, was to be compensated, because her economic value had been diminished as a result of her pregnancy. “Laws which provided financial compensation to the master of an impregnated slave were motivated by a desire to avoid economic loss to the master and to take into account the possibility of death in childbirth.”39 The offspring produced by such liaisons could be dealt with in a variety of ways. The master could rear the child or could turn the newborn infant over to its father. The child could be exposed. The mother could be rented out as a wet nurse for a high price.40 Infants lucky enough to be taken in by orphanages were free, and the hospital itself was listed as the father, thus conveying legitimate status on the orphans.41 The compensation owed to the master varied in different Italian cities. In Genoa it was one-half of the slave’s value. In Siena and Florence it was one-third of the value in addition to the full costs of the birth if the woman lived. If she died, the master would be entitled to her full value, though such laws were not always enforceable.42 In 1462 a statute came into force in Florence carrying a fine of one thousand florins if two eyewitnesses or four people with secondhand knowledge of the event swore that a man came into another’s house “to visit a slave or to lie with her.”43 Given the penalties that could be imposed, it 37  Ibid. 38  Jacques Heers, Gênes au XVe siècle: Activité économique et problèmes sociaux (Paris, 1961), p. 554. 39  Pistarino, “Liberti e schiave,” pp. 361–62. 40  Gioffrè, Mercato degli schiavi, p. 96 41  Ibid., pp. 100–104; Pistarino, “Liberti e schiave,” p. 362. 42  Origo, “Domestic Enemy,” pp. 345–46; Pistarino, “Liberti e schiave,” p. 361. 43  Origo, “Domestic Enemy,” p. 346.

Slavery In Late Medieval Europe

681

is not surprising that insurance contracts could be purchased in Genoa by the fathers of the unborn children of slaves to idemnify the master in the case of his slave’s death.44 The antinatalist policies of the Italian slaveowners is additional evidence of the domestic use they made of their slaves. The owners of rural slaves often encouraged procreation as a long-term measure to gain new workers, whereas those who employed domestics did not want to lose the short-term advantages. Aside from domestic service, there were other jobs for slaves, but they were not especially numerous. Artisans bought a few slaves as helpers in manufacturing, since the artisan workshops were small and required few workers. Michel Balard believed that one or two slaves in artisan and merchants’ families were typical in thirteenth-century Genoa. The will of Guglielmo Caffaraino, a cloth maker of that city, provided for the freeing of his six slaves upon his death. Five were male and probably were employed in their master’s workshop, while the woman probably worked as a domestic in the household.45 In the fifteenth century there were some black slaves who were gondoliers in Venice, but they were a quite small minority. Most of the gondoliers, one to two thousand in number, were self-employed. Origo portrayed oarsmen in the Italian galleys as slaves, but this has been effectively refuted by F. C. Lane, who reported that while there may have been some few slave oarsmen used on private vessels, the Venetian laws prohibited slaves on board the municipally controlled vessels.46 Some other uses for slaves were possible. They could be pledged as security for loans. They could be rented to craftsmen, and they would later return to their masters with enhanced values once they had learned their crafts. After acquiring a craft, slaves could usually be sold only to masters in the same craft. Certain occupations were closed to them. In Venice, slaves could not be taught certain specialized procedures in the manufacturing of fine cloth. Genoese pharmacists were prohibited from teaching slaves of Turkish or Tatar origin the secrets of their profession.47 The services of slaves could be rented out. “In 1281, for example, Bartolino de Asture sold to Opossino de Aldono the services of his slave Martino dragumannus from Slavonia, for a period of fifteen years…. The said slave, used as an interpreter, swore on the Bible to observe the contract.”48

44  Heers, Gênes, p. 554. 45  Balard, “Esclaves à Gênes,” pp. 651–52, 664–66. 46  Origo, “Domestic Enemy,” p. 333: Lane, Venice, pp. 333, 347. 47  Gioffrè, Mercato degli schiavi, pp. 90–91. 48  Balard, “Esclaves à Gênes,” pp. 673–74.

682

Phillips

The methods of acquisition of slaves in the Italian cities were recorded in deeds of sale, which specified the details of the transactions, such as the price and origin of the slave, as well as the slave’s complete physical description. The seller guaranteed that the slave was free of physical defects or pregnancy. If a purchaser later discovered that his slave had been pregnant at the time of sale, he could have his purchase price refunded. The new owner got full control over the new slave, including the right to discipline his new slave by beating or whipping. He had the same rights over his own family, a legacy from Roman family law. As a result, slaves were not in too different a position from that of other family members. The master could treat his slaves harshly, but definite limits existed. No master could kill his slave with impunity, and the state dealt strictly with violations.49 The new slaves entered quickly into a process of integration with the family. If they had not been baptized at an earlier point on their journey to Italy, they were baptized and given Christian names. It was important that the baptism came after the enslavement, because Christians were not permitted to be enslaved. Violations of this rule occurred with some regularity, as impious slave dealers passed off Greek Orthodox and even foreign Catholic slaves as nonChristians.50 Once baptism took place, the new slave was on the way to becoming a member of the new household both legally and physically. Because slavery in Italy was above all an urban phenomenon, and space in the cities was at a premium, with no possibility of separate slave quarters, slaves shared their masters’ houses. In the more elaborate households, they lived in the attics, in small rooms between the main floors, or in Venice on the sea floor (the ground floor).51 Many slaves resented their status and made efforts to run away, but there was not much hope for successful flight. The church provided only temporary sanctuary and in most cases turned the runaways over to their masters. An illustration of a failed attempt appeared in a merchant’s letter. We hear from Ibiza that Ser Antonio Dello has arrived there with many Moorish captives on his ship, and twelve of them ran away with his rowing boat…. But because of the weather, the said Moors came here [to Majorca] and for the present have been imprisoned, which is a great piece of good fortune.52 49  Origo, “Domestic Enemy, pp. 333–34, 340. 50  Ibid., pp. 334–36. 51  Lane, Venice, p. 206. 52  Origo, Merchant of Prato, p. 92.

Slavery In Late Medieval Europe

683

Pursuit could be far-reaching. A slaveowner in Barcelona wrote to an Italian living in Avignon asking him to be on guard for two runaways. One is named Dmitri, a big man and very handsome. His flesh is good, fresh and rosy. [The other] lacks a tooth in front and has a rather greenish skin…. I pray you … to have them caught, let them be strongly fettered, and send them back by boat to me.53 Iris Origo found an insurance policy dating from 1401 for the shipment of a Tatar slave, Margherita, from Porto Pisano to Barcelona. The value of the slave was set at fifty gold florins, and the owner would receive reimbursement if the slave met misadventure at sea. But the policy specifically excluded the illness or death of the slave and “any attempt at flight or suicide—‘if she throws herself into the sea of her own accord’—a risk which would hardly have been specified, had not such incidents occurred fairly often.”54 This illustrates the desperate lengths to which some slaves would go to escape their bonds. There were other avenues to freedom, and among them manumission was the most common. The slave population in the Italian cities was constantly changing as manumissions freed existing slaves, who then were replaced by newly imported people. The major reason for this regular turnover is that the majority of the slaves were young females destined for domestic work. Their proximity to the family and their racial background, generally white, meant relatively easy assimilation. We have discussed the sexual ties that developed between slave and free. They were only part of a wider array of often affectionate attachments that in some cases linked the slaves and the families they served.55 Women slaves could attain freedom by matrimony. If a master married his own slave, she was considered to be free, and this was also true in the case of a female slave who married a free man with her master’s permission. A master who wished to make his own child by a slave mother legitimate could do so on his accord only if he were living openly with the child’s mother. If he were not, if he had a legal wife, he needed permission of the authorities.56 Francesco Datini, the famous merchant of Prato, had a daughter by one of his slaves.

53  Ibid. 54  Origo, “Domestic Enemy,” p. 331 55  Balbi, “Schiavitù a Genova,” 2:1029; Heers, Gênes, pp. 554–56. 56  Pistarino, “Liberti e schiave,” p. 364.

684

Phillips

He later gave his mistress her freedom and arranged a marriage for her. He also adopted his daughter and left her two hundred lire in his will.57 As we noted, most slaves were baptized at or shortly after their sale, and thereafter they were expected to conform to Christian practice. But adherence to Christianity was not by itself a path to freedom. Manumission always required the volition of the master, who often would specify some conditions for the newly freed person to fulfill. In Genoa in 1186 records indicate that a female slave of Muslim origin was freed on condition that she remain in her master’s service for ten years. In 1190 Otto Guercio and his wife Anna freed Elena, a Sardinian slave woman, contingent upon her resisting the sexual advances of one Armanino until he married her.58 Monetary payments were often required before manumission was permitted, and this fact indicates that some slaves had the opportunity to acquire their own money. One Genoese slave was freed when she was able to pay ten lire for her freedom, while another bought her manumission for eight lire and the promise to pay her former master one lira for each of the next six years. Genoa may have had small groups of freedmen from various ethnic groups who aided their countrymen in purchasing their freedom.59 Manumissions by testament were common, usually by the will of the master, but occasionally by the will of his widow. The newly freed were often required to serve the masters’ heirs at low pay. As one example, a Genoese resident in Cyprus left instructions in his will to free three female slaves. Two obtained freedom immediately, as well as living quarters and money. The third was to serve the widow for four years, after which she would receive a small house, some furnishings, and ten barrels of wheat. After freedom, the newly manumitted could rise in the social hierarchy, sometimes very quickly and far.60 The mobility can be explained by the ethnic origin of the majority of the slaves in Italy. They were mostly white and not easily distinguishable from the native born, once they had learned the language and the customs of their host society. We can account for their ethnic origins by an examination of the recruitment of slaves for the late medieval and Renaissance Italian cities. The slave trade to Italy was almost exclusively in the hands of long-distance traders who purchased people in distant markets. True, the Normans had enslaved some women and children as they conquered areas in the Italian south and in Sicily in the late eleventh century, and in later centuries prisoners 57  Origo, “Domestic Enemy,” p. 345. 58  Balbi, “Schiavitù a Genova,” 2:1027–28. 59  Balard, “Esclaves à Gênes,” p. 678. 60  Ibid., p. 676; Verlinden, Modern Colonization, p. 91; Heers, Gênes, pp. 554–55.

685

Slavery In Late Medieval Europe

captured in the wars among the Italian states were held for ransom at times and threatened with slavery, but these were the exceptions that proved the rule. Italian merchants were among Europe’s most energetic and successful traders in distant regions. An important part of their trade was devoted to slaves, and while the bulk of their slave trade was concerned with supplying third parties, especially the Muslims of Egypt, with slaves purchased in other regions (the Balkans and the northern ports of the Black Sea especially), they also brought slaves home to Italy to be sold. In the early Middle Ages the Italians, particularly the Venetians, had supplied the markets of the eastern Mediterranean with wood and central European slaves in return for luxury goods purchased there. This began to change in the twelfth and thirteenth centuries. Wood gave way to manufactured goods, especially woolen cloth, and the Italian merchants shifted from central and eastern Europe to the Black Sea in their search for slaves. This important ChristianMuslim trade was vigorous, despite papal attempts to limit it by forbidding exports of arms, war materiel, and slaves to the Muslims. R.-H. Bautier, however, believed that “the blockade does not seem to have been very effective since in 1154 Pisa undertook to deliver precisely those forbidden commodities to the sultan.”61 In Genoa the slave market seems to have come into full flowering around 1190. Before that there had been only sporadic sales of slaves, including some probably Muslim black slaves who had been brought in by non-Genoese sellers. In the 1190s a regular series of slave sales began. Balbi studied the sales in Genoa in the period 1186–1226 and found recorded sales or manumissions of fifty-eight slaves. The accompanying tabulation summarizes Balbi’s findings.62

Muslims Sards Corsicans Greeks Unspecified

Males

Females

17 (3 black)  3  1 —  2 23

13 (1 black) 10 —  1 11 35

61  R. H. Bautier, The Economic Development of Medieval Europe, trans. Heather Karolyi (New York, 1971), p. 101. See also Lane, Venice, p. 60. 62  Balbi, “Schiavitù a Genova,” 2:1025–29.

686

Phillips

These figures, although they record a relatively small number of cases, are consistent with scholars’ views of slavery in Italy. Women outnumbered men, no doubt because domestic slavery was more important than industrial or agricultural. Italians bought slaves to act as servants and preferred women for domestic work. In these years we see a recruitment pattern based solidly on the western Mediterranean; the Black Sea became important for Genoa only later. Muslims from North Africa or Spain, and non-Muslims from the islands predominated. Only one Greek appeared in the records. The relatively low numbers of slaves are interesting, too. They reflect the strictly occasional demand for slaves as domestic laborers in the thirteenth century, when the general European population was on the rise and when Italian cities could fill most of their labor needs with the native free. Only in the fourteenth century, when an escalating series of disasters caused a severe decline in population, was there a substantial rise in demand for slaves.63 In the thirteenth century, the Genoese were especially active in the slave trade. They had participated in the First Crusade, and thereafter they maintained a steady commerce with Muslim Egypt. Until the second half of the thirteenth century, they played a secondary role to Venice, whose doge persuaded the leaders of the Fourth Crusade to divert their armies to the conquest of Constantinople from the Byzantines. As a consequence, the Venetians received favored trading rights in the Byzantine Empire and easier access to the Black Sea. When Michael Paleologus drove the Westerners from the Byzantine Empire in 1261, the Genoese, thanks to a previous treaty with the new Byzantine ruler, replaced the Venetians in the eastern imperial trade and received a near monopoly on trade to the Black Sea. In Pera, a suburb of Constantinople, and in Tana and Caffa on the north side of the Black Sea, the Genoese merchants maintained commercial posts, from which they obtained slaves among other goods, until the fifteenth century. During the second half of the thirteenth century, Genoa regularly supplied Mamluk Egypt with slave recruits for its army, due in large part to the fact that the normal overland caravan routes from the Black Sea across Anatolia to Egypt were cut.64

63  Balard, “Esclaves à Gênes,” p. 680. 64  Andrew S. Ehrenkreutz, “Strategic Implications of the Slave Trade between Genoa and Mamluk Egypt in the Second Half of the Thirteenth Century,” in The Islamic Middle East, 700–1900: Studies in Economic and Social History, ed. A. L. Udovitch (Princeton, 1981), pp. 335–45. See also Elena C. Skrzinskaja, “Storia della Tana,” Studi Veneziana 10 (1968): 3–45.

687

Slavery In Late Medieval Europe

The shift had repercussions back to Italy. Thirteenth-century slave recruitment in Genoa illustrates the common Italian pattern. In the first threequarters of the century, Muslims made up around 75 percent of the slave population, reflecting the well-established trading links with Iberia and North Africa. The Aragonese conquest of the Islamic kindgom of Valencia in the 1230s placed numerous Muslims on the market, and many of these found their way to Italy. Most of these Muslims were sold in Italy before they had been baptized and given Christian names. In the final quarter of the thirteenth century, the situation began to change dramatically, as the primary area of recruitment shifted to the Black Sea. In 1275 the first recorded Russian slave, a man named Balada, appeared in Genoa. He was the first of many slaves brought from the still-pagan areas of the Black Sea. Their numbers grew rapidly when the Genoese set up trading stations in Pera beginning in 1281 and in Caffa around 1289. Most of the slaves brought in from the Black Sea reached Genoa after having been baptized.65 This set the pattern for the fourteenth and fifteenth centuries. Until the Turks cut the straits connecting the Black Sea and the Mediterranean in the mid-fifteenth century, the bulk of the slaves came from the region of the Black Sea. This is borne out by Origo’s analysis of the official list of slaves sold in Florence between 1366 and 1397. The ethnic distribution of the 357 slaves listed is shown in the accompanying tabulation. Tatars Greeks Russians Turks Circassians Bosnians or Slavs Cretans Muslims

65  Balard, “Esclaves à Gênes,” pp. 630, 634–49.

274 30 13 8 4 5 1 22

688

Phillips

Origo reported that 329 of the slaves were female, either women or young girls, while only 28 were males, and only 4 of them were over the age of sixteen.66 Michael E. Mallett also found that most of the slaves in Florence were from the region of the Black Sea. Many came overland to Florence from Ancona on the Adriatic, where several Florentine merchants had set themselves up as slave traders. With Florence’s conquest of Pisa and its port, the city’s merchants had maritime access to the Black Sea. Thereafter they could purchase their chattels directly in the Black Sea slave markets and were no longer dependent upon the Venetian or Genoese traders. After the mid-fifteenth century, when the Black Sea markets could no longer be reached, Florentine galley masters began to import more blacks from markets in Catalonia and Portugal.67 The Black Death changed the situation drastically in the middle of the fourteenth century and created a heightened demand for slaves. The plague probably originated in central Asia and was brought to the Crimea across the Asian steppes with the caravan trade. It reached the Black Sea in 1346; two years later it entered Italy and spread to most of the rest of western Europe. Probably onequarter to one-third of the European population died in a three-year period.68 The consequences were catastrophic, and one was an increase in slavery. Because of the high death rate, the workers who survived easily found good jobs in the countryside or in the cities. They could not be induced to become household servants. Death had not spared the elite, but the rich who remained had a larger supply of money available, as fortunes were consolidated through inheritance. It is not surprising that slavery grew as a result and supplied supplemental workers and domestics. In 1363 the Florentine government allowed the unrestricted importation of slaves from outside Italy, stipulating only that they be of non-Christian origin.69 Venice took a different course, probably because of the large influx of slaves over the preceding two decades, and

66  Origo, “Domestic Enemy,” p. 336. 67  Michael E. Mallett, The Florentine Galleys in the Fifteenth Century, with the Diary of Luca de Maso degli Albrizzi, Captain of the Galleys, 1429–1430 (Oxford, 1967), pp. 114, 127–29. 68  On the Black Death, see William H. McNeill, Plagues and Peoples (Garden City, N.Y., 1976), pp. 151–70; William M. Bowsky, ed., The Black Death: A Turning Point in History? (New York, 1971); Philip Ziegler, The Black Death (New York, 1969). The Black Death in England had a notable impact on the conditions of rural workers. In its aftermath the remaining peasants were able to throw off the last vestiges of serfdom. See R. H. Hilton, Bond Men Made Free: Peasant Movements and the English Rising of 1361 (London, 1973); id., The English Peasants in the Late Middle Ages (Oxford, 1975). 69  Origo, “Domestic Enemy,” p. 324.

Slavery In Late Medieval Europe

689

prohibited slave auctions in 1366, no doubt to reduce sales. Venetians could still import slaves, but sales had to be by private contract.70 Tatars, Russians, and Circassians continued to enter Genoa in this period, but their numbers declined in the second half of the fifteenth century, as access to the Black Sea diminished. In response, the Genoese turned back to the western Mediterranean. Muslim slaves in increasing numbers came to be purchased in North Africa, Spain, and Portugal. Because of the varied ethnic structure of Islam, the Genoese divided their captives into several categories: white, black, indaco (indigo), lauro (mulatto), and olivegno (olive). Whereas black slaves were fairly common in southern Italy and Sicily, relatively few of them reached Genoa or other northern cities. Domenico Gioffrè found them mentioned in only ten of the 1,600 notarial acts referring to slaves in fifteenthcentury Genoa. A few Guanche slaves from the Canary Islands also found their way to Genoa.71 With the recruiting grounds for non-Christian slaves reduced by the political changes in the greater Mediterranean region, but with demand for domestic servants still high, the Italians turned to indentured servants. The Venetians were in the vanguard in fostering this temporary servitude. Euphemistically called “souls” (anime), indentured servants were predominantly young, purchased by Venetian shipmasters from their parents in Dalmatia, Albania, Istria, and Corfu. Their legal situation distinguished them from ordinary slaves. Bound only for a defined period of service, after that they could buy back their freedom, if they had the money. Theoretically they could not be exchanged or sold to another master, but the Venetian government had to take steps to ensure that they were not taken to other Italian cities where they ran the risk of falling into perpetual slavery. The Venetians were not totally disinterested in keeping them from being taken elsewhere, for the prohibition was enacted partly “in view of the shortage of male and female slaves for serving our own gentlemen and citizens.”72 The existence of slavery in Renaissance Italy was pervasive. The account books and notarial records have preserved evidence of the presence of numerous slaves of various ethnic origins in the peninsula, and their existence also intrigued the portrait painters of the Renaissance. Yet as dramatic and well documented as their presence was, it was not important in a statistical sense. The Venetian census of 1563 listed only 7 to 8 percent of the population as

70  Lane, Venice, p. 133. 71  Gioffrè, Mercato degli schiavi, pp. 13–15, 17–22, 27, 31–32, 36–37. 72  Origo, “Domestic Enemy,” p. 332; Lane, Venice, pp. 332–33.

690

Phillips

servants, and this included free as well as slave.73 Within Europe in the late Middle Ages and the Renaissance, slavery was never crucial for social or economic development, and it was not practiced on any great scale in Europe proper. Rather it was Europe’s colonial areas—first the Atlantic islands and then the Americas—that would witness the rise of slavery to heights it had not reached since the end of the Roman empire. The motor behind that rise was plantation agriculture, of which sugar cane growing and sugar processing were the most important components. Because of their early acquisition of new lands, it was the Spaniards and the Portuguese who created the great expansion in sugar plantation agriculture.

Slavery in Spain and Portugal

Christian Spain and Portugal were part of the European world where slavery gradually declined as serfdom and free labor grew. For most of the early Middle Ages the Christian states of Iberia were too poor to purchase slaves in large numbers. Their economies were weak, far closer to the subsistence level than those of the Muslim states. As a result, the possibility of slave use in large-scale ventures producing goods and commodities for urban markets was restricted. Yet, unlike the rest of western Europe, the Iberian kingdoms were frontier states, sharing borders with non-Christian states whose inhabitants could be raided and enslaved with complete legality. This meant that slavery persisted there longer and more vigorously than elsewhere in Christian Europe. When the Iberians began their overseas expansion in the fifteenth century, slavery was a living institution that could be transplanted with ease to the Atlantic and Caribbean islands and the American mainland. In the early phases of the Christian reconquest, from the eighth to the twelfth century, there had been two distinct slave systems at work in the Christian states. One was a continuation of traditional Roman and Visigothic slavery. Its victims were primarily Christians, whose conditions were virtually identical to those of the Visigothic slaves. But Christian slaveholding, except in the isolated Asturias in the northwest, gradually underwent decline. Slowly but resolutely the slaves and their descendants blended into the “free” class of rural peasants. Since the reconquest and repopulation of newly acquired territories created a need for rural labor, peasants could almost always secure favorable legal and tenurial conditions in return for their participation in the resettlements. By the twelfth century the use of Christian slaves had almost ended, but slavery 73  Lane, Venice, pp. 332–33.

Slavery In Late Medieval Europe

691

as a system persisted. Even during the early period, numbers of Muslim slaves had been present in the Christian states, but after the twelfth century most of the slaves were Muslim prisoners of war, captured in the fall of conquered cities and in the course of Christian raids into Muslim territory. In fact, many raids had the capture of slaves as their principal goal. As one example, after the eleventh-century reconquest of Avila, some two hundred Muslim slaves were put to work in chains to build the town’s famous walls.74 During Portugal’s reconquest in the twelfth and thirteenth centuries, the Portuguese crown and nobles conformed to general Iberian practice and employed Muslim prisoners of war as slaves. But after the Algarve (southern Portugal) was absorbed early in the thirteenth century, Portugal no longer bordered on Muslim territory. Although at times the Portuguese crossed into Castilian territory to participate in raids on Granada and gained slaves there, the most fruitful source of slaves was Africa. There the Portuguese purchased slaves or raided to secure them. Even before they began their maritime expansion down the African coast in the fifteenth century (which we will examine in detail later), African slaves came to be used in Portugal. In 1317 King Diniz of Portugal gave command of the kingdom’s fleet to a Genoese, Manuel Pesagno, who was licensed for privateering in Moroccan waters and could retain onefifth of all the slaves captured.75 In the central kingdom of Castile during the twelfth, thirteenth, and fourteenth centuries, slaves were almost exclusively Muslim in origin. Positive and negative factors account for this. Muslim lands lay on Castile’s southern boundaries, and border raids and larger military engagements provided numerous captives. As one example, after the significant Christian victory over the Muslims at Las Navas de Tolosa in 1212, several thousand defeated Muslim warriors entered the market as slaves. The normal procedure was to offer the prisoners of war the possibility of repatriation if ransoms were paid; if they were not redeemed they could be sold as slaves. Before the fifteenth century, Castile was not an important Mediterranean power; therefore the wide range of ethnic groups sold as slaves in the Mediterranean reached Castile only infrequently.76 74  Charles Verlinden, L’esclavage dans la Europe médiévale, vol. 1: Péninsule ibérique—France (Bruges, 1955), 1:105–06, 108–09, 116–18, 122, 128–29, 135–38, 172–74,; J. Font Ruis, “La sociedad en Asturias, León y Castilla en los primeros siglos medievales,” in Historia social y econômica de España y América, ed. Jaime Vicens Vives, 5 vols. (Barcelona, 1957–), 2:319–20. 75  Verlinden, Esclavage, 1:546–47, 550; Verlinden, Modern Colonization, pp. 38–39. 76  Verlinden, Esclavage, 1:548–49, 551–59, 589, 593.

692

Phillips

After the thirteenth-century conquest of the cities of the south, far more Muslims remained in Christian territory than could be profitably used as slaves, and large populations of free Muslims came to be governed by Christian rulers. As a result, free Muslims were more numerous than Muslim slaves, who tended to be concentrated in the frontier regions where raids ensured a steady supply. They were less numerous, though not nonexistent, in the parts of Castile more distant from the frontier.77 The Christians themselves were not immune from Muslim raids and could be enslaved in Muslim territory. There they were offered for ransom, and Christian rulers and subjects sought to raise and deliver the ransoms. In the twelfth century Alfonso VIII of Castile entrusted redemptionist efforts to the military religious orders. By the thirteenth century, the Trinitarians and the Mercedarians had ransoming as their main activity, and they coordinated fund raising for that purpose. Cities such as Cuenca regularly taxed their citizens to raise funds for ransoms, and the monastery of Santo Domingo de Silos sent agents to ransom captives in the Muslim cities of the kingdom of Granada and in North African centers such as Tangier and Ceuta.78 If Castile was cut off from the Mediterranean until late in our period, the lands of the Crown of Aragon—Aragon proper, Catalonia, and later Valencia— were open and receptive to the influences of the inland sea. Benjamin of Tudela, a Spanish Jew who made a fabulous journey around much of the known world and described it in prosaic terms, found the port of Barcelona in the mid-twelfth century packed with ships from around the Mediterranean, with cargoes—according to Jaime Vicens Vives—consisting mainly of gold and slaves. The trade in gold and slaves predated the cloth trade, the late Middle Ages’ great generator of commerce. For Vicens this traffic in humans and precious metals explains why in 1062, for example, the merchants of Barcelona owned slaves, why in 1104 Ramón Berenguer III collected a tithe on the profits from the capture and sale of slaves, and why in 1148, when Ramón Berenguer IV was planning the conquest of Tortosa, he found support among … the bourgeoisie of Barcelona…. All this justifies our thesis that between the ninth and the eleventh century Barcelona was one of

77  Verlinden, Modern Colonization, pp. 37–38. 78  James W. Brodman, “Military Redemptionism and the Castilian Reconquest, 1180–1250,” Military Affairs 44 (1980): 24–27; Verlinden, Esclavage, 1:606–10.

Slavery In Late Medieval Europe

693

the great points of contact in the western Mediterranean between the Christian world on the one hand and the Muslim world on the other.79 All the customary ways of enslaving were prevalent in the Crown of Aragon: by birth, marriage, judicial sentence, and debt. Until the end of the thirteenth century, conquest and raids supplied many slaves; thereafter Aragon had completed its mainland expansion, and the chief sources of slaves were piracy and trade.80 The great period of Aragonese expansion was the thirteenth century, when James I and Peter of Aragon took over the kingdom of Valencia and cooperated with Castile in the conquest of Murcia. Prisoners of war were regularly sold into slavery, and James I was particularly well supplied with slaves. He sent some two thousand as gifts to kings, emperors, the pope, the cardinals, and nobles.81 The capture of large numbers of slaves was a continuing tradition. In 1280, after Peter took Montesa, “slavers continued for at least a year and a half their purchases among the multiple prisoners of war….”82 Slavery in the newly conquered kingdom of Valencia conformed to the normal Mediterranean pattern—urban rather than rural, domestic rather than agricultural, even though a minority of the slaves certainly lived and worked in the countryside. There was a wide spectrum of slave owners; Robert I. Burns has identified slaveholders ranging from artisans to bishops to the king.83 Strong legal limitations bounded the existence of the slaves in the thirteenth-century kingdom of Aragon. Their legal status was almost nil, though they could appear in court as plaintiffs or defendants in special circumstances. Muslim slaves could be ransomed or purchase their freedom, but if they remained as slaves they were subject to the domestic jurisdiction of their owner, except that those accused of serious crimes were judged in the royal courts, as were cases of disputed servile status. A century later localities had promulgated harsh regulations for slaves, based on the assumption that slaves were

79  Jaime Vicens Vives, with Jorge Nadal, An Economic History of Spain, trans. Frances M. López-Morillas (Princeton, 1969), pp. 148–49. 80  Verlinden, Esclavage, 1:251–73. 81  Robert I. Burns, S. J., Islam under the Crusaders: Colonial Survival in the Thirteenth-Century Kingdom of Valencia (Princeton, 1973). p. 111; Robert I. Burns, S. J., The Crusader Kingdom of Valencia, 2 vols. (Cambridge, Mass., 1967), 1:2, citing Muntaner, Crónica catalana, ch. 13. 82  Burns, Islam under the Crusaders, p. 351. 83  Ibid., pp. 109–10, 407; Burns, Crusader Kingdom, 1:22, 128.

694

Phillips

rebels and prisoners. Prohibitions on their drinking in taverns and public places kept them segregated from the general population.84 So far we have been concerned with Christian slaveholders, but slave owning by non-Christians continued in a restricted fashion. After the conquest of Valencia neither Jews nor Muslims could hold Christian slaves, but it is quite likely that Muslims continued to own Muslim slaves. There was a constant attrition as Muslim slaves fled their Muslim masters to accept baptism (which did not mean automatic manumission), and this loss to local Muslim slaveowners could not be made up. Muslims were, after the conquest, cut off from the Muslim slave trade, and their worsening economic situation meant that they could not easily purchase slaves from Christian suppliers. Free Muslims, too, often ran the risk of being enslaved. Merchants from the Crown of Aragon took slaves to southern France, and Muslim envoys who visited Barcelona often bought Muslim slaves there. The economic benefits Christians could derive from the slave trade meant that they used legal and extralegal means to increase the supply of Muslims to put on the market.85 Because of the exposed Aragonese coastline, Christians were in danger of being carried off by Muslim pirates and being held for ransom. The numbers of captives grew so large that the crown and the church began to organize means of ransoming the hostages. A crown official, the exea, was charged with coordinating these activities, but what the government could do was limited. As a result, the church assumed a major portion of the burden. The military orders in the Crown of Aragon were active in the effort, as in twelfth-century Castile, but the new redemptionist orders were more important. The most prominent was the Mercedarian order (the Order of Our lady of Mercy), founded in the 1230s by St. Peter Nolasco, who himself was credited with the redemption of some fourteen hundred captives. From their motherhouse in Barcelona, the Mercedarians spread out along the Christian frontier and into Castile, as we have seen. They were present at the capture of the city of Valencia, and, once the conquest was completed, they expanded their activities to the Castilian

84  Charles Emmanuel Dufourcq, La Espagne catalane et le Maghrib aux XIIIe et XIVe siècles (Paris, 1966), pp. 72–73; Burns, Islam under the Crusaders, pp. 109–10, 250, 252, 260; Verlinden, Esclavage, 1:292–95. 85  John Boswell, The Royal Treasure: Muslim Communities under the Crown of Aragon in the Fourteenth Century (New Haven and London, 1977), pp. 49–56, 313, 316–17; Burns, Islam under the Crusaders, pp. 111–12, 128; Burns, Crusader Kingdom, 1:115; Verlinden, Esclavage, 1:291–92, 300.

Slavery In Late Medieval Europe

695

frontier with Granada and to North Africa, aided in their efforts by pious donations.86 By the end of the thirteenth century, the Iberian Peninsula no longer occupied the unique position regarding slavery that it had held during most of the middle Ages. Portugal had completed its reconquest and was still a century away from beginning its African conquests. Castile still had Muslims across its southern frontier, but outside the border regions slavery was on the decline. For both Portugal and Castile, slavery would only experience an upsurge when Africa and the Atlantic became important theaters of action in the fifteenth century, a topic we will treat in later chapters. For the Crown of Aragon, the conquest of Valencia, together with the joint Aragonese-Castilian conquest of Murcia, precluded further expansion within the peninsula and deprived them of a Muslim presence across their frontier. Thereafter, the traders and seamen of Catalonia and Valencia turned their complete attention to the Mediterranean, where their forebears had been active for centuries. One of the most important preconditions for the emergence of New World slavery was the recovery and assimilation of Roman law into the legal systems of the medieval kingdoms. The elaborate slave laws contained in the Code of Justinian could easily be applied when new conditions called for them. In the eleventh century the Italian legal scholar Irnerius had begun the academic study of the Roman code, and in the next two centuries knowledge and application of Roman law spread widely in western Europe. In Iberia the Castilian king Alfonso X in the mid-thirteenth century produced a new code for his kingdom, known as the Siete Partidas, with heavy influences derived from Roman law. Although it never fully became law in Castile, the Siete Partidas still had a significant influence on late medieval and early modern legislation in Spain, both for the home country and for the American colonies, and thereby ensured that many Roman rules for slavery entered Spanish law.87 By the end of the Middle Ages, Iberia had ample historical experience with slavery and a legal code for operating a slave system. The third element necessary for a great expansion of slavery was sugar cane agriculture, also well known to the Iberians. The peninsula had long been a sugar-producing region. From the time of Abd ar-Rahman I, who ruled Islamic Spain from 755 to 788, sugar cane was grown in the southeast. Abd ar-Rahman was a member of the recently overturned Umayyad dynasty and had escaped Damascus following the Abbasid 86  Burns, Crusader Kingdom, pp. 248–52. 87  J. A. Doering, “La situación de los esclaves a partir de las Siete Partidas,” Folia Humanistica 4 (1966):337–61; E. N. van Kleffens, Hispanic Law until the End of the Middle Ages (Edinburgh, 1968).

696

Phillips

revolution. After fleeing across North Africa to Morocco, he was able to unite sufficient political and military support to set himself up as the head of an independent emirate in Spain. In an attempt to build Spain’s economy, he sent for seeds and cuttings from the Levant and introduced a number of new crops into Spain, including sugar cane. Like many other plants, sugar cane was first grown in Spanish soil under carefully controlled conditions in royal gardens, initially in Córdoba and later in Toledo and Almería, where the water requirements were carefully monitored to ensure that the best growing conditions could be determined. By the tenth century, cane prospered in southern Spain, and, at the beginning of the fourteenth century, sugar from the Muslim city of Málaga was sold as far away as Flanders. Sugar from Morocco at the same time was exported to England.88 Sugar could be produced readily in southern Spain, but attempts to spread its cultivation northward were relatively limited until the late Middle Ages. King James II of Aragon (1291–1327) undertook to grow sugar in Valencia, the kingdom on Spain’s eastern Mediterranean coast. To do so he secured plantings from Sicily together with a Muslim slave skilled in sugar techniques. It is interesting to note that the second wife of James II was Mary of the Cypriot ruling house of Lusignan; part of her dowry was paid in sugar and sold by Barcelona merchants at the behest of the king. After James II’s initiative, sugar took quite a time to prosper. The first production mentioned was in the 1380s. In 1433 the cathedral chapter of Valencia attempted to secure the right to collect a tithe on sugar production, and in the process they provided an important account of the growth of production in Valencia. The canons reported that since the 1380s and 1390s both Christian and Muslim farmers had planted sugar cane as a secondary crop and sold their cane in its raw state as a delicacy for children and adults. In the second decade of the fifteenth century, planting increased as grain gave way before cane in many places. In 1407 the Valencian government gave financial aid to Nicolau Santafé, a sugar expert, to set himself up in Valencia. The first evidence of a mill in Valencia comes from 1417, when the master potter Thahir Aburrazach contracted to move to Burriana and make ceramic forms and vessels needed for the mill (trapig de les canyes mels in the Valencian dialect) owned by the merchant Francisco Siurrana. Nobles, too, constructed mills. The knight Don Galceran de Vich built one at Jeresa and later another at Gandía, and the soldier-poet mosén (Sir) Ausias March owned 88  Thomas F. Glick, Islamic and Christian Spain in the Early Middle Ages (Princeton, 1979), pp. 77–78; S. M. Imamuddin, Some Aspects of the Socio-Economic and Cultural History of Muslim Spain, 711–1492 A.D. (Leiden, 1965), pp. 88–89; Pérez Vidal, Cultura de la caña, pp. 11–13.

Slavery In Late Medieval Europe

697

a sugar plantation and later built a mill on his property. By the 1430s numerous mills were in full operation. This expansion in Valencia took place just at the time when eastern Mediterranean sugar production was faltering and commerce there was being threatened. In these circumstances it is obvious why Valencian sugar should draw the attention of a large German merchant house, the Ravensburger Handelsgesellschaft. After its foundation in the 1380s the company—which, unlike other German houses such as the Fuggers and the Welsers, concentrated on commerce and avoided banking—expanded rapidly and established branches in various parts of Europe. By 1420 they had agents and warehouses in Valencia and exported Valencian products, including sugar. The production and export of Valencian sugar generally increased in the first half of the fifteenth century, and by 1460 the company’s profits were great enough to encourage them to enter the production side. They acquired land from Hugo de Cardona along the river Alcoy near Gandía and built a mill and refinery managed by maestre Santafé, probably the son of Nicolau de Santafé. The manufacturing complex prospered at first, and the quality of sugar produced there was extremely high. But in the 1470s difficulties intervened. A lawsuit involving Hugo de Cardona slowed production, and this coincided with the company’s loss of some of its markets, difficulties in transportation, and competition from Madeiran sugar. The company’s directors sold the facility in 1477 and a few years later rejected a proposal to reopen it.89 Portugal also played a vital role in the spread of sugar production. In the later Middle Ages, sugar was being produced with some profitability in the Algarve (southern Portugal), and the desire to tap the wealth of Morocco’s sugar plantations may have entered into the motivations for Portuguese expansion in Africa.90 Beyond this, the major importance of Portugal and Spain in the expansion of sugar production came as a result of their discovery and exploitation of the islands of the eastern Atlantic. The Azores and the Cape Verdes, to a lesser extent, and Madeira and the Spanish-held Canaries, to a much greater extent, were to be staging areas on the eve of the discovery of America. Before concluding, we should examine the methods of growing sugar cane and of obtaining marketable sugar from it. The cultivation of sugar cane remained more or less constant in the Mediterranean and Atlantic growing regions. Cane was planted from cuttings of mature stock, not from seed. The 89  Pérez Vidal, Cultura de la caña, pp. 14–15, 37–38, 41–45. 90  Antonio H. de Oliveira Marques, History of Portugal, 2 vols. (New York, 1972), 1:140; Henrique Gomes de Amorim Parreira, “História do açúcar em Portugal,” Anais: Estudos da História da Geografia da Expansão Portuguesa 7 (1952): 18.

698

Phillips

“setts,” as these roughly foot-long cuttings were called, needed a slightly acid soil, and cane did best with a rainfall of some sixty to seventy inches per year. Few, if any, Mediterranean lands had that much rain; therefore irrigation projects were used to increase the water supply to the setts, which were planted at the bottom of the furrows of the plowed fields to receive the maximal amount of water. When the plants reached maturity, the long stalks were cut back near the ground, leaving the bottom (or ratoon) in the ground to produce the next crop. The next stage was the transportation of the cane from the field to the mill, carried on the back of draft animals, in carts, or by humans. At the mill it was cut into appropriate lengths and crushed, with the juice dropped from the crushed cane collected for refining. The fresh juice contained a small percentage of foreign matter that had to be removed if the sugar were to attain its pure form. Boiling the juice with an additive made the separation of the pure sugar easier. In the late Middle Ages natron and potash were commonly used as clarifying additives, as the juice was boiled in successive stages to drive off the water, until it was ready to be dried in loaf-shaped molds of pyramidal or cone shape.91 In the fourteenth century the Italian Pegolotti, in his famous manual of commerce La Practica della Mercatura, described the varieties of sugar on the market. The best was mecchera, in pyramidal shape, white and dense. Next in order came caffetino, shaped like a cone with a rounded top; then bambillonia, musciatto, and domaschino, the least preferred. Powdered sugar was pressed into the form of a loaf, but it usually turned to dust during transport. Perfectly transparent candy sugar was also produced and sold.92 Sugar found a ready market as European demand constantly expanded. The Mediterranean producers could not keep up with that demand, and they had to face the competition of sugar coming from the Atlantic islands and from America in the fifteenth and sixteenth centuries. These two new regions were where the complete identification of sugar with slavery would take place, a development we shall examine in later chapters.

91  Galloway, “Mediterranean Sugar Industry,” pp. 182–88; L. A. G. Strong, The Story of Sugar (London, 1954), pp. 22–31. 92  Francesco Balducci Pegolotti, La Practica della Mercatura, ed. Allan Evans, (Cambridge, Mass., 1936), pp. 297, 362–65.

The Identity of the Slave in Scandinavia Ruth Mazo Karras Part of the definition of slavery in chapter 1 stated that the slave is always in some sense an “other” to those who dominate the society. The medieval historian cannot help adopting this viewpoint because the sources do not allow us to see medieval culture through the eyes of anyone outside the small group who left records. This does not mean that we must accept the dominant culture’s view as an accurate representation of the reality of its age. We can examine the way in which the cultural elites constructed their world, as reflected in the various kinds of texts that have survived; from this we can make inferences about the world views of other groups and about what material reality was like. We must always remember that the sources we have to work with are artifacts but not mirrors of the society and culture that produced them (see the appendix for a discussion of some considerations in interpreting particular kinds of texts). A slave system requires a supply of slaves, and the source of that supply will affect the way that culture constructs slavery. There are many different ways in which slaves are “others” or outsiders. Masters may locate the slaves’ otherness in their different ethnicity, their lack of membership in the dominant race, people, or nation; in their moral qualities, perceived slavish behavior or lack of courage or virility, particularly for slaves captured in war; or in their birth to other slaves in a society that puts a premium on genealogy or hereditary status. The otherness may also be located in the tasks the slaves perform, if particular kinds of work are considered appropriate only for slaves, or in the slave’s legal disabilities or lack of legal rights. Particular duties or legal disabilities are assigned to the slave because of his or her enslavement, but contemporaries may see the causal relationship the other way around, particular kinds of work or particular legal disabilities giving rise to particular moral or personal qualities perceived as justifying enslavement. This chapter deals with the identity of the slaves in two senses: where they actually came from, and how the culture constructed an identity for them as slaves.

Source: Karras, Ruth Mazo, “The Identity of the Slave in Scandinavia,” in Ruth Mazo Karras, Slavery and Society in Medieval Scandinavia, New Haven: Yale University Press, 1988, 40–68.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_024

700

Karras

The Terminology of Slavery

By any definition, including those discussed in chapter 1, the institution called þrældom in Scandinavian sources is slavery and the person called a thrall is a slave.1 Of course, it is the way the thralls appear in law and saga that makes it clear to us that our term slave is applicable to them; it could be that the laws about them were written as they were because the redactors thought it appropriate to identify the thrall with the Roman servus and apply the same sorts of laws (see chapter 4). The influence of Christianity and Latin learning upon Scandinavian culture could have transformed the conception of the thrall.2 We do not know anything about the status of the thrall before the time when the sources were written; since all the sources make thralls slaves, we must conclude that they were so by the time under discussion here. The sources also use other terms for slavery, none of them unambiguous (even the word thrall can appear in a figurative or pejorative sense referring to someone who is not actually a slave). Many of the terms are used interchangeably, and the meaning in each case depends on the context.3 When the sources are in Latin, we must ask what vernacular terms the writer had in mind. Danish and Swedish diplomatic material uses Latin until the fourteenth century, while in Norway and Iceland most of this material is in the vernacular. Even in Denmark and Sweden law codes were written in the vernacular. When people made a will or agreed to sell land or granted a privilege they probably did so in their native tongue; translation by a clerk to the nearest Latin equivalent might not convey the exact meaning they intended. One cannot assume that a given Latin term like curia, colonus, or servus meant exactly what it would mean in an English or Continental context. One way of determining which terms in Latin and Old Norse meant the same thing is to look at direct translations. One of the extant law codes, that of Skåne, exists in a Latin paraphrase (see appendix). Surviving vernacular Bibles from Scandinavia, like the Old Swedish paraphrase of the Pentateuch from the beginning of the fourteenth century or the Norwegian Bible paraphrase Stjórn

1  Hereafter the English derivative thrall will be used for the Old Norse forms þræll, træll, etc. 2  See Prakash, “Terms of Servitude,” for a discussion of how terms from their own language applied by conquerors to agrarian groups in a colonized culture can construct a new version of social reality. 3  Throughout this work, whenever I have quoted or cited a source referring to slaves, I have noted what term appears in the original if (1) the meaning is at all ambiguous or (2) the meaning is unambiguous but the term used is unusual.

The Identity Of The Slave In Scandinavia

701

from the same period, provide another opportunity to see what terms were considered equivalent by comparing the terms used with those in the Vulgate.4 Thrall—Olc þræll, and variants in the other Old Norse languages—is the most common term for “slave” in the law codes and in the sagas. The thrall is the one bought, sold, manumitted, and compensated for with payment to the master instead of the kin. The etymology of the word provides little information about the origins of slavery or possible connotations. It has few cognates in other Germanic languages: the English thrall (OE þræl, ME thrawl) is derived from the Norse, though the OHG drigil is cognate. The word seems to derive from a PON *þrahilaR, PGmc *þrǎhila or *þreghila, “run” (giving also Go. þragjan), derived ultimately from PIE *trek-, “run,” “pull,” “draw.”5 The original connotations of thrall—servant or messenger—were presumably those of work rather than of status. The Old Swedish paraphrase of the Pentateuch often uses thræl to translate the Latin servus, though in other places it translates servus and ancilla as thiænisto swen and thiænista qwinna (serving man and woman), even when they are commodified by being given as gifts. It sometimes renders servus as man, though this occurs where the word seems to refer to a retainer. Where it seems clear that the Latin servus refers to slaves rather than servants, in legal provisions for buying and selling Hebrew and foreign slaves and freeing them in the sabbatical year, the Old Swedish Pentateuch uses the terms thræll and amboot. In one passage where mercenarius (hired man), servus, and ancilla appear in the same sentence, it uses thiæniste swen to translate mercenarius.

4  Slavery still existed in Sweden at the time of the Old Swedish Pentateuch. Though it was not current in Norway by the time of Stjórn, the terms used there still indicate what statuses Norwegians equated with Latin terms. Servus is rendered more consistently as þræl there than in the Swedish text, but sometimes it seems that the translator felt he had to explain: in 2 Kings 4:1 (Stjórn 612) “til þanar oc þionkanar oc þrældoms” for the Vulgate’s “ad serviendum” indicates that thrall might have lost some of its specific meaning. Belsheim, “Bibelen paa norsk-is-landsk,” has collected Biblical citations from other texts but few of them are passages dealing with servitude. 5  Michalsen and Štech, “Det slaviske ordet ‘otrok,’” 235; Hellquist, Svensk etymologisk ordbok, s.v. “träl”; Oxford Dictionary of English Etymology, s.v. “thrall”; Holthausen, Wörterbuch des Altwestnordischen, s.v. “þræll”; Falk and Torp, Norwegisch-Dänisches etymologisches Wörterbuch, s.v. “træl”; Marstrander, “Træll”; De Vries, Altnordisches etymologisches Wörterbuch, s.v.” þræll.” De Vries also suggests the possibility that the word goes back to a PGmc *þranhila, with a meaning of “forced,” but the evidence of the cognates argues against this.

702

Karras

The West Norse version, Stjórn, uses a similar mix of terms to translate servus, though þræll is more common here than in the Swedish.6 While Latin servus in the Bible has various Old Norse translations, depending on the context, in Latin sources written in Scandinavia it usually seems to be a translation of thrall and to mean slave. Andreas Suneson used it to mean slave throughout his Latin paraphrase of the laws of Skåne. When Innocent III made a pun about how servi who had committed a crime should not be exempted from a trip to Rome for absolution, since servi should be absolved by the servus servorum, the men to whom he referred were clearly slaves. He also called them mancipii.7 Though servus in the twelfth through fourteenth centuries elsewhere in Europe may be translated as serf (see chapter 1), it did not have this meaning in Scandinavia. Testamentary manumission of Scandinavian servi indicates their unfree status, but no documents refer to servi in the context of land tenure. When servus did not have the specific meaning of slave, it had the more general one of servant, but in any case did not mean serf. There was no group in medieval Scandinavia who could be called serfs; the landbo or leiglending, called a colonus in Latin, was juridically free, not under his landlord’s jurisdiction although he had contractual obligations to him.8 The term is best rendered into English as tenant, although it means a tenant of a specific sort. Landbo or colonus is not often found in direct opposition with thrall or servus. The former are terms of economic status and the latter terms of legal status. Often connected with colonus is villicus, translated in the vernacular as bryte or some variant. This word could be rendered into English as steward. By the thirteenth century it generally seems to refer to some sort of royal official, but in some of the earlier law codes it refers to a slave who supervises the master’s household.9 Ambátt and its variations usually mean female slave; ambátt is the female equivalent of thrall. It is sometimes used of women who are concubines but not necessarily slaves, but its use there seems to be derogatory rather than denotative (see chapter 3). The Latin equivalent is ancilla, which, when it appears 6  For thræl, e.g., Gen. 9:25 (Svenska medeltidens Bibel-arbeten 1:173; Stjórn, 63, has þræll ok pionustumadr); for thiænisto swen, e.g., Gen. 24:35 (Svenska medeltidens Bibel-arbeten 1:205, though Stjórn, 137, has þræla ok þionustukonur); for man, Exod. 12:30 (Svenska medeltidens Bibel-arbeten 1:311, where Stjórn, 281, has þionustumenn). For mercenarius, Exod. 21:20 (Svenska medeltidens Bibel-arbeten 1:336; Stjórn, 302); Lev. 25:6 (Svenska medeltidens Bibelarbeten 1:372). 7  Innocent III to Andreas Suneson, Archbishop of Lund, 13 January 1206, dd 108, 1:4:213–14. 8  L.-A. Norborg, KLNM, s.v. “Landbo”; H. Bjørkvik, KLNM, s.v. “Leiglending.” 9  See Riis, “Villici og Coloni,” 1–20, for the use of these words in documents from Denmark. See also chapter 3.

The Identity Of The Slave In Scandinavia

703

in Scandinavian law codes or wills, is usually the female equivalent of servus, meaning slave.10 Like thrall, ambátt and another word for a female slave, Olc þý, have no etymological connotations of legal status, ethnic origin or captivity. Olc þý is cognate with OE þeow, “slave,” and Go þius/þiwi, “servant, female servant,” going back to a PIE *þihu- “hurried”; ambátt is either cognate with or derived from Celt ambactus, a slave or dependent, with connotations of “one who serves” and with Ger Amt as its modern relative.11 Both these terms appear in other Germanic languages in the masculine but in Old Norse only as feminines. Perhaps the Old Norse words are not cognates but rather borrowings into Old Norse or proto-Old Norse from foreign languages. It may be that at some linguistically formative period most of those enslaved in wars or raids were women and their own word for slave was adopted into Old Norse as a feminine. Male slaves, for whom the various forms of thrall were used, would have been not captives but their descendants. One word in the vocabulary of Scandinavian slavery, OS annöþogher, OD annøþoghær (Olc ánauðigr, but the word is much less common than in East Norse), does have an etymology that sheds some light on what the culture saw as the distinctive feature of slavery. The word comes from some relative of the Old Icelandic ánauð, “oppression,” and though it denotes slave the connotation is that the person has been enslaved or forced to labor.12 The term occurs frequently in the Swedish and Danish provincial law codes of the thirteenth and early fourteenth centuries. By this time it may have lost its root meaning and come to be synonymous in connotation as well as denotation with thrall, but it does provide a hint that at some stage the distinguishing characteristic

10  On Saxo’s use of this term to mean female slave, see Steenstrup, “Saxo Grammaticus,” 106. Glossarium til medeltidslatinet i Sverige, s.v. “ancilla,” gives only the meaning tjänarna (female servant), but offers no citations (except for ancilla dei), ignoring testamentary manumissions of ancillae. An Icelandic reference to Ps. 123:2 uses ambátt for the Vulgate’s ancilla: Heilagra Manna Søgur 2:215. 11  Brugmann, “Zu den Benennungen,” 381 and 390; De Vries, s.v. “arnbátt”; Uhlenbeck, Kurzgefasstes etymologisches Wörterbuch, s.v. “andbahti”; Grimm, Deutsches Wörterbuch, s.v. “Amt.” 12  Cleasby-Vigfusson, s.v. “ánauðigr”; De Vries, s.v. “nauð”; Schlyter, Ordbok, SGL 13, s.v. “annøþogher”; Holmbäck and Wessén, Svenska land-skapslagar 5:43 n. 40; G. F. V. Lund, Det ældste danske skriftsprogs ordforråd, s.v. “annøþoghær.” It appears as a nickname in Lnb, S351, where a man is called “Ormr ánauðgi” (which Pálsson and Edwards, Book of Settlements, 133, translate as “Orm the Unfree”) although he does not appear to be a slave. There may be a story behind his name that Landnámabók does not tell.

704

Karras

of slaves was their powerlessness and domination by force rather than their ethnicity or their function as servants. It is surprising that in one instance the OS Pentateuch should use annœdogha to translate famula, since famulus and famula generally seem not to mean slave.13 In wills from Denmark and Sweden, if servi or ancillae receive anything it is their freedom, while famuli or famulae receive gifts—clothing or horses—and famuli sometimes serve as witnesses. A famulus could be a member of a bishop’s familia; a famulus also appears in municipal law codes as a representative who carries on his master’s business.14 The term seems to have a generalized meaning of servant rather than the specific meaning M. M. Postan has suggested for twelfth- and thirteenth-century England.15 Swen (Olc sveinn) seems to mean “servant” in general, either free or unfree. It appears in the OS Pentateuch to translate servus, when servus refers to a respected manager who does not appear to be bought or sold.16 In a few instances swen clearly refers to a slave: Archbishop Absalon of Lund manumitted his bathsuen, Eskil, in his will.17 A legoswen, on the other hand, is clearly a hired man; the OS Pentateuch uses this word to translate mercenarius.18 There are many other words whose main meaning is that of service and not specifically slave status. Hjon or hion appears in many of the law codes to mean members of the household, either slaves or free servants, but legohjon clearly means a hired servant. Huskona, too, appears in the Danish laws referring to an unfree woman, but does not always imply unfreedom. Two other words often found in Icelandic sagas sometimes refer to men who are also called “thralls”: man and huskarl. Someone’s man is usually a slave, but not always; a huskarl or retainer is usually but not always a free man. No one can be identified as slave

13  Exod. 21:7 m the Vulgate reads “si quis vendiderit filiam suam in famulam non egredietur sicut ancillae exire consuerint” (“If someone sells his daughter as a famula she shall not go out as the ancillae do”). Svenska medeltidens Bibel-arbeten 1:335 uses annædogha for famula and ambot for ancilla. Stjórn, 302, has “if a man sells his daughter as an arnbátt she shall not go out as other ambáttir do.” Both Norse texts suggest that they did not make the fine distinction the Latin did between types of servants; all were conceived of as unfree, so the same terms could be used to translate all. 14  Wills are found throughout dd and ds. For municipal law codes, see that of Tonder (1243, based on that of Lübeck), 10, dgkl 1:217. 15  Postan, “The Famulus.” 16  Gen. 24:2 (Svenska medeltidens Bibel-arbeten 1:203; Stjórn, 134, here uses radmaðr). On the term in Icelandic sources see Krag, “Trelleholdet,” 366–67. 17   d d 32, 1:4:62. 18  Lev. 25:40 (Svenska medeltidens Bibel-arbeten 1:375).

The Identity Of The Slave In Scandinavia

705

or free solely because he is labeled with one of these terms.19 The Norwegian laws also use the term mansmaðr (man’s man). By its etymology this need not mean a slave, but in no known example does it clearly indicate a nonslave.20 The etymological denotation is that of possession rather than of work or ethnicity. Some of the Swedish provincial law codes—those of the Svear provinces, dating from the late thirteenth and early fourteenth centuries—often use the term ofræls (unfree) instead of a specific word for slave. This may indicate a growing emphasis on freedom, a notion that it was the normal state, but it more likely represents the gradual dissolution of the simple slave-free dichotomy. A term for slave may no longer have covered all the unfree. This shift in terminology and its meaning for the society is discussed in chapter 5. An analysis of when these laws use unfree and when they use thrall is appropriate here, however, since we are dealing now with the connotations of specific terms. The Uppland law and the Södermanland law have only a few references to slavery, several of which occur in new ameliorative legislation possibly inspired by the Church (see chapter 5). The Dalarna law has hardly any; the Västmanland law has more, only some of which are taken from the Uppland law. The Västmanland law uses thrall in cases where it does not follow the Uppland law but sets different levels of compensation for slaves and free people.21 The Uppland and Södermanland laws use thrall in instances where it can only refer to slaves, such as sale as a chattel, but in the ameliorating legislation, which may be assumed to be new in 1296, unfree may have a wider meaning.22 Besides the provision on the selling of a Christian slave the other instance in which the Uppland law uses thrall instead of unfree is in the law providing for the torture of a slave who attacks his master.23 This law applies to anyone who 19  Johannesson, Die Stellung der freien Arbeiter, 100; Wilde-Stockmeyer, Sklaverei auf Island, 48–52. 20  Wilde-Stockmeyer, Sklaverei auf Island, 51; Krag, “Treller og trellehold,” 210n.; Sandnes, “Tolv kyr, to hester og tre træler,” 81. See also Cleasby-Vigfusson, s.v. “man.” 21  VmL Manh 24:8; VmL Manh 6:4; VmL Arf 17. 22  The Uppland law prohibiting the sale of a Christian slave and the Söderman-land law that followed it use thrall. UL Köp 3 and SdmL Köp 3. UL Arf 19, recognizing the marriage of two slaves, uses unfree, this is repeated, also using unfree, in VmL Arf 14 and SdmL Arf 3:2, though DL Gipt 4 uses thrall; UL Manh 6:5, providing for identical compensation for the killing of a slave and a free man, uses unfree; as does UL Arf 22, on the penalty for unlawful intercourse being the same. The Söderman-land laws setting levels of compensation for slaves and free people, different in some instances but the same in others, also uses unfree. SdmL Byg 24:1, and SdmL Manh 26:8. 23  UL Manh 15:1.

706

Karras

kills his rightful lord, and it mentions specifically a householder’s or steward’s hired or slave man or woman. It could be that the general rubric unfree would have covered more than just the slave; perhaps it would not have been clear exactly whom it included, so the law named the specific groups. Even if unfree did not have any meaning beyond slave and was merely a synonym for thrall, the fact that, rather than the term thrall, which might be thought of as a status having an existence of its own, a word was used that was nothing but the negative of free represents a break from the other laws. While the main feature of a slave’s status, as Norwegians or Icelanders perceived it, might have been his lack of freedom, he was also a member of a particular group, thralls, not just someone who did not have freedom. Slaves may have been a less clearly defined group in the society of the Svear laws, and freedom may have been thought of as a more tangible quality, the possession or nonpossession of which defined the person. A dichotomy between free and unfree is different from one between free and slave: the latter implies that slavery is something in itself, not just a lack. One reason for the use of a broader term for slave in the Svear laws might have been the influence of Latin. The Latin word servus had a broad range of meaning, which could lead to a certain vagueness of legal status. A Swedish writer, seeing that the Latin word servus sometimes meant servant in general, might have doubted whether thrall was the proper translation for it and preferred unfree with its greater vagueness. A striking absence, in light of the vocabulary of unfreedom in other European tongues (including the word slave itself), is the lack of ethnic terminology for Scandinavian slavery. The word sclavus, according to Verlinden, took on the juridical meaning slave as well as the ethnic meaning Slav in Germany during the tenth and eleventh centuries. This meaning fell out of use when the slave trade via Germany to Muslim Spain dropped off, but by the thirteenth century it became common in Italy, and by the fourteenth century in Spain.24 Yet the term does not appear in either Latin or vernacular documents up to the fourteenth century in Scandinavia, despite the presumed role of the Swedish Vikings in the trade in Slavic slaves to the Muslim world (see below). Nor does wealh, an Anglo-Saxon word for slave with the original meaning of Celt or Welsh, appear in Scandinavian languages, despite the presumed large-scale slave raiding by Norwegian Vikings in Ireland and Wales.25 24  Charles Verlinden, “L’origine de sclavus” and L’esclavage 2:999–1010. 25  Pelteret, “Late Anglo-Saxon Slavery,” 51 and 480–84; Bosworth and Toller, An Anglo-Saxon Dictionary, s.v. “wealh.” Wealh may originally have meant “foreigner” before it came to mean “Celt.” Faull, “The Semantic Development,” 38 n. Some scholars have suggested

The Identity Of The Slave In Scandinavia



707

Slave Raiding and Slave Trading

Whether or not the Vikings engaged in large-scale slave trading in their day, they certainly have the image of slave traders in the eyes of modern scholars.26 The extent of this trade is very far from clear. Scandinavians spread through eastern as well as western Europe: to England, Scotland, Ireland, France, the Low Countries, Germany, the Baltic coast, and Russia. The nature of the contacts varied with the different regions; the Vikings, or Varangians as they were called in the east, may have taken captives in all these places, but we do not know whether this amounted to large-scale enslavement, and if so whether they used the slaves in the Scandinavian homeland or as trade goods. Most evidence for slavery in the Viking Age refers to Scandinavians as slave traders but not as slaveholders in their own society. For the activity of Scandinavians in Slavic lands there exists both archeological and documentary evidence: the latter consists mainly of Arabic accounts that refer to the Rus’, whom most scholars have equated with Swedes. No one contests the archeological evidence for some Scandinavian presence, though scholars disagree on what degree of penetration and settlement it represents.27 The Swedes pushed south along the Dniepr to Kiev, the Black Sea, that the word est in later Old Swedish and early Modern Swedish came to mean slave or serf, and that this development is a relic of a trade in Estonian slaves. Lindroth, “Estnisk bosättning,” 196–98, points out the many Swedish place-names beginning with Est- and suggests that these settlements derive their names from Estonians originally brought to Sweden as slaves. The ear liest use of Est for serf or slave apparently comes from the sixteenth century (Svenska Akademien, Ordbok öfver svenska språket, s.v. “Est,” cites examples from 1523 and 15 69). It is not a relic of medieval slavery; rather, it is due to the fact that Sweden took over Estonia in the early modern period at a time when the landlords were German and the peasants (serfs) Estonian. 26  Rosén in Carlsson and Rosén, Svensk historia, 1:76, Foote and Wilson, The Viking Achievement, 200, and Skovgaard-Petersen, “The Coming of Urban Cul ture,” 8–10, exemplify the many modern scholars who refer to this slave trade but are unable to give many specifics. 27   I accept here the equation of the Rus’ with Scandinavians (at least there were Scandinavians among those referred to as Rus’), but this identification is not universally accepted. There is dispute over the meaning of both this term and “Varangian.” The Russian Primary Chronicle, 59–60, tells of the Varangian Rus’ coming to rule at Novgorod in the ninth century. The controversy, known as the Normannist debate, over the origin of the Rus’ became entangled in fierce polemic. For discussions in English of the origin of the name, of the historical worth of the Primary Chronicle, and of the whole controversy, see Riasanovsky, “The Varangian Question”; Riasanovsky, “The Embassy of 838”; V. Thomsen, The Relations between Ancient Russia and Scandinavia, 37–86; Vernadsky, The

708

Karras

and Byzantium, or along the Volga to the lands of the Bulgars and the Khazars, dealing in furs, wax, amber, and slaves.28 Even if one does not accept that the term Rus’ in Arabic sources refers to Swedes, a trail of numismatic evidence leading from the Near East through Russia to Gotland indicates this was an important trade route. The vast numbers of dirhams found in Scandinavia, mostly Gotland, are not proof of trade in any particular commodity, but scholars have often posited a slave trade to explain the hoards.29 Such an explanation implies that Swedes were selling Slavic slaves to the Muslims and bringing home their profits. Ahmad Ibn Faḍlān, who served as ambassador from the Caliph at Baghdad to the Bulgars of the Volga in 921–22, wrote a detailed report on the slave-trading activity of one of the bands of Rus’ trading along the Volga.30 According to Origins of Russia, 174–209; Paszkiewicz, The Origin of Russia, 109–81; Cross and SherbowitzWetzor, introduction to The Russian Primary Chronicle, 38–50; G. Jones, A History of the Vikings, 244–50; Boba, Nomads, Northmen and Slavs, 102–32; Pritsak, The Origin of Rus’. On the archeological side, much of the recent material from excavations in the NovgorodSmolensk-Lake Ladoga area has not yet been fully published, and even less is available in Western European languages. See Stalsberg, “Scandinavian Relations with Northwest Russia.” For a summary of recent work in Russian, see Dejevsky, “The Varangians.” See also Avdusin, “Smolensk and the Varangians,” and comments by Klejn, Lebedev and Nazarenko, and Bulkin; the papers in Rahbek Schmidt, ed., Varangian Problems; and for an older summary of the archeological material, Arbman, Svear i Österviking. On the arrival of the Swedes in Grobin (Latvia), see Nerman, Grobin-Seeburg, 199. 28  Besides Ibn Faḍlān, discussed below, other Arabic sources mention the Rus’ as traders in different commodities: furs, swords, honey, wax, skins, tin, lead, mercury, and slaves. These sources, collected in Birkeland, Arabiske kilder, are by no means all eyewitness accounts—they date from as late as the fourteenth century—nor are they all independent. 29  Sawyer, Kings and Vikings, 123–25, denies that these coins arrived via the slave trade, or indeed any trade. See also Sawyer, “The Viking Perspective.” On the dirham hoards in Russia and the Ukraine, see the series of articles by Thomas Noonan. On the coins found in Sweden itself see Noonan, “When did Islamic Coins First Appear?”; Hovén, “Ninth-Century Dirham Hoards”; Herschend, “Om vad silvermynt kan vara uttryck för”; Rasmusson, “An Introduction to the Viking-Age Hoards.” The dirhams from Scandinavia—86,000 in all, of which 80,000 are from Gotland—have not yet been studied in detail to determine whether the patterns of distribution among various mints are the same as those found in Russia (this would indicate whether the coins were arriving in Scandinavia via Russia or directly from the Islamic lands without circulating on the way). Such a study will have to await the completion of Maimer et al., ed., Corpus Nummorum Saeculorum IX–XI. In the meantime, the best summary is Stenberger, Die Schatzfunde Gotlands. See also Granberg, Förteckning över Kufiska myntfund. 30  The passage for which Ibn Faḍlān has become famous among archeologists and Beowulf scholars is his description of the ship burial of a Rus’ chieftain accompanied by a female slave. See chapter 3.

The Identity Of The Slave In Scandinavia

709

Ibn Faḍlān’s description, the Rus’ were selling only female slaves: a Rus’ merchant would prostrate himself before his wooden idols and ask for help in selling his wares, saying, “Oh, my Lord, I have come from a far country and have with me so many girls and so many sable furs.” The Rus’ did have male slaves of their own, for Ibn Faḍlān discusses the differences in the treatment of the ill depending on whether they were slaves or free men.31 Ibn Faḍlān’s stress on female slaves may be not an accurate reflection of Viking slave trading patterns but rather a product of his voyeurism.32 Ibn Rusta, also writing about the Rus’, reports that “they treat their slaves well” and that when they defeat another tribe “they take their women captive and make them [the men] into slaves.”33 Other sources referring to Slavic slaves in Islamic lands and the importance of the Scandinavians in the acquisition of these slaves do not stress female slaves in particular. In the ninth century an Arabic writer, al-D̲ j̲āḥiẓ, noted that the majority of white eunuchs in Iraq were Slavs.34 If Swedish merchants carried on a large-scale trade in Slavic slaves, they probably brought some home as well, though there is no evidence at all for 31  Ibn Faḍlān, “Reisebericht,” 85–86. My English translation is based on the German translation. Amin Razi’s Encyclopedia, a sixteenth-century work containing an epitome of Ibn Faḍlān’s report that may possibly be a witness to an older text than the existing manuscript, omits any mention of slaves in this passage. “Ibn Fadlan’s Account,” trans. Smyser, 97. 32  Ibn Faḍlān describes in detail the sexual use of the girls the Rus’ bring with them (Ibn Faḍlān, “Reisebericht,” 83); it is not explicit in the Arabic that they are slaves (John Boswell, pers. comm., Dec. 1983), although the fact that they are to be sold certainly indicates this. For another example of Ibn Faḍlān’s interest in sexual matters relating to the Rus’, see Ibn Faḍlān, “Reisebericht,” 93, an account of customs he does not claim to have witnessed, which reveals a good deal about his method and approach. 33  Birkeland, “Arabiske Kilder,” 16–17. Birkeland notes (136n.) that others have suggested that it is the women who are made slaves, but actually the word in this instance has a masculine ending. 34  One late Icelandic legendary saga refers to a slave eunuch belonging to the daughter of the King of Permia (Bósa saga ok Herrauðs 13). This does not mean that Scandinavians had any eunuchs—there is evidence for castration as a punishment for slaves, but no references to it as a normal occurrence—but it indicates that by the thirteenth or fourteenth century they did connect slaves with eunuchs when writing about exotic lands. In Spain the Arabic word ṣiklabī, “Slav,” came to take on the meaning of eunuch. R. Brunschvig, EI, s.v. “‘Abd.” There is no reason to think that male Slavs were not also sold along the Volga to traders from the Muslim East, but the trade route via Germany and France to Umayyad Spain was quite distinct from the route via the Volga and the Caucasus to Iran. Of over 62,000 Muslim coins found in Scandinavia from the Viking period, only thirteen were from Spain. Linder-Welin, “Spanish-Umaiyad Coins found in Scandinavia,” 15.

710

Karras

Slavs as slaves in Sweden in the Viking Age.35 Whether or not they restricted their raiding to women would have a bearing on the types of slaves they imported and the role of those slaves in their society. Women slaves alone would assimilate more quickly than a slave community of both sexes. Certainly it is unlikely that Viking traders brought Slavs into Sweden in large enough numbers to form the basis of the agrarian economy, the substratum of laborers whose surplus product allowed free men the leisure to adventure abroad. If Swedish society during the Viking Age was in fact based on slavery, slave raids cannot have been the only source of slaves. There may have been a native class of hereditary slaves whose origins go back to a time for which there are no sources, but any suggestions that the entire society was based on slavery must be strictly speculative. We have some shaky evidence for Swedish slave-trading activities abroad in the Viking Age, but not for slaveholding at that period. Danish Vikings enslaved many captives from Germany, France, and England. Saints Willibrord, Anskar, and Rimbert, according to their biographers, all ransomed boys from the Danes. Frankish chroniclers refer to slave raids by the Vikings.36 Danish Vikings also took each other, and Norwegians, captive. The skald Valgarð described a raid by Harald Harðráði and Sveinn Úlfsson, future kings of Norway and Denmark respectively, on Sjælland and Fyn, in which many women were enslaved, and according to Snorri Sturlusson the Norwegian king Óláf Tryggvason was captured and enslaved as a young boy by Vikings (presumably Danish) based in Estonia.37 If the account is true presumably not all the slaves sold at that market were captured Norsemen but many were Estonians, and the buyers need not all have been Danes. There is little evidence that the Danes in the Viking Age used slaves to work their lands at home. The slaves seized in raids may have been mainly sold abroad. A much later reference to slave raiding, in the early thirteenth-century laws of Skåne, envisions that people from free families would be captured as slaves, 35  There is evidence of Slavic influence (metalwork and pottery) in the Viking Age level of the settlement at Eketorp on Öland, as well as in a few other excavations, but this can be best explained by trade or a group of Slav settlers. Stenberger, “Eketorp,” 18. 36  Alcuin, Vita Willibrordi 9; Rimbert, Vita Anskarii 15; Vita Rimberti 17. The ransom of captives was a literary topos in Merovingian saints’ lives and may have been taken over into later vitae, so it is well not to rely too much on saints’ lives for evidence of when slavery was common. Pelteret, “Slave Raiding and Slave Trading,” 105. For chronicles, see, for example, Annales Bertiniani, s.a. 842; Annales Fuldenses, s.a. 840 and s.a. 885; Annales Xantenses, s.a. 845; Annales Vedastini, s.a. 879, 880, 881, 884, and 885; d’Haenens, Les invasions normandes, 89–90. Most accounts of Viking raids refer to burning and killing, not capturing of slaves. 37  Snorri Sturluson, Haralds saga Sigurðarsonar 19, and Óláfs saga Tryggvasonar 6–7.

The Identity Of The Slave In Scandinavia

711

presumably within the region governed by the Skåne law: “If a freeborn man comes into slavery, because he is captured and afterward sold, and he is killed, then his relatives and nearest kinsmen have the right to take full compensation for him and give to the master who owned him as much of the compensation as the dead man cost him when he bought him.”38 Whether a survival of earlier nonwritten law, a reflection of late twelfth- or early thirteenth-century practices, or a compiler’s idea of what the rule ought to be, the law implies that free men of Skåne might become slaves of other residents of Skåne who bought them from foreign raiders. Other evidence points to the enslavement, probably by violence, of free people in the twelfth century. The will of Archbishop Absalon of Lund, who died in 1201, freed “Christian the cook who, unjustly captured, was in slavery.”39 Similar unjust enslavement appears in the story of the abbot William of Æbelholt who lived in the second half of the twelfth century: his thirteenth-century biographer reported that his enemies planned to sell him to the Slavs.40 Adam of Bremen observed that Danish pirates did not hesitate to sell each other into slavery.41 In the post-Viking period the Danes and Swedes would at least have had access to slaves from across the Baltic. The chronicle of Henry of Livonia, from the early thirteenth century, reports numerous instances of the Letts, Livs, Lithuanians, and Estonians fighting each other and enslaving the women and children of the victims’ tribes.42 Many of these slaves could well have found their way into the hands of the Danes who were at that time occupying part of Estonia. According to Henry, the German crusaders too took women captives. Captives are not necessarily the same as slaves, but there is no indication that any of these people were being held for ransom, and the men who might have ransomed them had been killed. One of Henry’s fellow priests had been enslaved as a boy and then freed by Bishop Meinhard and educated in a monastery.43 If the Teutonic Knights captured and used Slavic and Baltic slaves it would be surprising if the Danes did not do the same. 38  SkL 129. The Latin paraphrase makes it even clearer that the text refers to the victim of a raid. ASun 75. 39   d d 32, 1:4:63. It may be that “Christianus” was a description and not a proper name and that his enslavement was “unjust” because he was a Christian. 40   Vita Sancti Wilhelmi Abbatis et Confessoris 17, in Vitae Sanctorum Danorum, 328. 41  Adam of Bremen, Gesta 4:6. Krag, “Treller og trellehold,” 215, argues that this is a moral comment on Adam’s part and not a reflection of the actual situation. 42  Henry of Livonia, Chronicon Livoniae 1:5, 11:5, 11:6, and passim. 43  Henry of Livonia, Chronicon Livoniae 13:4 and 10:7. On slaves and captives, see SkyumNielsen, “Saxo som Kilde,” 175.

712

Karras

Swedish expansion, farther to the north, may also have involved enslaving the colonized. In a will of 1310, Asmund Lang freed a slave named or referred to as Karelus; Clara Nevéus suggests that he may have been a Karelian.44 A treaty of 1323 between Magnus Eriksson and the Novgorodian Grand Prince Yuri Danilovich, setting the border between Sweden and Novgorod in what is now Finland, recognized slavery. Slaves, debtors, and criminals who fled from one party to the other were to be returned. This treaty may indicate the presence of slaves among Swedes in Finland and probably the enslavement of Finns.45 Slaves taken in Magnus’s campaigns may have been brought back to Sweden.46 It is in texts referring to Vikings from the West Norse region that the clearest indications appear of slave raiding as a source of slaves for the Scandinavian lands themselves. Once again numerous reports from Britain and Ireland lament the depredations of the Northmen: “Many were the blooming, lively women; and the modest, mild, comely maidens; and the gentle, well brought up youths, and the intelligent, valiant champions, whom they carried off into oppression and bondage over the broad green sea.” The Norwegian slave trader was an important enough figure to appear in the twelfth-century tale of Tristan as the one who lures the young prince from home. Icelandic literature also provides numerous references to raiding in Ireland as a source for slaves.47 Norwegian Vikings made slave raids not only against the Irish and Scots (who are often called Irish in Norse sources) but also against Norse settlers in Ireland or the Scottish Isles or even in Norway itself. Arneið, daughter of 44   d s 1656; Nevéus, Trälarna i landskapslagarnas samhälle, 137. Karelus in this document need not be a proper name. In the original will (Svensk Riksarkivet, Pergamentbrevsamlingen) there are no initial capitals on proper names. It could be a misspelling: the will also refers to a priest by the name of Karulus, and it may be that these are both variants of Karolus. The possibility that Karelus is an ethnic distinction used as a name is tantalizing indeed, but we cannot make too much of it. 45   d s 2418 and Finlands medeltidsurkunder 313. The Swedish translation found in a copy from 1537 includes slaves, but a Swedish version in the same hand as the Latin (preserved in a fifteenth-century copy) has tiænara (servants) instead of träla. 46  They may have included Novgorodians as well as Finns. The former, of course, were Christian, but Orthodox rather than Catholic, and the Swedes might not have considered them coreligionists. On conflict between Sweden and Novgorod, including the religious aspects, see E. Christiansen, The Northern Crusades, 171–91. The distinction between Christian and non-Christian slaves appears in some of the Svear laws: UL Köp 3 and SdmL Köp 3. 47   Cogadh Gaedhel re Gallaibh, trans. Todd, 43; Bromberg, “Wales and the Medieval Slave Trade,” 264; Pelteret, “Slave Raiding and Slave Trading,” 106–10; Gottfried von Strassburg, Tristan und Isold, 11. 2149–2757; Lnb SH6, S84 (H72) and S123(H95).

The Identity Of The Slave In Scandinavia

713

Ásbjǫrn, Jarl in the Hebrides, was captured and enslaved when a Viking fleet attacked her father’s homestead. Freystein the Fair, a slave of Þórstein uxafót (Oxfoot), had been captured as a child in a raid on his parents’ home in Norway. The story of Rögnvald, in a fourteenth-century manuscript, indicates that a child without a guardian might be wrongly enslaved by someone who wanted to seize his inheritance.48 Most of the slaves to whom the Icelandic literature refers, however, are Irish (or Scottish) by ancestry, not just Norse settled in Ireland or Scotland. Modern population geneticists have established that the population of Iceland, which has not undergone any substantial immigration since the tenth century, today contains a substantial Celtic component, some studies even suggesting that the Celtic element was in the majority.49 Even if these data are 48   Lnb S278; Droplaugarsonar saga 1; “þórsteins þáttr uxafóts” 7, in Fjörtíu Íslendinga-þættir, 452; Þttr Rögnvalds, Óláfs saga Tryggvasonar 243, Flateyarbók 1:320. In the stories of Arneið and Freystein, the high birth of the slave is recognized and the slave is freed and marries a free person. This could be indicative of the special treatment accorded to slaves from good Norse families, or the high lineage could be a creation of the storyteller to account for the fact that a former slave married a free person. 49  The Celtic genotypes are expressed in the phenotypic distribution of A and O blood types, which differs greatly from that on the Scandinavian mainland and more closely resembles the distribution found in Ireland and Scotland, although, as Steffensen, “The Physical Anthropology of the Vikings,” 92–95, points out, looking at a high frequency of the O gene as a Celtic trait “would amount to saying that the Icelanders had more Celtic blood than the Irish and the Scots.” Bjarnason et al., “The Blood Groups of Icelanders,” 448, suggest that the blood-group data show the proportion of Norwegian genes as below 25 percent, even taking into account the high O frequency in the Trondelag (Western Norway) from where many of the Icelanders came. More recently and cautiously, ConstandseWestermann, “Genetical Affinity,” 161–63 and 176, concluded from other traits as well as the ABO system that “we might very tentatively postulate the existence of a genetic tie between the populations of Scotland, Ireland and Iceland.” Other measures besides blood types have been used as well. Saugstad, “The Settlement of Iceland,” uses the frequency of the gene for phenylketonuria to suggest that the Norwegian element was around 25 percent; comments by Berry, Edwards, Rafnsson, Steffensen (suggesting a maximum of 40 percent for the Celtic population), and Thompson, and Saugstad’s reply, follow. Potts, “History and Blood Groups,” and Sunderland, “Comment,” discuss the methodology of population history by blood groups. A. C. Berry, “Non-Metrical Variation,” 355–56, uses physical anthropology to determine phenotypic patterns and concludes that there are strong resemblances between the Hebrides and Orkneys and Iceland. Steffensen, “Knoglerne fra Skeljastaðir,” argues that cranial differences show a much larger component from the British Isles in the Icelandic population than is indicated by Landnámabók. Thompson, “Icelandic Admixture Problem,” 69, suggests that, based on gene frequencies, the Icelanders were “of a wholly Irish origin,” very different genetically from Norwegians,

714

Karras

reliable, a high Celtic element in the population does not necessarily imply a high number of Celtic slaves. Many of the early settlers of Iceland came not direct from Norway but from Ireland, the Orkneys, Shetlands, Hebrides and Faroes and Caithness, where they had intermarried with as well as enslaved the native population.50 Most or all Icelanders may have carried enough Celtic genes that even slaves of pure Celtic ancestry would not be physically distinguishable from the free population. Many of the free settlers themselves are described as Irish, not merely “from Ireland,” even when they have Scandinavian names. Others are the sons of Norsemen and Irish or Scottish women.51 These free Irish settlers and freed Irish slaves provide a possible explanation for the high Celtic proportion in the population. The Celtic influx could have come at the time of settlement rather than from a continuing importation of Celtic slaves. Slave trading was a major commercial activity of the Viking Age and later. Vikings not only sold slaves to the eastern Islamic countries but possibly traded captives taken in Britain and Ireland to Muslim Spain as well.52 There were no doubt some slave markets within Scandinavia itself. Irish slaves certainly found their way to the western part of Scandinavia, and since the Danish and Swedish Vikings had traded in slaves it is likely that some of these foreign slaves did find their way to the eastern part of Scandinavia too. It is not possible to conclude from this that Viking Age Scandinavia depended upon a slave economy, but it is possible to see why Scandinavians came to think of all slaves as foreigners or outsiders.

Hereditary and Debt Slavery

By the time the main body of extant sources—literary and legal—was written down, Scandinavia does not seem to have had a continuing supply of foreign Danes, and Swedes. Her mathematical model comes up with a 2 to 7 percent Norse component (p. 79). Obviously this would apply to free people as well as to slaves. While there is plenty of room for disagreement both about the quality of these data and about their interpretation, it seems established that there is a strong Celtic component in the Icelandic population even if the evidence is not conclusive enough to put a percentage on it or to state that the Celtic component outweighed the Norwegian. 50  The source for the origins of the settlers is Landnámabók; its value for computing percentages of Icelanders of Norse and Celtic origin is defended by Steffensen, “Tölfræðilegt mat.” See also J. Jóhannesson, Íslendinga Saga, 15–24. 51  Examples from Landnámabók: S20 (H20); H21; S83 (H71); S366 (H321); S392. 52  Smyth, Scandinavian Kings, 159–64.

The Identity Of The Slave In Scandinavia

715

slaves, except perhaps for some sporadic acquisition of captives. To the extent that slavery survived beyond the Viking Age into the later eleventh, twelfth, thirteenth, and even fourteenth centuries (see chapter 5 for its decline), it must have relied to a great extent on inherited slave status and possibly other forms of internal recruitment; the hereditary slaves need not only have been the descendants of Viking captives but could also have descended from an indigenous enslaved agricultural population. Icelandic literature views slavery, and the moral character associated with it, as hereditary. In the story of Ǫgmund dytt, for example, Ǫgmund’s father is a freedman and his mother a kinswoman of the well-known Icelander VígaGlúm. Not only does Ǫgmund come of good family on his mother’s side, but he is also Glúm’s foster son as well as a kinsman. His father was freed long before Ǫgmund’s birth, so he has never lived in slavery, yet the taint of slavery lingers for Ǫgmund: when Glúm wishes to accuse him of cowardice, he says that Ǫgmund proves the validity of the proverb, “The unfree kindred often is lacking in courage.”53 The phrase indicates that medieval Icelanders classified slaves as a single group, whom they imagined as a kindred, distinguished genealogically from the rest of society despite the fact that the unfree in reality did not come from a particular ethnic group. The law codes of all the Scandinavian countries envision a hereditary slavery. Not all of them state explicitly that the child of two slaves is a slave, but the discussions of the status of the child of a mixed (slavefree) union assume that if both parents are unfree the child will be too. In Sweden, the law of Östergötland makes this explicit: if slaves of two different masters marry, the husband’s owner gets two-thirds of their property and children and the wife’s owner one-third. In the case of a mixed marriage the children are free although the owner of the slave parent might have a right to some of the free parent’s property.54 In discussing the status of children of a mixed marriage all four 53  “Ǫgmundar þáttr dytts,” 108. Glúm’s son Vigfús also refers to Ǫgmund’s ancestry, accusing him of taking after the slave kindred more than after the people of Þverá, Glúm’s kinspeople (107). 54  ÖgL Gipt 29. This whole law uses the terms fostre and fostra and so applies only to homeborn slaves, who could possess some property (see discussion of this group in chapter 3). The Scandinavian laws on the subject may be compared with Roman law, where if the two parents belonged to different masters the child belonged to that of the mother. There was no legal recognition of the parents’ relationship, so the father or his owner could have no claim (CTh 12:1:6). There were, however, exceptions; Constantine laid down rules about the division of property providing that slave families were to be kept together. Buckland, The Roman Law of Slavery, 78; CTh 2:25:1; C 3:38:11; Inst 3:6:10; Nov 22:9:10. The Lombard law also provided that the child of a female slave was a slave but did not state it

716

Karras

Svear laws use the same phrase as the Östergötland law, “let the children go to the better half.”55 None of these Svear law codes explicitly discusses the status of the child of two slaves, except the Uppland law, in a very problematic passage: “If an unfree man takes an unfree woman, betroths her and is wedded to her and begets a child with her, that child is [not] free.”56 Whether the word not appears obviously makes a great deal of difference. All five of the extant fourteenth-century manuscripts do indeed say “that child is free,” but a printed edition from 1607, based on a now-lost manuscript that is thought to be closer to the original than any of the extant manuscripts, has “a ofrælst wæræ” (“is not free”).57 If one assumes that the variant reading is a copyist’s error, the principle of lectio as an explicit principle that the child of a mixed union followed the status of the mother. LLomb 156; similarly EdTh 65. (Cf. lhp 77: 1–2a, where the compiler seems to be confused as to whether a child of a mixed union follows the status of the father or the mother; this confusion continued throughout the history of English serfdom.) The Visigothic law, on the other hand, provided that the child followed the lower of the parents’ statuses—the opposite of the Swedish laws. Verlinden, “L’enfant esclave,” 110; LVis 3:2:3 and 3:3:9. 55  DL Gipt 4; UL Arf 19; VmL Arf 14; SdmL Arf 3:2. It is clear from the Swedish laws that “the better half” refers to the parent of higher status. The canon law has a different rule: Gratian, Decretum, Pars II, Causa XXXII, Q. IV, c. 15, states that the child shall take “the worse part.” This was sometimes interpreted as meaning that the children followed the parent of lower status, the exact opposite of the Swedish law (for example, Rufinus, Summa Decretorum, 487, took this interpretation, as did much German custom, using the phrase “der ärgere Hand”). Where there was no legal marriage, however, as Gratian said in the next sentence, the child followed the status of the mother. Under Roman law, where one partner was a slave there could never be a legal marriage, so in any case of mixed marriage the child automatically followed the mother; see Watson, Roman Slave Law, 10–12. Some decretists interpreted “deteriorem partem” to mean the mother, so that the rule would agree with the Roman law rule. Stephanus of Tournai, Die Summa, 244, explained that the woman was always the inferior part of a marriage, so the sentence meant the children always followed her. See Gilchrist, “Medieval Canon Law,” 293. Ivo of Chartres had the rule that a child follows its mother’s status, apparently whether or not the parents are married: Decretum 16:185. 56  UL Arf 19. 57   Uplandz Laghen, fol. 21r–21v, also edited in Upplandslagen enligt Cod. Holm. B199 och 1607 års utgåva, 101. Holmbäck and Wessén, Svenska landskapslagar 1:2:85n., think that the 1607 edition must be in error. On the dating of the texts in the mss. and printed edition, see Henning, “Upplandslagens redigering,” 146; Ståhle, “Några frågor,” 91–139. Three of the mss. were used in Schlyter’s edition (SGL). The other two have been edited separately: Upplandslagen efter Ängsö handskriften, 47, and Upplandslagen enligt Codex Esplunda, 52. Another medieval fragment of the inheritance section, “Ett fragment af Upplandslagens Ärfdabalk,” is missing the page with this provision. The inheritance section of UL also

The Identity Of The Slave In Scandinavia

717

difficilior would require the acceptance of the reading “that child is free.” Given that the law recognizes slavery and has to specify that in the case of a mixed marriage the child follows the better half, it is hard to imagine a scribe changing the reading to “that child is free” because the opposite made no sense to him. It is more likely that the phrase “that child is free” made no sense to some scribes, who altered it. This seems to have happened in a sixteenth-century copy from one of the extant medieval manuscripts, where the scribe has inserted eij (not) before frælst.58 On the other hand, if the error was accidental it is more likely to have involved omitting a word than inserting one. The difference in the texts may be due not to an alteration, intentional or not, by a copyist on his own but to a deliberate change by the redactors of the law or their successors. If the law when first issued was meant to read “that child is free,” the following statement, that the offspring of a mixed marriage were free, would be meaningless; the law could just have stated that any child of a legal marriage was free. This makes it likely that the reading “that child is free” is a later alteration and that the original reading was that the child of two unfree people inherited their unfreedom.59 The original statement, that the child of two slaves is a slave, is unique in the Scandinavian laws and seems unnecessary. The key may lie in the phrase “is betrothed and wedded to her.” The law may originally have been intended to reassure slaveowners that they would not lose their rights to their slaves’ children if they allowed the slaves to enter a legal, Christian marriage.60 appears in the only extant ms. of HL (Ups. B49), from the fourteenth century, and has the provision that the child shall be free (fol. 77r). 58  Ups. B53, fol. 39v. See Holmbäck and Wessén, Svenska landskapslagar 1:2:85n. 59  The second provision would not be wholly redundant, because it also states that the slave woman who marries a free man with her owner’s permission may go free along with her children. 60  Hasselberg, “Den s.k. Skarastadgan,” 60–65. The clearest reason for believing that the reading “that child is free” is not original lies in the treatment of the law by the other two codes that draw heavily on the Uppland law, those of Västmanland and Södermanland. Had the original reading been “that child is unfree” it is easy to explain why it might have been omitted in the other codes: it could well have been seen as redundant. If, however, the reading were “that child is free,” one could explain why Västmannalagen did not take it up, but it is much more difficult to explain about Södermannalagen. The Västmanland law is in general much harsher on slavery than the Uppland law: for example, it does not copy the provision that a slave shall be compensated for in the same way as a free man, but has low levels of compensation similar to those in the other Scandinavian laws. (UL Manh 6:5; VmL Manh 24:8.) The reason for the harsher laws on slavery despite the Västmanland law’s overall reliance on the Uppland law may not be a difference in the older, unwritten laws of the two provinces, but rather that the Uppland law was an issued

718

Karras

The somewhat earlier Danish laws do not explicitly discuss the status of the children of two slaves; they simply assume that those children are slaves. Where they discuss the children of a mixed marriage they adhere to the Roman law rule that the child follows the status of the mother.61 The Norwegian and Icelandic laws do not make explicit rules about the offspring of mixed marriages, let alone the offspring of two slaves, although both the Frostaþing law and Grágás imply that the child of a free mother and a slave father is free.62 Although the implication of the Icelandic and Norwegian laws, like those of eastern Scandinavia, is that the children of two slaves are slaves themselves, there is some evidence that heredity was not considered as major a source of slaves there as in Denmark and Sweden. Both Norwegian and Swedish laws refer to slave children being reared in the owner’s household.63 Unlike the Swedish laws, which in dividing the children of the slaves of two different masters assume that it is to one’s advantage to have the children, the Norwegian laws treat possession of one’s slaves’ children as an unwelcome obligation.64 This could be because the Norwegian laws are earlier and go back to a period when it was easier to obtain adult slaves than rear slave children. A Frostaþing law provision on the exposure of slave children does not refer to a master exposing the children, but it punishes a slave if he exposes his children of his own accord, implying that it would be normal for him to do so at his master’s code, whose redactors, including churchmen, were probably more interested in creating social changes than the private compiler of the Västmanland law. The Södermanland law is also an issued law, confirmed by Magnus Eriksson in 1327. It takes up some of the improved conditions for slaves from the Uppland law (the prohibition on selling a Christian slave: UL Köp 3; SdmL Köp 3). That it omits the phrase from the Uppland law about the child of a marriage between two slaves would be surprising if the original reading were “that child is free”; the redactors of the law of Uppland in 1296 would hardly have been ready to take a major social step that their counterparts thirty years later were not willing to take. The most likely interpretation is that the reading “that child is free” is later than 1327, after the redaction of the Södermanland laws that rely on the Uppland laws. It is probably a deliberate change, not scribal error. 61  ASun 73; ÆVSjL Tr 14. 62  FrL 2:1; Gr la 118. The Grágás provision says not that such a child is free but that it cannot inherit; since slaves could not inherit anyhow, the provision would not be necessary if the child were not free. Grágás does refer in passing to the child of two slaves being a slave: in referring to penal slavery, it says that a thief shall be a slave as if he had been born of two slaves (Gr Ib 229). 63  GuL 223 (male slaves over 15 years of age, raised at home, may be given in payment of wergeld); ÄVgL Þiuf 17 and YVgL Þiuf 52 (if a slave has been stolen, the owner claims it by saying he has nourished it and brought it up). 64  GuL 57; FrL 2:6.

The Identity Of The Slave In Scandinavia

719

command.65 Many scholars have suggested that exposure of slave children was common in Iceland, particularly in light of the fact that neither law nor saga makes much reference to them.66 Exposure of children was one of the two pagan practices specifically allowed to continue at the time of the adoption of Christianity in Iceland in the year 1000, but saga evidence is not adequate to determine how widespread the practice was.67 There are a number of instances in the sagas of the exposure of children, but the sagas imply disapproval of these incidents. Exposed children in the sagas are sometimes placed where they will be found by kindly people, a common folk-tale motif that one may not take as accurate reporting on Icelandic society. The exposed child is usually illegitimate, the child of an unmarried woman of good family or of the concubine of a free married man. There are no literary examples of a child of two slaves being exposed, though such an event probably would not have been significant to the saga authors. Árni Pálsson, among others, has suggested that the vast majority of slave children in Iceland were exposed at birth, because it was not economical to bring them up.68 The supply of adult working slaves who were economically productive would have been replenished through raids and purchase. The interest in ownership of children in the Swedish laws, which date from a later period, might be due to a decreased supply of slaves from overseas. There is no evidence for large-scale infanticide in Iceland, but it is a possibility. The literature cannot be taken as proof one way or the other. The presence of an indigenous (or at least assimilated) slavery is more marked in the laws of Denmark and Sweden than in those of Norway and Iceland. Emancipation procedures in the Danish and Swedish codes imply that many slaves would be expected to have free relatives, so presumably they were neither recent captives nor a completely distinct class. The laws 65  FrL 2:2. 66  The lack of references to slave children in law codes and sagas does not, of course, mean that they did not exist. Exposure of female children has also been suggested as a possible explanation for the lack of unmarried women in sagas, but in this case as well as with slave children it may merely be that saga authors took no interest in them. Frank, “Marriage in Twelfth- and Thirteenth-Century Iceland,” 475. 67  Ari Þórgilsson, Íslendingabók 7. See Boswell, “Expositio and Oblatio” and The Kindness of Strangers, for the general context of the practice. Foote, “Þrælahald á Íslandi,” 62, points out that when the continuation of the practice of exposing chil dren was sanctioned as one of the conditions for the acceptance of Christianity, it probably referred to free children, because the right of the slaveowner to kill his slaves was not in question (see chapter 4). 68  Pálsson, “Um lok þrældóms.”

720

Karras

of Västergötland allowed a man to buy out of slavery another man’s slave to whom he was related. The laws of Östergötland similarly provided for freeing one’s relatives from slavery, as did those of Skåne, and Andreas Suneson’s paraphrase makes it clear that a free man might have relatives who were born into slavery, not just captured.69 These laws could indicate the enslavement of formerly free men within their own province, since they are still within reach of their kinsmen. One cannot assume that these relatively late laws represent a survival of old laws for the benefit of raid victims. They might rather reflect the results of Christian influence. By the thirteenth century in Sweden, free people and slaves could marry (see chapter 4), and in such a case a slave might well have free relatives. None of the provincial law codes provides for the redemption of a slave by kinsmen from another province or kingdom. The laws of Norway make explicit distinctions between native and foreignborn slaves, implying that both were present. The Frostaþing law provides a penalty of flogging for a native slave who tries to escape, but castration for a foreign-born slave. A foreign-born slave could be tortured if charged with a crime; the law gives no corresponding rule for a native slave. Under the Gulaþing law, a native slave could be beheaded for theft, a foreign slave, male or female, merely flogged. A native female slave or freedwoman who stole would lose an ear for the first offense. The Gulaþing church law also punished a foreign, but not a native, slave for working on Sunday. Native slaves only could be used in payment for the redemption of oðal land.70 The native slaves to whom the various law codes refer could be the descendants of captured foreigners or of formerly free native people who fell into slavery through debt or crime. The latter does not seem a very likely explanation, because the law codes from all four countries make clear distinctions between debtors and other slaves. Still, it is entirely possible that the categories occasionally got blurred, that things were more fluid than the laws make them appear, and that a long period of time or several generations of debt would lead one to be thought of as a permanent slave. Penal slavery is common in 69  ÄVgL Arf 22; ÖgL Arf 17; SkL 131; ASun 77. 70  FrL 10:40; GuL 259; GuL 16; GuL 266. The fact that a penalty for working on Sunday appears only for foreign slaves might indicate that by the time Christianity appears in the laws slave and foreign were linked concepts. In an Icelandic text quoting Gal. 3:28, that there is “neither slave nor free” in Christ, the phrase is “né útlendr þræll né frelsingr,” “neither foreign slave nor free.” (Leifar, 1.) This does not necessarily imply that all slaves were foreign or that in the twelfth or thirteenth century when this was written they were imagined to be so, but in emphasizing the contrasts that come together in Christ, the phrase does indicate the connection of slave and foreigner.

The Identity Of The Slave In Scandinavia

721

continental Germanic as well as Roman law, and it seems to be considered permanent; debt slavery appears much less frequently and is not the same as true slavery.71 Several Swedish laws refer to debt or penal servitude. One Swedish law code, that of Östergötland, prohibits debt slavery, though in another provision it explicitly acknowledges the existence of the practice by providing compensation for the killing of a debt slave. The other laws that mention debt slavery (Upplandslagen and Södermannalagen) do so only to prohibit it.72 The Östergötland law also provides for slavery as a punishment for theft, possibly limited in term according to the amount stolen. The Uppland law provides penal slavery, clearly temporary, in cases where a convicted criminal cannot pay his fine. The law of Gotland, in its provisions on manumission, implies that all slavery was for a fixed term, which may have been for debt. The law of Västergötland implies that thieves might be condemned to penal servitude, serving not the person from whom they stole but the king.73 The Swedish laws on debt slavery do not imply that people were enslaved by force when they could not pay their debts but rather that they would voluntarily become slaves in return for payment of their debts. This “voluntary” debt slavery might have applied to relatively poor people convicted of offenses who could not pay the compensation. It might have its roots in a practice akin to commendation, where someone who did not have any land or could not pay his debts put himself in the service of someone wealthier. There did exist another status, that of flatföring, by which people gave up their property and certain rights in return for support but apparently did not work and did not become unfree.74 The difference between a debt slave and a flatföring, in 71  On penal slavery in Roman law, see Buckland, The Roman Law of Slavery, 277–78. In LVis enslavement is the punishment for several crimes: e.g., LVis 7:2:13, for theft; similarly, LBurg 47; LBav 7:4. See Thompson, The Goths in Spain, 137. 72  ÖgL Arf 11: “and [one may] not give oneself as a voluntary slave, because Birger Jarl abolished that”; ÖgL Dr 17:2; UL Köp 3:3; SdmL Köp 3. 73  ÖgL Uaþ 4, 35 and 38; UL Þing 7:3, echoed in VmL Þing 15; gl 16:2; ÄVgL Þiuf 3:1; YVgL Þiuf 27. 74  Nevéus, Trälarna i landskapslagarnas samhälle, 32–33. The word comes from OS flät, Olc flet, the benches on which guests or followers sat, as opposed to the householder’s high seat. Holmbäck and Wessén, Svenska landskapslagar 5:144n. Provisions for the flatföring are found in ÄVgL lord 3:1; YVgL lord 5. ÖgL has the same sort of status, though it is not called flatföring. It is clear that these people received their keep in return for their property, not for work (ÖgL Arf 12). Similarly, DL Gipt 16; VmL lord 17; UL lord 21; HL lord 15. All these laws provide for indigent people to be taken care of by their relatives first, and since the stress is on the sick or incapacitated these people can hardly have been slaves.

722

Karras

practice, probably came down to the fact that the latter was under the protection of his or her kin, whereas the former might be in less friendly hands. The Danish laws do not mention debt slavery at all. They do provide for the status of flatföring.75 A charter of Valdemar I, from circa 1177, shows that debt slavery was practiced outside the law. The charter, protecting some church liberties in the province of Halland, adds at the end a command to two men to release a woman whom they were apparently holding in slavery because of debts her brothers owed.76 The Skåne and Sjælland laws also provide for penal slavery in cases of theft. The thief became a slave of the king but did not lose all his rights as a free man. He could not be sold to another owner and might keep some rights to his property, but he did lose the right to a free man’s compensation.77 The will of Archbishop Absalon includes clauses freeing several people not born in slavery but rather enslaved. Besides Christianus the cook, mentioned above, there were two women “whom Nicholas the stablemaster took from freedom into slavery,” and a woman who was “taken in” (acceptam). It seems likely that these women were enslaved either for debt or theft, unless they were simply forcibly enslaved without a legal pretext.78 Perhaps those who could not pay their debts commonly resorted to some form of voluntary slavery, and an attempt to make this involuntary explains Valdemar’s reaction to the actions of Bare and Otti, the two men who were keeping a woman falsely enslaved for debt. Several laws from Norway and Iceland imply the existence of debt slavery, although the Frostaþing law does not mention it. Under the Gulaþing law, a debtor who could not pay had to go into slavery, though his kinsmen had the option of paying his debt and taking him as their dependent. A father could put his child into slavery for a debt of up to three marks. He could not by law mortgage more of his child’s future than that amount, though one may question the strict application of such a limit in practice. A debt slave still retained some rights of a free man. He, or his heirs, were to be compensated for his wounding or killing. He was entitled to full compensation, not that of a slave, though the master could take that portion of the compensation that would be due him 75  JL 1:31–32; ESjL 1:38–41; SkL 42–43. All these laws show great concern for offering the heirs the first chance to undertake the support of the dependent, which indicates that the laws did not envision the dependents as penniless. 76  d d 66,1:3:1:100. See Skyum-Nielsen, review of Trälarna i landskapslagarnas samhälle by Nevéus, 358–59. 77  SkL 151; ASun 95; SkL 130; ASun 76; YVSjL 87:8–10. 78   d d 32, 1:4:62.

The Identity Of The Slave In Scandinavia

723

for the death or wounding of a slave. The debtor could not be sold abroad or at a public sale.79 Icelandic law did not so carefully describe the rights of debt slaves, but it did recognize their existence by providing that if they were killed their relatives could receive compensation, paying off the debt.80 The redactor of Grágás did not see the status of debt slave in Iceland as just a sort of commendation, a means of providing protection and maintenance for the indigent in return for services, as the detailed laws separately providing for such support show.81 The Norwegian and Icelandic laws make clear the distinction between the slave and the debtor. Unlike the Swedish laws they use a different term for the debtor, not þræll but rather skulldarmaðr or lögsculldar maðr. The debt slaves’ retention of their compensation and inheritance rights and the prohibition against selling debtors in “heathen lands” where they might not be able to get free once their term was up show their distinction from real slaves. Though the laws clearly differentiate the debtor from the slave, however, practice may have blurred the distinction, with debt slaves being treated as full slaves. The sagas treat as slaves people who were not born into that status and who may have been debtors, like Skíði in Svarfdœla saga, who “was known as [a slave] not because he was such by kin or nature.”82 The distinction in the law, of course, postdates the importing of large numbers of foreign slaves in the Viking Age. It is impossible to say whether the status of the debtor was different from that of the slave in earlier times, if indeed it is possible to speak of debt in a premonetary economy. Under Icelandic law a man could also be enslaved for theft. This servitude, which included confiscation of property, seems to have been permanent.83 Norwegian law provided penal servitude not for theft but for a variety of other offenses. Since these include witchcraft and sexual transgressions, the Church may have had a hand in drafting these laws. Under the Gulaþing law, a woman became the king’s slave if she slept with a slave. She had an option of redeeming herself by paying a fine. The other provisions for penal slavery did not 79  GuL 71. This text is also found in a fragment of about the same time as the Codex Rantzovianus: ngl 1:15–117. 80   Gr la 96; Gr la 118. 81   Gr lb 128–43. 82   Svarfdœla saga 13. It does seem that at times slaves and debt slaves are equated in Icelandic law though different terms are used: e.g., Gr la 44, dealing with law cases about harboring them. See Maurer, “Die Schuldknechtschaft,” for a full discussion. 83  He is to become a slave “as if his father was a slave and his mother a slave-woman and he was born on earth a slave.” Gr lb 229.

724

Karras

provide the option of redemption once the person had been enslaved, and the provision here is probably merely a roundabout way of saying that she would be either fined or enslaved if she could not pay the fine, as was the case for witches under the Eidsivaþing law. Under the Frostaþing law a nun was enslaved for sleeping with a man.84 There are other potential means of enslavement besides foreign raids, trade, debt, and punishment. Unwanted children could be sold into slavery; there could be internal raids or kidnapping; those who held power could simply have denied those at the bottom of the economic pyramid any legal rights and gradually turned them into slaves by law rather than by conquest. Perhaps we should phrase the question in terms not of where the slaves came from but of how people became slaves under the law. The issue then comes down to the power to make and enforce the laws and to grant or deny legal rights. In effect, anyone becomes a slave when someone else can make the law classify him or her as such. From the evidence we have it seems that most of those who were so classified became slaves by capture, birth, or debt, but the evidence is part of the social construction built by the slaveholders who needed to justify the institution. That justification came largely through labeling slaves as inferior by birth, whether for reasons of race or of character; but the labels need not accurately reflect the reality of how people became slaves.

The Image of the Slave

Ethnic Background The means of acquiring slaves can say a good deal about their ethnic background and the reasons society viewed them as other. As it is impossible to say what proportion of the slave population came from raids or from penal servitude, it is also impossible to tell what proportion of slaves came from what ethnic group. The stereotypical slave was small and dark and therefore of a different ethnic background from the tall, ruddy Norse; there is evidence, however, that the reality did not bear out this stereotype. As discussed above, even if slaves in Iceland had a high proportion of Celtic genes, free people did too.

84  GuL 198; EþL 1:45; FrL 3:14. The latter provision is repeated in the so-called King Sverre’s Christian Law, which actually dates from the late thirteenth century, so the punishment may have remained in force until that date, although the later law more likely just copied obsolete provisions from the earlier ones. This law cannot antedate the establishment of female monasticism in Norway in the twelfth century.

The Identity Of The Slave In Scandinavia

725

The stereotype of the slave as ethnically different was part of the social construction that made the slave an other. The Danish and Swedish laws do not make any explicit distinctions between native and foreign slaves and do not in any way imply ethnic differences. There is not as much literary material from the East Norse as from the West Norse region to supplement the laws and give a fuller idea of racial attitudes about slaves, but one source provides some help. Saxo Grammaticus, writing at the beginning of the thirteenth century, hints in several instances that a slave is physically distinguishable from a free person, not on ethnic grounds but because social class is reflected physically. In one story he recounts, two Swedish princes are forced by their evil stepmother to become shepherds, royal slaves. Svanhvita, the daughter of the Danish king, comes upon them and when one of the two, Regner, identifies himself as a slave, she replies: The shimmering glow of your eyes pronounces you the progeny of kings, not of slaves. Your form reveals your race, just as in your glittering look nature’s beauty shines out. Your sharp sight displays the splendour of your birth and there is no indication of humble origin when the handsomeness which graces you is a manifest token of your nobility. The outward keenness of your glances betrays a bright quality within, while your visage testifies your true family, for in your gleaming countenance may be observed the magnificence of your ancestors; no unworthy begetter could have shed on you such a gracious, aristocratic appearance. The glory of your blood bathes your brow with a kindred glory, and the mirror of your face reflects your innate rank.85 Even if Saxo means Svanhvita to speak words of love and not of social commentary, it tells us something of his attitude to slaves that he has her connect beauty with nobility of birth and ugliness with slave ancestry rather than with character traits. A case of the reverse effect, the identification of slave ancestry in a person of high social status, occurs when Amleth, prince of Denmark, visits the court of the king of England. Amleth, whose feigned madness contains a good deal of method, notes among other criticisms of the dinner just served that the king “had the eyes of a slave” and the queen “displayed three mannerisms of a 85  Saxo, Gesta Danorum 2:2:5, trans. Fisher, 44. Cf. the Icelandic Hrólfs saga kraka 23, in which Helgi falls in love with Yrsa, who has been brought up by peasants but is (unknown to either of them) his daughter. She tells him she is a peasant’s daughter and he responds that she does not have the eyes of a slave.

726

Karras

maidservant.” The king subsequently discovers that his mother had committed adultery with a slave, and that his wife’s mother “had been captured and reduced to slavery.”86 These passages manifest Saxo’s extreme class pride rather than reflecting any real physical distinguishability of slaves. Throughout the Gesta Danorum Saxo extols those of noble blood at the expense of the base-born. He expresses great admiration for the hero Starkather, who, though dying, “preferred the torture of his agonizing wounds before the ministrations of those in low walks of life.” Starkather rejects aid from a bailiff (praeco), from a man who had married a female slave (ancilla), and from a female slave herself. He tells them he considers it “utterly degrading” to accept their aid. Saxo comments: “What a great man we must now judge Starkather!”87 Given Saxo’s view of the importance of social status, one cannot infer anything from his tales of identifying slaves from their looks. The English queen, daughter of a woman captured and enslaved, is revealed as a slave by her manner—her blood may have been tainted by her mother’s enslavement, but this is reflected in her behavior, not in her features. Saxo nowhere explicitly refers to slaves as foreigners. It is the social fact of enslavement that led Saxo to attribute what he saw as degrading characteristics to slaves. While Saxo depicted the slave in the eastern part of Scandinavia as a degraded and ignoble person but not necessarily of a different nationality than the Danes, the image of the slave in Iceland and Norway was that of a foreigner. Most references in Icelandic literature to the nationalities of slaves call them Irish. Sometimes the sagas include Irish slaves because they add an exotic element. In Laxdœla saga, the beautiful woman whom Hǫskuld Dala-Kolsson purchases for three times the normal price turns out to be Melkorka, the daughter of King Mýrkjartan (Muircheartach) of Ireland. The romantic motif of the kidnapped princess and her son who returns to his grandfather’s country with tokens of his ancestry served a literary function that may have led the author to include it regardless of its historical accuracy. Icelanders of Celtic and slave descent might have wanted to envision themselves as of royal origin.88 Several Irish slaves play major roles in other Icelandic sagas. In Njáls saga, Gunnar’s wife Hallgerð uses the Irish slave Melkólf, who already has a bad reputation, to steal food from his former master’s household for her. In the story of 86  Saxo, Gesta Danorum 2:6:18–20, trans. Fisher, 88–89. 87  Saxo, Gesta Danorum 6:7:11–13, trans. Fisher, 182. 88   Laxdœla saga 12–13 and 21; more briefly, in Lnb, S105. Other examples of royal slaves are Erp and Muirgeal, son and wife of Jarl Meldun of Scotland: Lnb S96 (H83). Royal women of Norse origin were also enslaved in Ireland and Scotland: Lnb S84 (H72); Lnb S278.

The Identity Of The Slave In Scandinavia

727

the dreams of Þórstein, son of Síðu-Hall, an Irish slave named Gilli kills his master who has castrated him.89 The sagas depict these slaves as generally nasty sorts; their Irishness may be part of this negative image. Even when they do not play a particular role in the story and are only mentioned in passing, slaves or freedmen are often given an Irish origin or Irish names. Landnámabók lists many Irish freedmen, most of whom came to Iceland with Norse settlers from the Isles. Landnámabók reports that their masters freed them and gave them land and that the places they settled were named after them. The compilers of Landnámabók probably included these accounts to explain place-names.90 That slaves in Iceland were perceived as being mainly Irish does not necessarily mean that they were ethnically different from the dominant class. The presence and assimilation of free people of Celtic ancestry would have made the slaves of Irish ancestry not very distinct from the free population. This is probably less true for Norway than for Iceland, since fewer free people had ancestors who had spent time in Celtic lands and intermarried with non-Norse, yet both Norwegian codes contain provisions implying that slaves would not as a rule be immediately distinguishable from free people. The Frostaþing law considers the case of a corpse of which no one knows the status, and the Gulaþing law sets free a slave who has lived as a free man for twenty years without arousing complaint.91 These laws, of course, do not imply that there were no foreign slaves, but they do envision that slaves would not automatically be distinguishable from free people by their physical characteristics. Where specific provisions about foreign-born slaves do appear in the Norwegian laws, the penalties are not always harsher for foreign slaves; the differences may derive from practical reasons rather than ethnic prejudice. The castration penalty for a foreign-born slave in the Frostaþing law could be an attempt to control and quell, not merely punish, the unruly. Foreign-born slaves might be more likely than native ones to flee, because not having grown up as slaves they might be less used to the state of slavery or to heavy work or they might be given less attractive tasks. On the other hand, while native-born slaves would have no home to escape to, they would have a greater chance of passing as free men in a different part of the country than would foreigners. 89   Brennu-Njáls saga 47–49; “Draumr Þórsteins Siðu-Hallssonar,” 321–26; Foote, “Þrælahald á Íslandi,” 59. The latter incident is noteworthy not only because the slave is Irish but also because this is the only literary reference to castration of slaves in Iceland, and because this incident, which supposedly took place around 1055, is the latest event including a slave described in the family sagas. 90  Examples from Landnámabók: S95–103; S135; S350 (H309, M14); S390 (H345). 91  FrL 4:5–6; GuL 61.

728

Karras

In the Gulaþing law, the more severe penalty for theft by a native than by a foreign-born slave could stem from the consideration that a master was not responsible for the training of a foreign-born slave and thus should not have to suffer by losing such a valuable piece of property.92 Though the distinction in the laws between native and foreign slaves was not an ethnic one, and though at least in Iceland a large proportion of both the slave and the free population had some Celtic genes, the stereotype persisted that the slave was distinguishable physically, not just by manner, from the free person. The stereotype is at issue in two legends, the stories of Geirmund Heljarskin and of the god Ríg, both of which survive in Icelandic versions. The story of Geirmund Heljarskin (Hell-skin; the goddess Hel was supposed to be black-skinned), which appears in both Landnámabók and Sturlunga saga, is set in the time of Harald Fairhair, king of Norway in the late ninth and early tenth century. Ljófvina, the daughter of the king of Bjarmaland (Permia), has been taken captive by King Hjǫr, who marries her. She is clearly not a slave or concubine but his queen. The queen gives birth to twin sons, but they are so dark and ugly that she is ashamed. At the same time one of her slaves bears a child, Leif, who is fair and even noble in appearance. The queen forces the slave to trade children with her and brings up the princely-looking Leif as her own, not telling her husband what she has done. One day, however, Leif reveals his physical cowardice: he is playing with a gold ring in the king’s hall; the other two boys take the ring from him and he cries as they mock him. The skald Bragi, present in the hall, tells the queen that he can tell the lads’ true nature by their behavior, that Leif is the son of Loðhǫtt the slave and not of Hjǫr and his queen. The queen is forced to admit what she has done.93 The moral of the story lies in the fallibility of the slave stereotype. Slavishness in the tale is certainly determined by heredity but is expressed only in ignobility of character, not in physical appearance. The queen thinks that noble appearance as well as character should mark noble birth, but the tale proves her wrong. The story does not describe her own appearance, but it is strange that as a Permian she would have exactly the same idea of noble beauty as the Scandinavian stereotype. It is never made explicit that the dark skin of the queen’s two sons makes them look foreign and therefore slavish, but it is clear 92  Gjessing, “Trældom i Norge,” 215, suggests that the penalty was lighter for the foreign-born slave because he would not be familiar with local custom; this explanation is not convincing, as theft is considered wrong in most societies. 93   Lnb, S113 (H86); a fuller version in “Geirmundar þáttr Heljarskinns” 2, Sturlunga saga 1:5–7. The story also appears in Hálfs saga ok Hálfsrekka 17:1–3, with a different mother for the boys.

The Identity Of The Slave In Scandinavia

729

that they do not measure up to the ideal of princely beauty. Yet their noble birth shines through. According to the story, Geirmund, one of the two darkskinned sons, grew up to be a great Viking and one of the most prominent settlers of Iceland. At least by the thirteenth century, when the extant versions of the story were written down, and probably much earlier, stories circulated in Iceland making the point that despite the stereotype one cannot tell a slave by his looks. The point would no doubt have appealed to many Icelanders of Celtic descent, but the story still indicates that the stereotype existed. Rígspula: A Legend of Origins Making nearly the opposite point—that looks, rather than innate character, reveal social class—is the poem Rígsþula (the Lay of Rig). Rígsþula, though classed as an Eddic poem, does not appear in the Codex Regius, the manuscript containing what is usually called the Elder Edda. It appears only in Ormsbók, a manuscript of Snorri’s Prose Edda from the mid-fourteenth century, in fragmentary form.94 In the poem, a god called Ríg visits three households in order of increasing prosperity, sleeping with the wife and fathering a son in each household. The third son, Jarl (earl), marries and fathers the nobility and royalty; the second son, Karl (man, farmer, yeoman), marries and fathers the race of free men. Þræl (slave), the oldest son, has children with a woman named Þír (female slave), without the benefit of a wedding. Rígsþula’s description of Þræl and his family is frequently quoted; it reveals more about the Nordic view of the slave than about the actual identity of slaves. Karl is ruddy and Jarl fair; Þræl’s main identifying characteristic, his dark skin or hair (hǫrundsvartan or hǫrvi svartan depending on the emendation) marks his caste. Þræl did grow   and he throve well, there was wrinkled hide   upon his hands, crippled knuckles,   ….. thick fingers,   a foul face, a bent back   and long heels. He soon did more   to try his strength, doing his tasks,   tying up bast; he carried home brush   the bitter day long. To his farmstead   a wanderer came; her arms were sunburnt   and her soles dirty; her nose was crooked,   she was called Þír. 94  Sveinsson, Íslenzkar bókmenntir 1:287.

730

Karras

The names of the couple’s children describe their appearance (Tǫtrughypla, clothed in rags) or their work (Fiósnir, byre-man), or they are simply derogatory epithets (Fúlnir, stinking). The sons “built fences, // herded the swine / and spread manure, // tended the goats / and dug up turf.”95 Rígspula may reveal how Icelanders envisioned the slave’s life and duties and the contempt in which the ruling classes held him, but the real interest of this poem lies in its tripartite class division and the physical description of the slave. The god’s paternity of the three sons could mean that the poet saw slavery as divinely ordained with the slave class as genetically inferior, or it could mean that the poet wished to point out the essential brotherhood—or at least half-brotherhood—of all classes of human beings.96 Þræl’s dark skin is often taken to mean that the poet intended him as a foreigner, perhaps a Celt, but it is probably just meant to convey his unattractiveness or his manual labor under the sun. Þræl probably stands for “a fixed, distinctive class of remote native ancestry,” his physical characteristics the attribution by the elite of all ugliness to the underclass.97 The main obstacle to an understanding of what Rígsþula reveals about social class is the uncertainty of its date and place of origin. Some scholars have taken Rígsþula as an expression of a primitive Germanic social structure that had remained basically unchanged since Tacitus.98 The social structure depicted in the poem is not exactly tripartite, for it includes not only the three social classes but also a king, Jarl’s youngest son Kon the Young (Konr ungr, from which, the poet implies, the word konungr, king, derives).99 The poem does not reflect Icelandic social structure at any period: Iceland never had a king, nor did it have a hereditary aristocracy, although it did come to be dominated by a few great clans. Scholars have attempted to date Rígsþula based on its language and its parallels in other sources. It has been dated anywhere from the ninth through thirteenth centuries.100 Einar Ólafur Sveinsson dates it to the tenth century, suggesting that it was composed by a skald who had visited Ireland and who borrowed the Jarl/Karl distinction from the English tradition (eorl/ceorl). He 95   Rígsþula, Edda, 281–82, trans. Jeffrey Mazo (unpublished). 96  Hill, “Rígsþula,” 79–81, plumps for the former. 97  Foote and Wilson, The Viking Achievement, 66; Hastrup, Culture and History, 108. 98  Gjessing, “Trældom i Norge,” 43. Dumézil, Gods of the Ancient Northmen, 118–25, pushes it even farther back, relating it to general Indo-European social structures. 99   Edda, 286. This is a false etymology. 100  For an excellent summary see Harris, “Eddic Poetry,” 94–97. See also Anne Holtsmark, KLNM, s.v. “Rígsþula.”

The Identity Of The Slave In Scandinavia

731

discusses its archaic language, the parallels to the Eddic Hávamál in the education of Kon, the poem’s relation to the Hyndluljóð and Vǫluspá (which rank among the earliest known Icelandic poetry), and its irregular fornyrðislag meter. He also points out that Snorri Sturluson and the author of Skjǫldunga saga, which probably dates from about 1200, both knew the poem, because the Skjǫldunga saga and Snorri’s Ynglinga saga use the names Ríg and Danp.101 However, the archaisms in language could be deliberate. The literary parallels do not necessarily mean that Rígsþula influenced the other works. The influence could have gone the other way, or two contemporary works could have drawn on the same oral traditions.102 By social rather than linguistic criteria, Rígsþula is probably thirteenthcentury. The poem does not reflect Icelandic social structure at all, nor does it reflect the social structure in Norway much before the thirteenth century. A poem is more likely to depict a social class that is no longer significant, as 101  Sveinsson, Íslenzkar bókmenntir, 1:287–91; Snorri Sturluson, Ynglinga saga 17; Skjoldungasaga. See Bjarni Guðnason, KLNM, s.v. “Skjoldunga Saga,” on the dating. 102  Most recently Dronke, “Sem jarlar forðum,” has argued for an early date on the grounds of saga parallels to the description of Jarl. Even if one accepts the parallels, however, that does not mean the entire poem as it now survives must predate those sagas, nor need the date be pushed back to the Viking Age, only to the twelfth century. See comment by Von See, Edda, Saga, Skaldendichtung, 516. Nerman, “Rígsþula 16:8” and “Rígsþulas ålder,” attempts to prove an early date by archeological support for material objects described in the poem. Such dating, however, cannot be exact, for no one can rule out the possibility that the descriptions are old but the poem as it stands is not or that material conditions later correspond with it as well. Turville-Petre, review of Íslenzkar bókmenntir by Sveinsson, agrees with Sveinsson that at least the tradition behind Rígsþula predates Vǫluspá, composed just before 1000, that Rígsþula probably influenced Skjǫldunga saga rather than the reverse, and that nothing in Rígsþula proves that it is later than the ninth century. Heusler, Das altgermanische Dichtung, 97, claims that since neither Snorri nor Saxo used the myth of origins presented in Rígsþula, they must not have known it, and its composition therefore postdates 1230. There is no reason why Snorri and Saxo need necessarily have used it if they had known it, but it is certainly the sort of thing they would both have seized upon. Von See, “Das Alter der Rígsþula,” 88–92, argues that Snorri knew of a Ríg only in the context of royal genealogy and did not have a source relating him to mythology; he also argues that Rígsþula contains echoes of Vǫlsunga saga and is therefore late. Heusler also argues, in “Heimat und Alter,” 270–81, that many words in the poem, some of them loanwords, are not attested elsewhere until the twelfth century and that other elements in the poem, such as the pagan custom of sprinkling a child with water in naming it, are deliberate archaisms. Holm-Olsen, “Middelalderens litteratur i Norge,” 282, also dates the poem to the thirteenth century, basically following Heusler’s arguments, and Krag, “Treller og trellehold,” 212, concludes that “the poem has no value as a source for the period when slaveholding was a social reality.”

732

Karras

was the case with slaves in the thirteenth century, than one that is not yet significant, as with nobility in the ninth. Norway did have kings, and in fact some scholars have connected the poem with the development of a royal ideology in the thirteenth century.103 But a noble class, as distinct from free men as free men were from slaves, probably was not an important factor in Norwegian society during the Viking Age. The class of jarls was not socially distinct from that of the most prosperous bönder—yeoman farmers—in the tenth and eleventh centuries.104 The activities of Jarl and his sons as described in the poem do not include Viking raids or adventures as one might expect if the poem dated from the ninth or tenth century.105 The idea of all men as brothers, sons of one divine father, is a Christian conception, though it may be a pagan one as well. It appears in a twelfth-century Icelandic homily, so a twelfth- or thirteenthcentury Icelander could certainly have gotten the idea from a Christian context. There are European parallels as well, in stories of the different races of men descended from the sons of Noah.106 Even if we could firmly date the poem in its known form to the thirteenth century, this would not settle the question of what if any Scandinavian society it reflects. It could be a late version of a very old myth, or it could be a purely thirteenth-century composition. Some argue that it is not Scandinavian at all in its conception, but rather Irish. While the proposed Irish origin of some elements in the poem, like the god sleeping with the wife in each household he visits, may be somewhat tenuous, the amount of personal physical description

103  Von See, “Das Alter der Rígsþula,” 95, and “Rígsþula,” 96, connects the poem with a particular king of Norway, Håkon Håkonsson, who received the title of king in his father’s lifetime in 1240 and who was called the “young king,” like konr ungr in the poem. 104  A. Bøe, KLNM, s.v. “Jarl, “ notes that the only evidence for the position of the jarl in the Viking Age, besides Rígsþula, comes from sources like Heimskringla written in the thirteenth century. The jarl is mentioned in the Norwegian law codes as one of several classes with high wergelds, but the position does not seem there to be hereditary. GuL 200. See von See, “Das Alter der Rígsþula,” 93, and Gurevich, Norvezhskoe obshchestvo, English summary, 314, and Svobodnoe krest’ianstvo, English summary, pp. 260–73. Gurevich and others may go too far in arguing for a free peasant society in the Viking Age—almost a “primitive democracy”—but, though kings and magnates certainly existed, a hereditary aristocracy was not clearly delineated as it later became. See also chapter 5 on social classes in medieval Norway. 105  Von See, “Das Alter der Rígsþula,” 93. 106   Homiliu-bók, 33: “Our Father who art in Heaven. We come from a great family. Under this Father, slave and lord, king and knight, rich man and poor man are brothers.” For European analogues, see Hill, “Rígsþula.”

The Identity Of The Slave In Scandinavia

733

in Rígsþula is quite unusual for Norse poetry.107 The name Ríg is clearly derived from Irish rí, king, and no god or king by that name is known in any Scandinavian source except Rígsþula, Ynglinga saga, and Skjǫldunga saga.108 The presence of Irish elements in Rígsþula could mean that the legend is basically Irish in origin, but the lack of any very close parallels from Irish literature means that such an attribution must be purely speculative. There are no parallel tales in Germanic literature from which one could argue a primitive Germanic origin. The Irish elements could indicate an origin within a Norse community in Ireland or Scotland. The poet need not just have retold an Irish tale but could have incorporated Irish elements into a story from Norse oral tradition or of his own composition. The lay could have been composed in Norway or Iceland any time after the late tenth century and still have drawn on Irish elements, for Irish culture had by then had the opportunity to influence Norse culture through Scandinavian exposure to Irish courts.109 The use in Rígsþula of thirteenth-century social divisions and of Irish words and motifs casts doubt on the roots of the poem in an early Scandinavian oral tradition. So many features must be late or imported that it is impossible to say that the core is traditional. Rígsþula cannot be used to demonstrate a primitive Germanic or Viking conception of slavery. Nor can it reflect Icelandic or Norwegian social reality at the time the poem was written in its present form, because slavery no longer existed in Iceland and Norway by the thirteenth century. Rígsþula reflects not the way Norwegians or Icelanders viewed slavery at any particular point in history, but rather a thirteenth-century view of the origins of social classes that includes a class no longer significant. The poem asserts the brotherhood of men of all classes, but also stresses clear hereditary lines of demarcation between classes. The view of the slave as ugly, dirty, and menial does not reflect the attitudes of actual slaveholders to actual slaves but rather presents a figure of low status with whom Karl, the free farmer, contrasts. Free 107  Young, “Does Rígsþula Betray Irish Influence?” 101, points out that the custom of a visitor sleeping with the wife of the household is not attested anywhere else in Norse literature. See Chadwick, “Pictish and Celtic Marriage,” 84, 94–96, on Celtic traditions of a child fathered by a god, 107–09 on the possibility of this as a royal ritual, and 111–15 on relations to Rígsþula. There are no Celtic tales, however, that are very close analogues, and in general the custom seems to be treated as a dishonor to the husband. Robin Chapman Stacey, pers. comm., Oct. 1983. 108  Young, “Does Rígsþula Betray Irish Influence?” 99. The Irish adds a “g” in some declined forms. 109  Young, “Does Rígsþula Betray Irish Influence?” 106–07. See Hill, “Rígsþula,” 87–88, for the suggestion that the poem is an Irish (or insular) gloss on the Noah story.

734

Karras

peasants, threatened by the increasing power of large landholders in the thirteenth century, would prefer to be contrasted with those who did manual labor, and thus depicted as free, prosperous, and attractive. Karl is ruddyskinned, with lively eyes; his wife, Snœr (daughter-in-law), has keys dangling from her belt; their children have names like Halr (man), Drengr (young man, warrior), Þegn (young man, warrior, vassal), Smiðr (smith), Snót (woman, lady), and Brúðr (bride).110 The slave in the poem serves as a contrast to the honorable free farmer. The important point of the legend in Rígsþula for its audience is that the nobleman is innately noble and the free man innately free. The slave’s innate slavishness is only a foil for the other two. The description of the slave emphasizes by contrast the free man’s role in bearing arms and managing the farm rather than doing the dirty work. The physical characteristics of the slave contrast with the healthy good looks of the ruddy free man. It is possible that they also reflect the physical appearance of the typical slave in the Viking Age, but too many other possible reasons for the description, as well as the probable chronological distance from the age of slavery and foreign influences on the tale, seriously detract from the poem’s worth as evidence of a foreign ethnic origin for slaves as a class. Stereotypes and Slavishness In Iceland and probably Norway, which shared a West Norse literary culture, slaves were perceived as physically distinct from free men, though the tale of Geirmund Heljarskin shows that there was some recognition that looks were not everything. The stereotype could have arisen not from any actual difference but from a wish on the part of free people to distinguish themselves from slaves.111 Slavery was connected with blood even though the outward signs of it were not innate characteristics but behavior or typical tasks.112 Physical distinctiveness need not mean foreign origin. Dark coloring was not always seen as a sign of slavishness, but it was usually perceived as 110   Rígsþula, Edda, 282. 111  Similarly, in nineteenth-century Russia serfholders concocted a wholly fictitious theory of different racial origins for their serfs and themselves. Kolchin, “In Defense of Servitude.” 112  Saxo’s queen of England, mentioned above, behaves in a slavish manner because her mother was once captured: enslavement taints even noble blood. In Friðþjófs saga ins frœkna 6:9 (a legendary saga, not about real Icelanders), a man remarks during a storm at sea that he’d rather be at home bringing women breakfast in bed; Friðþjóf responds, “It shows you are of slave kindred, when you wish to work preparing food.” The type of labor one does is here, in this late saga, connected with servile blood or family.

The Identity Of The Slave In Scandinavia

735

unattractive. In Kormáks saga, Kormák has dark hair and people think him “black and ugly,” and in Fóstbrœðra saga a woman with black hair is not considered beautiful, but neither is she called foreign or slavish.113 The slave’s dark coloring in the literary sources may represent ugliness and lack of moral worth more than foreign origin.114 The stereotype of the slave presented in Icelandic literature had other elements besides the ugliness and degrading labor that appear in Rígsþula. The cowardice Leif displays in the story of Geirmund Heljarskin turns up elsewhere also. Proverbs that appear in Icelandic literature show how the degraded image of the slave pervaded the culture even into the thirteenth century. The wisdom poetry of the Edda gives the advice that slaves are unreliable: “a self-willed slave” is included in a list of things that one should never trust.115 Besides the proverb “the unfree kindred is often lacking in courage,” another states, “it is bad to have a slave as a special friend.”116 In the sagas people quote these sayings as ancient wisdom borne out by the situation at hand. That comments like these were current in Icelandic literature indicates that the idea of slavery carried with it all sorts of unfavorable connotations besides forced labor and lack of freedom. Even if slave in proverbs is used metaphorically to mean “no-good person in general,” rather than in a specific juridical sense, it is still significant that its use was so current. The “lack of freedom” connotation of slavery was certainly present and was responsible for the use of slavery as a political metaphor, for example by Saxo.117 Even after slavery had ended the image of the slave could carry powerful messages. Cowardice was the most important element in the unfavorable image of the slave’s moral character. The slave could not participate in the warfare that was glorified as the highest calling of any freeborn man in the Viking Age. While 113   Kormáks saga 3; Fostbrœðra saga 11. Kormák is of course an Irish name, but the Kormák of the saga is not an Irishman; because of intermarriage with Irish people, Irish names—for example, Njáll—became common in Iceland in general. Tomasson, “The Continuity of Icelandic Names,” 282. 114  On the relation of looks and character in Icelandic literature in general, see L. Lönnroth, “Kroppen som själens spegel,” esp. 24–31. 115   Hávamál, Edda, 30. 116  “Ǫgmundar þáttr dytts,” 108; Grettis saga 82. Grettir’s slave has just betrayed him to his enemy Þorbjǫrn ǫngul, and Þorbjǫrn and his men abuse the slave for his faithlessness. The latter proverb is also quoted in Brennu-Njáls saga 49, in reference to someone who is not a slave, though a thoroughly unpleasant sort. Saxo, Gesta Danorum 5:3:5, also has the proverb. 117  Skyum-Nielsen, “Saxo som Kilde,” 178–79. As Skyum-Nielsen points out, even though it is only used as a metaphor, it is not accidentally chosen but a very powerful one.

736

Karras

Óðin accepted warriors into Valhalla, slaves were claimed by Þór.118 The most vivid image of slave cowardice comes not from a tale of Iceland but from one of the legendary sagas, the Icelandic version of the Nibelung story. After the brothers Gunnar and Hǫgni have been defeated and captured, Gunnar refuses to tell Atli the whereabouts of the Niflung treasure until Atli presents him with the heart of Hǫgni. Instead of Hǫgni, Atli’s men kill the slave Hjalli, though Hǫgni says that he would rather die himself than listen to the slave’s shrieks. When they bring the heart to Gunnar, he knows it is Hjalli’s heart and not Hǫgni’s, because it is still trembling.119 In saga-age Iceland men displayed their bravery in private clashes of arms rather than full-scale battles between warring monarchs. No doubt the accusation that slaves were reluctant to fight has some truth to it, though prudence rather than cowardice could have accounted for a slave’s unwillingness to risk his neck when he personally had nothing to gain. With the prospect of being killed in following his master’s orders (as happens to Þórð the Coward in Gísla saga when he is mistaken for Gísli who has exchanged clothes with him),120 with little hope of being avenged after his death, and with no hope of fame for his deeds—the reason often given for a lack of fear among the Norsemen—there is little wonder the slave did not want to get involved in his master’s feuds. The image of the slave included not only inborn ugliness but also dirtiness. Slaves such as the tattered and infested Louse-Odd or the hunchbacked and stooping Kol are given the most demeaning appearances possible to go with their demeaned status. When the poet Kormák insults someone with the kenning “sooty slouch of seamy tatters, who spreads dung on homefields,” he is probably calling him a slave.121 The slave was also depicted as incompetent and uncoordinated.122 118   Hárbarzlióð, Edda, 82; this statement occurs nowhere else. The poem is a verbal duel between Óðin and Þór, and Óðin may be saying this just to provoke Þór. However, it was generally thought that slaves did not go to Valhalla: Gautreks saga 1 states that Óðin would receive a slave if his master took him along. 119   Vǫlsunga saga 39. Finally Hǫgni is killed and his heart is brought to Gunnar; Gunnar says that he can now die happy, since he can be certain that the secret of the treasure will die with him. 120   Gísla saga Súrssonar 20. 121   Fóstbrœðra saga 23; Vilmundar saga viðutan 2 (this saga is later and falls into the category of legend, but it does indicate what Icelanders of the fourteenth century thought slaves could be like); Kormáks saga 4. 122  When asked to join in a game because his master’s party is short a man, the slave Svart grumbles because he has work to do; when he joins in the game, his shoes keep coming

The Identity Of The Slave In Scandinavia

737

The stereotyped image of the slave as cowardly, ugly, and incompetent is not universal in the Icelandic sagas. Freystein the Fair in “Þorsteins þáttr uxafóts,” for example, is handsome and faithful, perhaps because he was born a free Norwegian. In Fóstbrœðra saga the slave Kolbak is “large and strong and fair in personal appearance,” and he does not hesitate to commit an attack for which he will be outlawed. The slave Asgaut in Laxdœla saga comes in for high praise: “He was a large and doughty man, and although he was called a slave, few might compare themselves to him, although they were free; and he knew well how to serve his master.” Two slaves of Víga-Glúm are killed when they shield him with their bodies in a battle.123 The exceptions to the general unfavorable image of the slave in Icelandic literature show that society was willing to give its approval to stories of slaves who behaved in ways their masters considered meritorious. Perhaps some slaves were depicted in this way because many Icelanders of the thirteenth century were descended from freed slaves and wanted to put their ancestors in a positive light.124 While individual slaves might be admired, however, slave as a general term did not conjure up images of faithful service. The existence of graves where slaves were buried with their masters, presumably so that they could serve them in the next world (see chapter 3), would seem to indicate that faithful service was part of the slave’s image. Yet despite these known graves, the literature reveals a contempt for slaves in stories of men who would rather die than be associated with slaves, who when about to be executed make a last request not to be led to death by slaves, or who cannot rest in their graves because they are humiliated at having slaves buried with them.125 The purpose of these recorded incidents seems to have been to show undone and the whole party laughs at him. This whole scene, in Havarðor saga Ísfirðings 17, has no function in the tale except comic relief. 123   Fóstbrœðra saga 9; Laxdœla saga 11 (Asgaut is later given his freedom for helping a kinsman of his master’s wife escape from his pursuers, and he goes to Denmark to settle); Víga-Glúms saga 23. The slaves in this last incident are not even named in the list of people killed in the battle. The incident is not used in the saga to show bravery on the part of the slaves; Glúm’s rival Már uses it to ridicule Glúm. 124  Foote, “Þrælahald á Íslandi,” 70–72. 125  Saxo, Gesta Danorum 6:7:11–13; Jómsvikinga saga 23; Lnb S72 (H60). In the Landnámabók tale, a man called Ásmund is buried with a slave beside him. In one of the redactions it is not stated how the slave died, but in another the slave has killed himself out of grief at his master’s death. When people walk past Ásmund’s mound they hear him singing a ditty about how he would rather be alone than in such bad company. Eventually the mound is opened and the slave’s body removed so that Ásmund can lie peacefully. Cf. the provisions in the laws about slaves being buried in a different part of the churchyard: EþL 50;

738

Karras

the pride of the free men involved rather than the low status of the slaves. That the literature viewed such extreme pride as an admirable attribute may indicate a longing on the part of the writers for the old days when there was a group on whom one could look down in that way. The names given to slaves in the sagas and Landnámabók only rarely betray a similar contempt for slaves. Some of the names are Irish; others, like Atli, Bjǫrn, and Þórð, are names like those of free people; most slaves mentioned in the sagas, especially women, are not referred to by name at all. The names given to slaves in Rígsþula cannot be taken as anything but symbolic and mythical; they may be contemptuous but real slaves were not known by such names. Some names that were particularly common among slaves in the literature, like Svart (black) or Kol (coal), refer to the stereotypical slave appearance. These names, however, were not found only among slaves: Iceland was full of free Kolbeins, Kolskeggs, and Kols, and Svart was the name of a man of good family, the great-grandfather of Sæmund the Wise.126 One of the slaves of Hjǫrleif, who took part in the revolt against him, was called Drafdrit (draf, dregs, husks, food refuse, giving English draff, and drit, dirt, excrement, giving English dirt). This was no doubt a derogatory nickname but not necessarily one given only to a slave: drit- appears as an element in several personal names of free men.127 Slaves in the sagas might be more likely than free people to acquire or put up with derogatory nicknames, but they could not be distinguished as a class by their names or nicknames, except perhaps that Irish names were more common among them than in the general population.128 In Sweden the slaves whose names are known from wills have normal Swedish names.129 Even after slavery no longer exists, thrall or ambáttarson (son of a slave woman) often appears in lists of insults for which compensation was required,

BþL 1:9. Archeology confirms the laws’ prescription: Gejwall, Westerhus, 77, reports that at this twelfth- and thirteenth- century cemetery, in Jämtland (on the Norwegian-Swedish border), the women were buried to the north of the church and the men to the south, as prescribed in the Norwegian laws, and also that the people buried closer to the church tended to be taller than those buried further away, suggesting that the latter may have been less well nourished, and therefore that they may have been slaves. 126   Lnb H296. 127  Gull-Þóris saga 1; Cleasby-Vigfusson, s.v. “drit.” 128   Svarfdœla saga 15 says that Skidi “bar þræls nam” which Williams, Thraldom in Ancient Iceland, 36, translates as “had this name because he was a thrall,” but this should rather be translated “was known as a slave” or “was a slave.” See Cleasby-Vigfusson, s.v. “bera.” 129  Nevéus, Trälarna i landskapslagarnas samhälle, 181.

The Identity Of The Slave In Scandinavia

739

showing the strength behind the contempt of slaves.130 In the Gulaþing law such an insult required full atonement.131 The only comparable insults were to say a man had given birth to a child or had been used as a woman, or to call him by the name of any female animal or use any derogatory term generally applied only to women.132 Fairly often in the sagas people insult others by calling them slaves, though when it is servants who are called slaves no one seems to get particularly upset about the insult.133 To call someone a slave was included in lists of punishable insults in most of the Danish and Swedish laws, even those of the fourteenth and fifteenth centuries. Some of these provisions imply not just a word spoken in insult but an attempt to have the person legally regarded as a slave.134 Even during and after the decline of slavery the term slave seems to have retained much of its opprobrium, although the laws cannot necessarily be taken as truly indicative of the worst slanders; surely the Danes of the fourteenth and fifteenth centuries could have thought of worse insults than “slave” or “chaff-back” (one who steals seed). At the same time Icelandic literature was developing its negative image of the slave, the word slave was appearing in a much more positive context, that of Christianity.135 The Biblical examples cited above show that the same terms were used for slaves in the Bible as for slaves in Scandinavia. That Moses and the children of Israel were slaves in Egypt may have taken some of the stigma away from slavery. The King’s Mirror, a didactic work of the thirteenth century, goes further than the Old Swedish pentateuch in calling Joseph in Egypt 130  See Meulengracht Sørensen, Unmanly Man, 25, for a comparison with other defamatory terms. 131  GuL 196. It is not included as a punishable insult in the Frostaþing law: FrL 10:35. Grágás does not specifically list punishable insults: see Gr lb 237. 132   Svarfdœla saga 13 shows the comparability of the insults of effeminacy and servility. Klaufi uses the term þræll of someone who has called him ragr (“coward,” with connotations of “effeminate, catamite”). 133  In Brennu-Njáls saga 17, Glúm calls his wife’s foster-father, Þjóstólf, who is a member of his household, a slave. Þjóstólf kills him, but it is clear from the story that he had in mind to kill him even before the insult. Calling servants slaves as an insult is reported even after slavery was no more in Iceland: e.g., Íslendinga saga 28, Sturlunga saga 1:257. 134  YVgL Retl 7; YVgL Add 2:13; ÖgL Arf 24–25; SdmL Manh 34; VmL Köp 3; Lund municipal code (later than 1326) 21–22, dgkl 4:7; Helsingborg (before 1346) 22, dgkl 4:153; Halmstad (after 1322) 20, dgkl 4:258; Skånske Biärkerät (provincial privilege as reissued in the fifteenth century) 23, dgkl 4:344. The inclusion of skalk in this last law shows that another social class has begun to join or replace slaves as the lowest in status. See also MESL (early fourteenth century), Radzstuffw balker 31. 135  On slavery as a metaphor in the New Testament, see Vogt, “The Faithful Slave,” 141–45.

740

Karras

and Esther and Mordechai in Ahasuerus’s kingdom alien slaves, stressing how galling it was to Haman to see his honors given to an alien slave.136 Even the king, it stressed, should realize that he is a slave before God, and in prayer one should refer to oneself as “thy slave begotten in sin by thy servant, the son of thy female slave.”137 Other sources refer to nuns as God’s ambáttar (female slaves). A book of homilies in Old Norse constantly refers to all men as slaves of God: “Thou, Lord, hast heard the prayer of Thy slave.”138 The force of such language in prayer was that it used the lowest possible human status to stress humanity’s humility before God. It not only brought out the lowliness of the slave but also fostered the idea that humility was a good thing. It recognized the equality of all people before God; the text “there is neither Jew nor Greek, slave nor free, male or female, but you are all one in Christ Jesus” (Gal. 3:28) was preached in Iceland.139 The equality of all people before God may not have been taken very seriously, however, when the Norwegian laws required the burial of slaves in a separate part of the churchyard. The contempt for slaves shown in Icelandic literature cannot necessarily be extrapolated to all Scandinavian society at the time when slaves were held. When slaves play a significant role in an Icelandic saga they are usually, though not always, depicted in a negative way. They are not reflections of slavery in actual life but literary characters created by an author for the purpose of his story, but it is significant that the author picks slaves for his negative characters. Generally, however, the slaves’ role in literature is not significant; they perform work, are sent on errands, accompany their masters on trips, but they are not named and do not really exist as characters. The fact that people in sagas constantly express contempt of slaves or use “slave” as a figurative term of opprobrium does not mean that such expressions were common in the tenth century but rather that thirteenth-century writers, composing perhaps for an aristocratic audience, imagined them as having been so. The thirteenth-century Icelanders wrote or heard about some slaves who were objects of contempt and about others who behaved honorably and were freed, perhaps founding a family that lasted until their own day. The range of 136   Speculum Regale 42. 137   Speculum Regale 54. Ancilla here could mean “maid.” Except for the prayer, the text of the King’s Mirror is in the vernacular. 138   Saga Guðmundar Arasonar 22; Codex AM 619 Quarto 14:4 and passim. This is a Biblical citation ( 1 Kings 8:28–29), but this does not make it less significant that the word þræll is used. The King James version has servant, the Vulgate servi tui. 139  It survives in a fragment of a moral tract found in a twelfth- or thirteenth-century collection of religious works: Leifar, 1.

The Identity Of The Slave In Scandinavia

741

images that survives probably reflects fairly well the range of ways slaves were viewed and perhaps the ways they were treated. The majority of slaves probably fell in the middle and did not make such a good story. Slaves were probably not as different from poor or powerless free people in appearance, in daily life, in courage, or in status as thirteenth-century glorifiers of the independent Icelandic free peasant would have liked to believe.

Slavery and Cultural Antipathy David Wyatt In the preceding chapters, I have highlighted how slavery was an ancient institution of some cultural significance for the societies of medieval Britain. Furthermore, during the late eleventh and early twelfth centuries the sources register a remarkable upsurge in reports of slave raiding activities within the ‘Celtic’ regions of the British Isles.1 The timing of this upsurge might, initially, appear surprising given that there was a growing antipathy towards such activities within English society following the Norman Conquest. Modern historians have attributed the high incidence of ‘Celtic’ slave raiding within this period to the corrupting influence of Scandinavian settlers within the Irish Sea region.2 In order to support this contention they have pointed to the increasing number of collaborative enterprises undertaken by mixed native and Scandinavian war bands in the ‘Celtic’ regions.3 Yet, this explanation fails to acknowledge the longstanding cultural significance of the institution of slavery within these native communities. Indeed, slave-holding and slave-taking were of fundamental importance for conceptions of power, ethnic/community identity and norms of gender within these warrior-centred societies. When this is taken into consideration then the notion that native war bands, comprising of naive and easily corrupted Christian individuals, had been exploited by pagan Viking sensibilities appears less convincing. The predominant reason for the alliance of ‘Celtic’ and Norse warrior fraternities were that both groups had similar social objectives and motivations; both occupied a similar conceptual world that related expressions of masculinity with physical and social power.

Source: Wyatt, David, “Slavery and Cultural Antipathy,” in David Wyatt, Slaves and Warriors in Medieval Britain and Ireland: 800–1200, Leiden: Brill, 2009, 337–394. 1  Holm, “Slave Trade”, pp. 317–345, 341, Wyatt, “Gruffudd”, pp. 595–617, see also Pelteret, Slavery, pp. 78–79. 2  Ó Croínín, Medieval Ireland, pp. 268–269, Holm, “Slave Trade”, p. 330, Bromberg, “Wales Slave Trade”, pp. 263–272, 263. A similar contention has been made regarding the Scandinavian influence upon Anglo-Saxon society, see Fisher, Anglo-Saxon Age, p. 333. For other examples see Wyatt, “Significance”, p. 329. 3  C. Etchingham, Viking Raids on Irish Church Settlements in the Ninth Century (Maynooth: St. Patrick’s College, Department of Old and Middle Irish, 1996), p. 49, Holm, “Slave Trade”, p. 328, Bromberg, “Wales Slave Trade”, pp. 263–264, 269, Wyatt, “Gruffudd”, pp. 595–617.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_025

Slavery And Cultural Antipathy

743

It was certainly true that Scandinavian raiders and traders were heavily involved in the medieval slave trade. Through their maritime connections they appear to have established an efficient slave supply system that fed lucrative market places across North Western Europe.4 The Scandinavians thereby established a maritime network that allowed warriors from a number of ethnic groups to ‘shunt’ slaves on to new locations. The Irish Sea region appears to have been particularly suitable for such an operation. The fragmented nature of the ‘Celtic’ polities together with the extensive coastlines bordering the Irish Sea facilitated the slave raiding activities of both ‘Celtic’ and Norse warrior fraternities. These warrior groups captured slaves as a means to assert their power over rival neighbouring territories. They were then able to isolate and alienate their victims from the vicinity of their origin communities by transporting them to the Scandinavian slave market places. By removing these individuals from their social milieu and shipping them abroad the slave raiders significantly reduced their motivations for escape and thus facilitated the process of subjection which, in turn, may have increased their value.5 The Scandinavian markets provided a venue at which human beings could be exchanged for silver, exotic goods or for more compliant foreign slaves. Native warriors, therefore, supplied slaves for the elaborate Scandinavian trade network, acting in a not dissimilar manner to the West African middlemen of the New World slave trade.6 Some modern historians have argued that the continuing significance of slave raiding and slave trading in the ‘Celtic’ regions during this period provides evidence that these societies were languishing in the economic backwaters of Europe.7 It may be true that the economies of this region were less 4  Wyatt, “Gruffudd”, pp. 601–602. 5  The late eleventh-century text Lebor Na Cert reveals the prestigious value attributed to foreign female slaves within Irish society, see LnC pp. 41, 69, 99. The early medieval Irish law codes also note that a higher value was placed upon foreign slaves, see ALI, vol. v, p. 111, see also Chapter 2. The Welsh Law of Hywel Dda reveals that the same was true in Welsh society, LHD p. 167. Similarly, Scottish slave raiders in the twelfth century are said to have sold their English captives on to other “barbarians” (presumably Norseman or other Scotsmen from the North) in exchange for cattle (“… vel pro vaccis aliis barbaris vendiderunt.”), Richard of Hexham, CRSHR, vol. iii, p. 157. 6  Patterson, Social Death, p. 154. See also R. Blackburn, The Making of New World Slavery: from Baroque to the Modern 1492–1800 (London/New York: Verso, 1997), p. 327. 7  For example, Gillingham has remarked that “in the less developed economies of the Celtic lands slavery continued to flourish.” See “Conquering”, pp. 72–73. See also Bartlett, Gerald of Wales 1146–1223 (Oxford: Clarendon Press, 1982), p. 162 for further examples see Wyatt, “Significance”, pp. 332–335.

744

Wyatt

developed than, say, the economies of England or France during this period. Yet, it is extremely dangerous to use the existence of slavery as a yardstick by which to measure the level of this economic stagnation. By doing this modern historians reveal more about their own post-abolitionist mindsets than they do about the twelfth-century economies of the Irish Sea regions. We have seen that arguments that regard slavery to be an economically degenerative institution have been discredited within the recent discourse concerning plantation slavery in the early-modern period.8 Similarly, attitudes that view slavery to be an unprofitable institution, which was incompatible with the growth of the medieval economy, find little support in the historical record.9 Within many medieval European societies slave ownership was an expression of an individual’s power, prowess and virility. Although slave labour may not have been a significant factor in medieval modes of production, slaves were important symbols of prestige and consequently the slave trade was extremely lucrative. The profitable nature of this trade was clearly recognised by contemporary medieval commentators. William of Malmesbury remarked that following the Conquest of England the reform-minded Archbishop Lanfranc and Bishop Wulfstan of Worcester, had to persuade William I to suppress the sale of Christian souls from England into Ireland. It seems that the Conqueror had initially been reluctant to prohibit this trade because he had “… enjoyed a share of the profits from this traffic.”10 Lanfranc’s and Wulfstan’s concerns regarding the bustling slave trade with Ireland were probably well founded. Dublin appears to have been the beating heart of the slave supply network during the tenth, eleventh and twelfth centuries.11 Archaeological excavations in the city have revealed the extraordinary wealth generated there during this period.12 The Irish adoption of coinage (even if for limited purposes) in the native kingdoms near to Dublin, reveals the significant economic impact of the Hiberno-Norse settlements 8  See chapter 1 and Wyatt, “Significance”, p. 335. 9  Ibid., see also McCormick, “New Light”, pp. 17–54. 10  “… qui regem pro commodo uenalitatis quod sibi pensitabitur …” GRA, pp. 497–498, William of Malmesbury also noted that Earl Godwine’s wife had made a vast fortune from trafficking attractive young female slaves from England to Denmark, ibid., p. 363. 11  Holm, “Slave Trade”, H. Clark and B. Ambrosiani, Towns in the Viking-Age (Leicester/ London: Leicester University Press, 1991), p. 168, A. Smyth, Scandinavian York and Dublin (Dublin: Templekieran Press, 1975), vol. ii, pp. 240–242, R. Power, “Magnus Bareleg’s Expedition to the West”, SHR, 65, 1986, pp. 107–132, 123, P. F. Wallace, “Archaeology and the Emergence of Dublin” in J. Bradley (ed.), Settlement and Society in Medieval Ireland (Kilkenny: Boethius, 1988), pp. 123–160, 158, Wyatt, “Gruffudd”, p. 602. 12  Clark and Ambrosiani, Towns, pp. 104–106, Wallace, “Archaeology”, pp. 123–160.

Slavery And Cultural Antipathy

745

upon the native economy.13 What is more, the increasing wealth of Dublin attracted the attention of the native Irish rulers. During the eleventh century over-lordship of the city became a prestigious and coveted title within Irish society.14 Any provincial king who wished to claim high kingship had to assert his control over this significant port and slave market.15 In 1052 Diarmait mac Maíl na mBó, king of Leinster, went one step further and assumed direct control over the city. Following this Diarmait and the subsequent Irish rulers of Dublin became increasingly involved in Irish Sea maritime operations that included extensive slave raiding activities upon the Scottish and Welsh coastlines. The lucrative nature of control over the Irish-Sea slave trade is revealed by the expedition of Magnus Barelegs, king of Norway (d. 1102) in 1098.16 In this year Magnus sailed west from Norway with a large fleet and his intention appears to have been to establish control over maritime trade routes in the Irish Sea region. He used his forces to ravage and then secure control over the Western Isles, Man and parts of Galloway. Furthermore, he made an alliance with the ruler of Dublin and may well have exercised political hegemony over the city for a brief period. Magnus appears to have deliberately attempted to halt the post-Conquest English advance into North Wales. In 1098 he attacked and successfully repelled the combined forces of Hugh of Chester and Hugh of Shrewsbury who were trying to secure control over the island of Anglesey. Magnus appears to have been aware that, if consolidated, their territorial gains in North Wales might pose a direct threat to prospective slave raiding activities there.17 Moreover, following the native Irish assumption of control over the city of Dublin we see an intensification of slave raiding activities within Ireland during the final half of the eleventh century.18 The contemporary evidence for such insular Irish slave raiding would appear to undermine the argument that the medieval Irish slave trade may be explained away by the external exploitation of a small community of Hiberno-Norse raiders and traders. Nevertheless, Scandinavian influence does appear to have acted as a catalyst stimulating the slave economies of the societies of medieval Britain. The 13  M. Gerriets, “Money among the Irish: coin hoards in Viking-Age Ireland”, JRSAI, 115 (1985) pp. 121–139 at p. 125. 14  Wallace, “Archaeology”, p. 159. 15  Ibid., p. 159. 16  Power, “Magnus”, pp. 107–132 17  Wyatt, “Gruffudd”, pp. 609–614. 18  The Irish annalists reveal an apparent upsurge in insular slave raiding activities conducted by local rulers during the late eleventh and early twelfth centuries, see Holm, “Slave Trade” for examples, pp. 335–337.

746

Wyatt

absence of a centralised money economy within the ‘Celtic’ regions does not appear to have negated the increasing utilisation of recycled coinage within these communities neither did it prevent their cultivation of wide-ranging trading contacts.19 The increasingly lucrative nature of the slave market within these regions is indicated by the direct involvement of the ‘Celtic’ royal elites in raiding and trading for slaves. Irish, Welsh and Scottish kings appear to have indulged in such large scale raiding activities during the late eleventh and early twelfth centuries that they became noteworthy for the contemporary chroniclers.20 This intensification of slave raiding activity in the Irish Sea region coincided significantly with the more general western European trends towards economic growth during this period.21 The contention that slavery was an economically degenerative institution practised only by economically ‘backward’ societies must, therefore, be called into question.22 Indeed, the complex and dynamic nature of the slave trade in the Irish Sea region during the late eleventh and early twelfth centuries would suggest that the institution was far from incompatible with the rise of the market economy.23 If we wish to understand the reasons why the burgeoning and vibrant Irish Sea slave trade came to an end then we must therefore follow an alternative line of enquiry.

19  See note 13 above. In addition, Crawford has argued that a standardised currency in the form of silver ‘ring money’ might have already existed in parts of the Western Isles and Scotland during the final decades of the tenth century, see Scandinavian Scotland (Leicester: Leicester University Press, 1987), pp. 133–134. See also J. Graham-Campbell, “Viking-Age Silver Hoards of the Isle of Man” in C. Fell et al. (eds.), The Viking Age in the Isle of Man (London: Viking Society for Northern Research, 1983), pp. 53–80, 63 and Barrow, Kingship and Unity, p. 20. 20  Wyatt, “Gruffudd”, pp. 595–617, Holm, “Slave Trade”, pp. 336–341. 21  Barber has argued that the growth of trade and trading networks was a clear indication of the vibrant nature of the economies within Western Europe during the eleventh and twelfth centuries. M. Barber, The Two Cities, Medieval Europe 1050–1320 (London: Routledge, 1993), pp. 60–85, 61. Nevertheless, modern sensibilities have prevented historians from recognising that the trade in human beings might have been part of this economic ‘revolution’. 22  For example, Gillingham’s characterization of the Hiberno-Norse towns of Ireland as “just a few coastal emporia” is unsubstantiated and misleading, see J. Gillingham, “Foundations of a Disunited Kingdom”, A. Grant and K. J. Stringer (eds.), Uniting the Kingdom? The Making of British History (London/New York: Routledge, 1995), pp. 48–64, 50. Indeed, the archaeological evidence from these settlements attests both to their size and economically vibrant character, see Clark and Ambrosiani, Towns, pp. 104–106. 23  A contention which is supported by McCormick’s investigation of the slave trade in the Carolingian empire, see “New Light”, pp. 17–54, see also chapter 1.

Slavery And Cultural Antipathy

747

We have already seen that cultural, rather than economic, factors had instigated a shift in attitudes towards the institution of slavery in post-Conquest England. Conversely, there may also have been significant cultural motivations behind the apparent increase of slave raiding activities conducted by the warrior fraternities of the ‘Celtic’ regions. We must acknowledge these cultural motivations if we are to understand the dramatic antipathy that had arisen between the English elite and their ‘Celtic’ neighbours during the twelfth century. This chapter will, therefore, examine the increasingly condemnatory attitudes of English writers towards the ‘Celtic’ populations in this period, particularly in relation to slave raiding activities. It will analyse the various methods by which the post-Conquest English elite attempted to remould the behavioural norms of their ‘Celtic’ counterparts. The cultural tensions that this ‘imperial’ acculturation created within the ‘Celtic’ communities will also be highlighted. This, in turn, will be related to an examination of insular responses towards the increasing English cultural influence being exerted upon ‘Celtic’ society. I will suggest that the increase in slave raiding activities during the twelfth century constituted a significant expression of this cultural antipathy. Finally, we will examine how the English elite were attempting to modify the behavioural norms of their ‘Celtic’ counterparts. Indeed, it will be argued that English ‘imperial’ domination brought about the end of the medieval British slave supply network, although it did not destroy the institution of slavery in the region.

The Dynamics of Cultural Antipathy

We have already seen how, during the first half of the twelfth century, postConquest English writers had begun to cultivate a new and ‘superior’ self-image for their countrymen.24 This ‘civilised’ self-image was based predominantly upon revised martial codes/values and social norms of the reform movement. English writers, therefore, regarded their own society as one that had recently emerged from the barbarous milieu of the Anglo-Saxon past. This was a past that was thought to have been characterised by savagely violent behaviour, sexual laxity and residual paganism.25 Yet, the ‘enlightened’ self-image of these twelfth century authors was not entirely forged in opposition to their AngloSaxon past. It was also sharply defined against the other communities of the 24  See also Thomas, English and Normans, pp. 307–315. 25  The tendency to compare contemporary ‘barbarians’ with the ancestors of a now ‘civilised’ society was also evident in Roman antiquity, see J. F. Killeen, “Ireland in the Greek and Roman writers”, PRIA, 1976, 76c, pp. 207–215.

748

Wyatt

British Isles that continued to adhere to more traditional modes of behaviour and systems of patriarchy.26 During this period, then, the English elite increasingly came to regard their ‘Celtic’ neighbours with an attitude of condescending disdain. This is illustrated by the negative portrayals of the social and political mores of these ‘Celtic’ communities found within the contemporary English narratives.27 The English sense of superiority was based upon the assumption that those communities that did not adhere to the revised behavioural norms which had been adopted within post-Conquest English society were to be regarded as barbarous and uncivilised. Significantly, the ideals and values of the reform movement provided the ideological tools necessary for the construction of this ‘imperialistic’ attitude. As Bartlett has observed, Not sharing the social patterns of western Europe meant not being a part of the Church. The images of exclusion and otherness available to those who formed and expressed opinions in twelfth century western Europe included not only the dichotomy Christian/non-Christian, but also that of civilized/barbarian, and the two polarities were often mutually enforcing … As the men of Frankish Europe intruded upon societies around and unlike their own, they found both non-Christians … and local variants of Christianity (notably in the Celtic countries). Their response was to equate the two …28 As we will see, the norms and objectives of the reform movement moulded and motivated the damning criticisms that the English commentators directed against their ‘Celtic’ counterparts. Furthermore, these criticisms were

26  See R. Bartlett, The Making of Europe: Conquest, Colonization and Cultural Change, 950–1350 (London: Penguin, 1994), p. 23, J. Gillingham, “The English invasion of Ireland” in B. Bradshaw et al. (eds.), Representing Ireland: Literature and the Origins of Conflict (Cambridge/New York: Cambridge University Press, 1993), pp. 24–42 and Thomas, English and Normans, pp. 307–315. 27  These negative twelfth-century portrayals may be juxtaposed against pre-Conquest Anglo-Saxon attitudes that appear to have regarded the ‘Celtic’ populations as cultural equals, ibid, and Gillingham “Conquering”, pp. 68–69. See also D. Bethell, “English monks and Irish reform in the eleventh and twelfth centuries”, Historical Studies, 8, 1971, pp. 117– 126 and R. Marsden, “Race and Imperialism: Twelfth Century English Attitudes towards Scotland” in S. Kehoe & I. MacPhail (eds.) A Panorama of Scottish History: Contemporary Considerations (Glasgow: Glasgow University, School of History 2004), pp. 10–23. 28  Bartlett, The Making of Europe, p. 23.

Slavery And Cultural Antipathy

749

extremely similar to the condemnations that the reforming papacy had hurled against Anglo-Saxon society prior to the Norman Conquest.29 Let us now, briefly, examine some examples of this superior English attitude in more detail. One of the most consistent condemnations directed by English writers against the ‘Celtic’ peoples was that they had failed to adhere to the reformers’ codes concerning acceptable sexual behaviour. The primary motivation for this criticism was the ‘Celtic’ elite’s continuing expression of power through displays of virility and female accumulation.30 This ‘Celtic’ recalcitrance in the face of the ascetic ideal first became an issue of concern for English ecclesiastical leaders in the years following the Norman Conquest.31 In 1073–4 Archbishop Lanfranc wrote to Toirrdelbach Ua Briain, king of Munster and Guthric, king of Dublin complaining about the incestuous and debauched marital practices of their subjects.32 Lanfranc had been prompted into this action by a letter from Pope Gregory VII that had instructed the archbishop … to extirpate serious moral offences wherever they occur, specifically and pre-eminently that you strive by every means open to you to ban the wicked practice we have heard rumoured of the Irish: namely that many of them not only desert their lawful wives but even sell them.33 Similarly, John of Salisbury wrote to the Pope Adrian IV during the central years of the twelfth century complaining that the Welsh despised the law of matrimony and remarking that

29  B. Smith, “The Frontiers of Church Reform in the British Isles, 1170–1230” in D. Abulafia and N. Berend (ed.), Medieval Frontiers: Concepts and Practices (Aldershot/ Burlington USA: Ashgate, 2002), pp. 239–254, 244. See also chapter 4. 30  See chapter 2. 31  It appears that Anglo-Saxon reformers, such as Archbishop Wulfstan of York, had been somewhat less critical of their ‘Celtic’ neighbours. Indeed, the pre-Conquest reformers would seem to have been far too busy attempting to modify the behaviour of their own countrymen, see Wormald “Wulfstan”, pp. 191–224. 32   LoL, 8 and 10, pp. 67–73. 33  “…  quatinus grauiora usquequaque resecare uitia stvdeat, et inter omnia et prae omnibus nefas quod de Scottis audiuimus, videlicet qvod pleriqve proprias uxores non solum deserunt sed etiam uendunt, modis omnibus prohibere contendat.” Ibid., 8, pp. 66–67.

750

Wyatt

… they (the Welsh) barter for a price their concubines, whom they have as well as wives; and ignoring the guilt of incest, they do not blush to uncover the nakedness of those who are kin by ties of blood.34 This ecclesiastical reaction towards the ‘Celtic’ resistance to the reform norms was to manifest itself in a particularly derogatory attitude towards the ‘Celtic’ elite during the twelfth century This ‘superior other’ attitude stereotyped the ‘Celtic’ populations as being lascivious, morally corrupt and, in the some extreme cases, less than human.35 This is, perhaps, best exemplified by Gerald of Wales’ views concerning the Irish. Gerald regarded the Irish in the following manner This is a filthy people, wallowing in vice. Of all peoples it is the least instructed in the rudiments of Faith … a people that is adulterous, incestuous, unlawfully conceived and born, outside the law, and shamefully abusing nature herself in spiteful and horrible practices.36

34  “Legem uero matrimonii contempnentes concubinas, quas cum uxoribus habent, commutant pretio, et crimen incestus ignorantes consanguinearum turpitudinem reuelare non erubescunt.” The Letters of John of Salisbury (1153–1161), W. J. Millor et al. (eds.) (Oxford: Clarendon Press, 1986), vol. i, 87, p. 135. In this letter John was relaying the concerns of Meurig, Bishop of Bangor (1139–1161). Meurig’s attitude, therefore, reveals that native Welsh ecclesiastics had adopted the values of reform and were attempting to impose these values upon the secular Welsh population with the support of the English and continental Church. 35  Bartlett has revealed how similar arguments were employed by reform-minded ecclesiasts to denigrate the non-Christian peoples of Eastern Europe during the same period, see Gerald, pp. 158–177 and Making of Europe, pp. 18–23. The use of sexual mores as a means to differentiate and dehumanise has been used by individuals involved in many diverse imperial enterprises in many different eras and geographical locations, for examples see Killeen, “Ireland”, p. 208. 36  “Gens enim hæc gens spurcissima, gens vitiis involutissima, gens omnium gentium in fidei rudimentis incultissima … Nec mirandum si de gente adultera, gente incesta, gente illegitime nata et copulata, gente exlege, arte invida et invisa ipsam turpiter adulterante naturam, tales interdum contra naturæ legem natura producat.” TH/EH, book iii, caps, xix and xxxv, pp. 164, 181, translation from O’Meara, HTI, pp. 108, 116. The Topographia was written during the final quarter of the twelfth century as a work of praise to Henry II. It attempts to provide a justification for his invasion of Ireland in 1171. Gerald also described Welsh sexual mores in a similar, if slightly more tempered manner, see JDW, book ii, cap. vi, pp. 262–263. Richard of Hexham also made similar complaints about the Scots, see below.

Slavery And Cultural Antipathy

751

Gerald’s criticisms of the Irish population frequently revolved around their sexual behaviour. He felt that the Irish were so given over to sexual excesses that many of them indulged in bestiality.37 This portrayal, like many of Gerald’s wilder observations probably had little basis in reality, yet his attitude is extremely illuminating. It is indicative of the author’s attempts to psychologically dehumanise the Irish populace by measuring them against the ‘civilised’ standards of the reform movement. As a result, Gerald’s ‘superior other’ attitude was constructed predominantly around the Irish unwillingness to adhere to the revised sexual norms expounded by the ecclesiastical authorities. In Gerald’s eyes the Irish populations’ failure to adopt the reformers’ norms meant that they were unable to access the revised avenues to social and spiritual power that the reform movement had opened up to his own elite. The English self-image of sexual temperance and piety might have been unfounded and hypocritical; nevertheless, it clearly provided the invaders with a means to assert their cultural hegemony over their Irish counterparts.38 An incidental anecdote in Gerald’s Topographica Hibernica provides us with evidence that such attitudes were also adopted and espoused by his secular counterparts. In chapter XXI of distinctio II of this text Gerald relates that a certain deformed Irishmen had attended the court of the author’s cousin, Maurice fitzGerald, in Wicklow. According to Gerald’s account this unfortunate individual had features and limbs that resembled those of an ox. Gerald, who clearly felt some pity towards this disabled individual, relates how this man met a violent end at the hands of the local Irishmen. The reason that he gives for the untimely death of this individual is rather illuminating. The Irish natives of the place, because the youths of the castle often taunted them with begetting such beings on cows, secretly killed him

37  For examples, see HTI, cap. 42, p. 64, cap. 54, pp. 73–74, cap. 56, p. 75, cap. 102, p. 110. Gerald was almost certainly drawing on classical accounts of the Irish in this respect see D. Scully “At World’s End: Scotland and Ireland in the Graeco-Roman Imagination” in E. Longley et al. (eds.) Ireland (Ulster) Scotland: Concepts, Contexts, Comparisons (Belfast: Cló Ollscoil na Banríona, 2003), pp. 164–170, 167–168 and D. Scully, “The Portrayal of Ireland and the Irish in Bernard’s Life of Malachy”, D. Bracken and Dagmar Ó Riain-Raedel (eds.), Ireland and Europe in the Twelfth Century: Reform and Renewal (Dublin: Four Courts, 2006), pp. 239–256, 242–243. 38  “One of the ways in which the Anglo-Normans … justified their expansionism was by using religion as a validation for their actions, and a religious element formed an integral part of their picture of the barbarian.” Bartlett, Gerald, p. 167, see also Davies, English Empire, pp. 127–128.

752

Wyatt

(the deformed man) in the end out of odium and malice—a fate which he did not deserve.39 This account suggests that the ‘superior other’ attitude that Gerald expressed in his texts (and which was based very much upon reform norms of sexual behaviour) had also rubbed off on his secular peers. Like Gerald, the English warriors at Maurice’s castle appear to have psychologically degraded their Irish neighbours because of a perceived sense of sexual laxity and immorality. More significantly the violent reaction of the Irish natives would suggest that they both recognised and despised this condescending attitude of cultural supremacy.40 In addition to concerns regarding the lax sexual behaviour of their ‘Celtic’ neighbours the English commentators also regarded the ‘Celtic’ societies to be wild, lawless and lacking in centralised government.41 William of Newburgh felt that the politically fragmented nature of Irish society meant that it was continually plagued by war and slaughter.42 Similarly, Gerald of Wales commented that the Welsh … live on plunder, theft and robbery, not only from foreigners and people hostile to them, but also from each other. When they see a chance 39  “Illusis vero sæpissime Hiberniensibus teræ illius a juventute castri, quod tales in vaccis genuissent, ex suorum malitia et invidia quam non meruerat occulta nece demum interiit.” TH/EH, distinctio ii, cap. xxi, p. 108, translation modified from O’Meara, HTI, p. 74. O’Meara translates invidia as ‘envy’, however, it seems unlikely that this unfortunate character would have been the object of that emotion. Odium would, therefore, appear to be a more appropriate translation, see C. T. Lewis and C. Short, A Latin Dictionary (Oxford: Clarendon Press, 1879, repr. 1966), p. 995. 40  “Divergences of sexual behaviour are clearly capable of provoking violent and irrational antipathy In medieval Wales and Ireland this antipathy was allied with national hostility, resulting in a deep-seated prejudice on the part of the invaders. This prejudice, in its turn, was reinforced by the highest principles and blessed by the Church, thus creating an ideology able to sanctify the aggressiveness of the Anglo-Norman invaders, reinforce their hatreds, and allay their self doubts. The metamorphosis of Church reform into an ‘ideology of colonization’ was complete.” Bartlett, Gerald, p. 45. I concur with Bartlett on this point but would substitute his ‘national hostility’ with ‘ethnic hostility’. 41  A criticism also aimed by reform-minded clerics at the Slavs in this period, see Bartlett, Gerald, pp. 165–167. 42   Historia rerum Anglicarum, CRSHR, vol. i, p. 167. William’s contemporary, Gerald of Wales held very similar views, see HTI, cap. 99, pp. 106–107, cap. 100, pp. 107–108, cap. 101, pp. 108– 109, cap. 102, pp. 109–110.

Slavery And Cultural Antipathy

753

of doing harm, they immediately forget all treaties of peace and ties of friendship.43 Other twelfth-century English chroniclers, such as Ælred of Rievaulx and Richard of Hexham, felt that Scotland was infested with brigands over whom the Scottish monarch exercised very little control.44 Furthermore, the majority of these English writers commented with some dismay that the ‘Celtic’ populations continued to participate in activities that were regarded to be unchivalrous and uncivilised. Gerald of Wales noted that Welsh society was particularly prone to dynastic feuds and vengeance killings. The most frightful disturbances occur in their (the Welsh) territories … people being murdered, brothers killing each other and even putting each other’s eyes out, for as everyone knows from experience it is very difficult to settle disputes of this sort.45 Such attitudes were clearly formed in direct relation to the reformers’ objectives regarding the imposition of monogamous practices and social stability. Internecine strife was stimulated by the continuing practice of resource polygyny within the ‘Celtic’ communities which, in turn, acted against the Church’s centralising impulses.46 Political fragmentation continued to facilitate the ravages of the ‘Celtic’ warrior fraternities and acted directly against the imposition of the Truce of God within these regions. This perceived lack of order appears to have provided the English elite with a ‘civilising’ ethos that motivated them to impose order and orthodoxy upon their unruly neighbours.47 For example,

43  “Ad hæc etiam rapimis insistere raptoque vivere, furto, et latrocinio, non solum ad exteros et hostiles populos, uerum etiam inter se proprium habent. Pacis quoque et amicitæ fædera, visa nocendi opportunitate non respiciunt: fidei sacramentique religionem turpi postponentes lucro.” Descriptio Kambriæ, lib. ii, cap. ii, p. 207, translation from Thorpe, JDW, p. 257. 44  See chapter 2. 45  “Per quod graves terris eorum toties, nec sine cœdibus multis et fratricidiis, seu fratrum exoculationibus crebris, virium quoque omnium et successum experientia difficile sedabiles, emergere solent turbationes.” DK, lib. ii, cap. iv, pp. 211–212, translation from Thorpe, JDW. See Davies, English Empire, pp. 132–133 and for other examples see Gillingham, “Conquering”, pp. 79–81 46  Which is why the reforming hierarchy was attempting to alter such practices. 47  “It was politically useful for the Anglo-Normans … to be able to adopt a high moral tone towards these peoples they were attacking. It improved morale, provided a justification for aggression, and reinforced feelings of superiority held by the invaders. Some kind of

754

Wyatt

the author of the Gesta Stephani felt that the English conquerors and settlers in Wales … had perseveringly civilized it (Wales) after they had vigorously subdued its inhabitants; to encourage peace they imposed law and statutes on them (the Welsh); and they made the country so abound in peace and productivity that it might easily have been thought a second (Roman?) Britain.48 In particular, English concerns regarding the ‘lawlessness’ and ‘barbarism’ of their ‘Celtic’ neighbours were characterised by antipathy towards slave raiding and trading activities. During the central years of the twelfth century John of Salisbury complained that the Welsh were … rude and untamed; they live like beasts and despise the word of life, and though they nominally profess Christ, they deny Him in their life and ways. For they carry on a regular slave trade and sell Christians into foreign parts …49 Furthermore, John of Salisbury goes on to directly associate such slaving activities with the continuing existence of concubinage and incestuous unions within Welsh society.50 In doing this he illustrates that a clear psychological connection associating slavery with sexual laxity and moral corruption existed religious deficiency was attributed to the barbarians and became an integral part of the ethnographic characterization.” Bartlett, Gerald, p. 170. 48  “…  propiis incolis uiriliter edomitis, constanter exoluere; ad pacem confouendam, legem et plebiscita eis indixere; adeoque terram fertilem omnibusque copiis affiuentem reddidere, ut fecundissimœ Britanniæ nequaquam inferiorem œstimares.” GS, pp. 14–15. Gillingham translates ‘Britanniæ as ‘England’ here, see Gillingham, “Foundations”, p. 61. 49  “…  rudis et indomita bestiali more uiuens aspernatur uerbum uitae, et Christum nomine tenus profitentes uita et moribus diffitenturi. Ab his enim Christiani usitato commercio in partes transmarinas uenundati ab infidelibus concaptiuantur.” LJS, 87, p. 135. Once again this account was a recitation of complaints made by Meurig, bishop of Bangor, see above. 50  Ibid., John’s complaints do not refer to incest as we would understand it but rather to marriage within degrees of kinship prohibited by the strict rules of the reform movement i.e. between cousins or the ex-wives of brothers; practices which were relatively common and indeed politically expedient in the kin-based societies of Ireland and Wales. Nevertheless, as Millersdaughter has highlighted accusations of incest carried great power as they “constitute a substantial portion of the “ideology of colonization” with which medieval and early modern England justified its various conquests”, Millersdaughter, “Geopolitics of Incest”, p. 276.

Slavery And Cultural Antipathy

755

within the contemporary ecclesiastical mindset. Richard of Hexham made a similar connection when describing a Scottish slave raid upon the north of England in 1138. Richard felt that this raid had resulted in Scots carrying away “noble widowed matrons” (“solas nobiles matronas”) and “chaste maidens” (“castas virgins”).51 He depicted their Scottish captors as being devoid of sexual restraint, moral character and, therefore, humanity. He described their actions in the following, somewhat colourful, manner. … these bestial men (bestiales hominess), who regard as nothing adultery and incest and the other crimes, after they were weary of abusing these most hapless creatures (the widows and maidens) after the manner of brute beasts, either made them their slaves or sold them to other barbarians for cows.52 Both Richard of Hexham and John of Salisbury clearly felt that their ‘Celtic’ neighbours’ failure to adhere to the reformers’ norms meant that they could no longer be regarded, in any real sense, to be Christian peoples. This attitude, which is evident in many of the ecclesiastical narratives of the period, is synonymous with the uniform classification of the ‘Celtic’ peoples as ‘barbarians’. Although such attitudes were most prevalent amongst English ecclesiastics, they are also evident in other parts of Christendom too. The continental bishop and saint, Bernard of Clairvaux, commented that when his friend, colleague and fellow reformer, Malachy, became Bishop of Connor in Ireland in 1137, … the man of God understood that he had been sent not to men but to beasts. Never before had he known the like, in whatever depth of barbarism; never had he found men so shameless in regard to morals, so dead in regard of rites, so impious in regard of faith, so barbarous in regard of laws, so stubborn in regard of discipline, so unclean in regard of life. They were Christians in name, in fact pagans.53 51  DGRS, CRSHR, vol. iii, p. 157. Richard of Hexham was the reform-minded chronicler and prior at the Augustinian priory of Hexham, Northumberland. His account of Stephen’s reign was written shortly before the prior’s death in the 1140’s. 52  “Denique illi bestiales homines, adulterium et incestum ac cetera scelera pro nichilo ducentes, postquam more brutorum animalium illis miserrimis abuti pertæsi sunt, eas vel sibi ancillas fecerunt, vel pro vaccis aliis barbaris vendiderunt.” Ibid., translation from Anderson, SAEC, p. 187. 53  “Cum autem cœpisset pro officio suo agere, tunc intellexit homo Dei, non ad homines se, sed ad bestias destinatum. Nusquam adhuc tales expertus fuerat in quantacunque barbarie: nusquam repererat sic protervos ad mores, sic ferales ad ritus, sic ad fidem impios, ad leges barbaros, cervicosos ad disciplinam, spurcos ad vitam. Christiani nomine, re pagani.”

756

Wyatt

John Gillingham has argued that the twelfth-century accounts that classify the ‘Celtic’ populations as barbarians are indicative of a significant ideological shift in the attitudes of the English elite. He feels that this shift signals the birth of a definitive ‘English’ identity that was based primarily upon a sense of economic and social superiority. Gillingham argues that this superior attitude first became evident within the writings of William of Malmesbury.54 He, therefore, feels that William was … an early exponent of the splendid English habit of regarding the course of English history as the triumph of civilisation over barbarism. William’s crucial intellectual step was to take the religious component out of the concept of barbarian and redefine it in terms of secular material culture … In the earlier period the word barbarus had been synonymous with paganus, but in William’s eyes the Christian Celts of his own day were ‘barbarians.’55 I would argue that Gillingham’s hypothesis is flawed in two crucial respects. Firstly, William’s negative attitude towards the ‘Celts’ was hardly the novel invention of that author and, secondly, the notion that William of Malmesbury defined his sense of English ‘otherness’ purely “in terms of secular material culture” fails to acknowledge either his motivations or context.56 Indeed, this appears to be an extraordinary statement with regard to a monastic author writing from within an ascetic milieu during an era of Church reform. William’s monastic background makes it extremely difficult for us to extricate his opinions on the secular world from his religious sensibilities. His opinions may well signal the emergence of a distinctly hybrid English identity in the post-Conquest era. They may also be suggestive of a fundamental shift in the attitudes of the English elite towards their ‘Celtic’ neighbours. They may, in addition, signal an increasing awareness of England’s economic hegemony within the British Isles (albeit drawing on similar notions evident in classical De Vita et Rebus Gestis S. Malachiæ, PL, J. P. Migne (ed.) (Paris: Migne, 1854), 182, p. 1084, translation from Lawlor, Life of St. Malachy, p. 37. St. Bernard wrote this Vita in around 1149 following Malachy’s death, see also chapter 2. The tone of the tract suggests that Malachy was a foreigner to Ireland yet, in truth he had been born and raised there. Nevertheless Malachy’s reform sensibilities clearly set him apart from his ‘barbarous’ countrymen in the mind of Bernard of Clairvaux. For other examples of similar attitudes see Gillingham, “Conquering”, pp. 67–84. 54  A contention supported by Thomas, English and Normans, p. 311. 55  Gillingham, “Conquering”, p. 69 and “Foundations”, p. 59. See also English, pp. 43, 104. 56  Gillingham, “Foundations”, p. 59.

Slavery And Cultural Antipathy

757

texts).57 Yet, despite all of this, there is nothing which is definitively ‘English’ about William’s attitudes in this respect, apart from the fact that he was an English exponent of the reform ideal. As such William was drawing on well established arguments and allusions evident in both the classical texts and the contemporary reform discourse of continental Christendom. We have seen how the continental bishop Bernard of Clairvaux, writing at around the same time, expounded attitudes towards the ‘Celtic’ peoples which were extremely similar to William’s. Moreover, Bernard was undoubtedly drawing on a well established tradition of anti-‘Celtic’ sentiment amongst reform-minded ecclesiastics in northern France.58 An early example of this tradition can be found in Warner of Rouen’s satirical poem Moriuht composed during the first quarter of the eleventh century.59 Warner was a scholar steeped in classical literature and as such would have almost certainly been aware of the highly pejorative Graeco-Roman representations of Ireland and its inhabitants evident in the works of ancient writers.60 Warner and his continental contemporaries were most probably also aware that these ancient scholars’ classification of the Irish as ‘barbarians’ had been dramatically modified and revised by early medieval ecclesiastics such as Bede. Bede had emphasised the rehabilitation and spiritual vitality of the Irish following their acceptance of Christianity in the fifth century.61 Yet, rather than following Bede’s more positive assessment of this fellow Christian people, a significant number of Norman and Frankish ecclesiastics, including Warner, chose instead to reinvigorate those classical sentiments which regarded the Irish as lustful, brutal barbarians beyond the pale of civilisation.62 It was no coincidence that Warner was writing at the same time and in the same geographical location which witnessed the dramatic expansion of the Peace of God movement and the infectious spread of reforming zeal.63 Indeed, Warner’s invective against the lascivious Moriuht clearly identifies him as a proponent of the ascetic ideal of reform.64 Furthermore, his text is littered with pejorative 57  See, Killeen, “Ireland” and Scully, “World’s End”. 58  See McDonough, Moriuht, pp. 48–51 for examples. Hugh Thomas has also noted that such sentiments were particularly virulent in pre-1066 Normandy, see English and Normans, p. 310. 59  For a fuller discussion of this text see Chapter 3. 60  McDonough, Moriuht, pp. 12–13, 20, 37–40, 45–51. 61  A perspective also taken by the ninth-century author Walahfrid Strabo, see Scully, “Malachy”, pp. 242–245. 62  For examples see McDonough, Moriuht, pp. 48–51. 63  Moore, Revolution, p. 8. 64  McDonough, Moriuht, pp. 32, 35–37.

758

Wyatt

comments about the Irish ‘Celts’, denigrating their appearance and dress and associating them with laziness, backwardness, treachery, paganism, beast-like qualities, sinfulness, sexual immorality and slavery: both spiritual and actual.65 In the opening section of his poem Warner provides the following ethnographic description of the inhabitants of Ireland and their shortcomings: Where the sun sets lies an isle, called Ireland, a fertile (land) though not well tended by its inhabitants. Many people say that if this (island) was occupied by a nation of any skill, it would surpass Italy in its riches. Regrettably, though this (land) affords a view of life in paradise from all sides, it rears and fosters Irishmen who lack refinement … Though these things be granted to them, in addition, under the frame of their body, their mind is crushed. They are also unaware of the light of God enthroned on high. How much they run after prostitutes with their bodies, aided by their mode of dress! Word has been brought to me: they couple like animals; they do not wear trousers, because they are constantly locked in sexual activity. For these people honour Venus because of her sexual passion for Mars. Phew! They defile beds, like a one-year old child … And up to the present time many facts have been reported to me about these Irish. It is immoral to record them and it shames (me) to recount (them).66 Warner of Rouen therefore issued forth a set of condemnations calumniating ‘Celtic’ social norms and moral fibre some hundred years before William of Malmesbury and subsequent twelfth-century writers like Gerald of Wales penned an almost identical set of negative assessments regarding the ‘barbarian’ nature of their neighbours. Although Warner was the product of a different society and cultural milieu he had one very clear thing in common with these twelfth-century English clerics; the ascetic norms and objectives of the reform movement. It should be no surprise, then, that when we examine the vociferous twelfth-century English condemnations of the ‘Celtic’ populations we 65  Ibid., pp. 74–77, 83, 85, 87. 66  “Solis in occasum iacet insula Scottia dicta, Fertilis, a populo non bene culta suo; Vt dicunt plures, hanc gens si gnara teneret, Vinceret Italiam fertilitate sua, Haec paradisiacam prospectans undique uitam, Heu! nutrit Scottos et fouet inlepidos … His sibi concessis, pressa quoque mente sub artu Carnis, et ignari luminis [et] Altithroni. Quantum per habitum sectantur corpore scortum! Est mihi perlatum: more cubant pecudum; Non braccas portant, ueneri quia semper adherent; Illis namque Venus Mortis amore decus. P!lectos uiolant, anni uelut unius infans … Svnt et adhuc de his Scottis mihi multa relata, Scribere quod nefas est quodque referre pudet.” Ibid., pp. 74–75.

Slavery And Cultural Antipathy

759

discover that the majority of them also emanated from ecclesiastical authors with reforming pretensions.67 Their foremost criticisms regarding the ‘Celtic’ populations were founded upon the underlying objectives of the reform movement that sought to eradicate traditional expressions of virility, curtail warrior violence and create a more stable and peaceful Christian society. It was the reform movement, therefore, that provided the ideological template for the English cultural and political domination over the ‘Celtic’ regions. Pope Adrian IV’s Laudabiliter, the papal privilege allegedly granted to Henry II in 1159 makes this clear. … we regard it as pleasing and acceptable to us that you (i.e. Henry II) should enter that island (i.e. Ireland) for the purpose of enlarging the boundaries of the church, checking the descent into wickedness, correcting the morals and implanting virtues, and encouraging the growth of the faith of Christ …68 Adrian therefore sanctioned any prospective English conquest of Ireland in the name of reform just as Pope Alexander II had sanctioned William’s conquest of England in 1066. The terminology employed by Adrian IV is strikingly similar to that used by Alexander II in his letter to William the Conqueror of October 1071.69 67  See examples below. 68  “… gratum et acceptum habemus ut pro dilatandis ecclesie terminis, pro viciorum restringendo decursu, pro corrigendis moribus et virtutibus inserendis, pro Christiane religionis augmento …” EH, lib. ii, cap. v, pp. 146–147. This was further reinforced by a privilege granted by Pope Alexander III, see ibid., p. 147. Both documents suggest that the Irish population were Christians in name only. There has been some historical debate concerning whether Laudabiliter was a contemporary forgery made in support of the interests of the English crown. However, it is now generally accepted that Adrian IV did issue a document sanctioning Henry’s expedition to Ireland. The redactor of the said document, Gerald of Wales, certainly believed in its authenticity. 69  This stated “We learn at this time of your outstanding reputation for piety among the rulers and princes of the world and we receive unmistakable evidence of your support for the Church: you do battle against the forces of simoniacal heresy and you defend the freedom of catholic rites and customs. But since the crown of reward is promised not to those who begin well but those who are proven at the end, we urge your excellency zealously to persevere in your most Christian devotion.” See also chapter 4. Brendan Smith has observed: “The papal banner carried to England by William, duke of Normandy, in 1066 symbolised an alliance between Church reformers and sympathetic princes which was to persist in the British Isles throughout the twelfth century and beyond. In 1171 the second conquest of a Christian kingdom by a neighbouring Christian power in the region

760

Wyatt

Attempts to disseminate the ‘civilised’ values of chivalry and the reform movement into the psyche of the ‘Celtic’ populations were not confined to ventures of militaristic and political domination. The ‘Celtic’ elite might be courted into the chivalric orbit of their English counterparts with the use of lavish gifts, flattery and prestigious marriage alliances.70 It appears that Henry I attempted to draw ‘Celtic’ nobles into his circle in the same way that his father had attempted to incorporate Anglo-Saxon leaders like Eadric the Wild and Waltheof of Northumbria into the post-Conquest elite. In order to illustrate this point in more detail let us briefly examine the career of the twelfth-century Scottish king David I (1124–1153). David, the youngest son of the prolific slave raider Malcolm Canmore, was brought up in England at the royal court of Henry I. Moreover, Henry appears to have spent a great deal of time and energy cultivating David in the revised manners and mores of the English elite.71 The English king is said to have lavished both gifts and attention on the Scottish prince.72 David also received a chivalric military training and subsequently served as a knight in Henry’s service in Normandy.73 In 1124 Henry bestowed upon David a wealthy English bride named Matilda. This was a prestigious union that was clearly intended to incorporate David into the English royal circle as Matilda was the daughter of Judith, the Conqueror’s niece.74 Henry’s strategy of acculturation should, by now, be extremely familiar indeed the twelfth-century English accounts that describe David’s character suggest that it was relatively successful. William of Malmesbury remarked that David was occurred with papal support in Ireland, reinforcing the message that within Christendom secular borders did not take precedence over the need to reform the Church.” Smith, “Frontiers”, p. 244. 70  For examples see Davies, Domination, pp. 6, 48–58. Davies argues that by doing this “… the kings and nobles of England were demonstrating their social superiority and the magnetism of their courts and their mores.” 71  Ibid., pp. 50–51. 72  See OV, vol. iv, pp. 274–275. 73  Henry subsequently gave David an estate at Querqueville near Cherbourg, see Davies, Domination, p. 50 and G. W. S. Barrow, David I of Scotland (1124–1153) The Balance of New and Old (Reading: University of Reading, 1985), pp. 16–17. 74  Henry’s choice of bride in this instance is interesting as Matilda was the daughter of Judith, the Conqueror’s niece and Waltheof the Anglo-Saxon earl who had failed to modify his uncivilised habits and subsequently been punished for treason. The union may have, therefore, been deliberately intended to remind David that his position at the English court depended upon his continuing adherence to chivalric mores. This appears to have been reinforced when King Stephen confirmed David’s possession of Waltheof’s old earldom in 1136, see DGRS, CRSHR, vol. iii, p. 146.

Slavery And Cultural Antipathy

761

… a young man of more courtly disposition than the rest, since he had from boyhood been polished by familiar intercourse with us (i.e. the English), and rubbed off the barbarian gaucherie of Scottish manners.75 The central justification that William of Malmesbury gave for his portrayal of David as a ‘civilised’ and enlightened individual was that the Scottish king had adhered to the behavioural norms of the reform movement. William noted that David and his brothers, Edgar and Alexander, had all been royal role models because … they successfully overcame the vice most prevalent in kings, and it is recorded that no woman entered their bedchamber except their lawful wives, nor did any of them bring a stain upon his innocence by keeping any mistress.76 A few years later William of Newburgh also commented upon David’s commendable sexual continence labelling him “a civilised king of an uncivilised race.”77 It, therefore, appears that the English elite deliberately attempted to ‘educate’ the rulers of the more traditional communities of Britain in the prestigious mores of the reform movement. This was probably done in the hope that leaders, such as David I, would conduct their affairs in a more tempered manner than had their fathers. Clearly the prohibition of resource polygyny and the imposition of the reform rules concerning monogamy were of central significance to this strategy. Both William of Malmesbury and William of Newburgh appear to have felt that if those individuals at the pinnacle of Scottish society could be made to adopt such norms then they might be more

75  “… iuuenis ceteris curialor et qui, nostrorum conuictu et familiaritate limatus a puero, omnem rubiginem Scotticæ barbariei deterserat.” GRA, pp. 726–727. William of Malmesbury also remarked that David had offered his subjects financial incentives in return for their adoption of more ‘civilised’ mores, ibid. 76  “… ita domesticum regibus uitium euicerunt ut numquam feratur in eorum thalamos nisi legitimas uxores isse, nec eorum quemquam pelicatu aliquo pudicitiam contristasse.” Ibid. 77  “… rex non barbarus barbarae gentis …” The History of English Affairs, Book I, P. G. Walsh and M. J. Kennedy (eds. and trans.) (Warminster: Aris & Phillips, 1988), pp. 102–103. Moreover, David’s reform-minded nature was evident even before he acceded to the throne. For example, in 1113 he was responsible for resettling a community of Frankish monks at Kelso on the Tweed, see Barrow, Kingship and Unit, p. 81.

762

Wyatt

easily disseminated into the middling and lower orders.78 Yet, as we shall see this strategy of English acculturation at the highest level was only partially successful. Indeed, as the secular and ecclesiastical elite within ‘Celtic’ communities were drawn into the cultural orbit of reformed Christendom violent and disruptive tensions were ignited within their communities. Whilst this acculturation was being undertaken the conflicts between traditional patriarchal values and the revised behavioural codes of the reform movement were played out, once again, in a dramatic fashion. In order to illustrate this let us now turn our gaze to events in Wales during the early decades of the twelfth century.

Acculturation, Antipathy and the Welsh Warrior

Due to geographical proximity and longstanding political and social associations the invasive tendrils of English cultural influence were, perhaps, felt most immediately and most keenly within the communities of medieval Wales.79 In the decade following the Conquest individual Norman lords began to make territorial inroads along the river valleys of the Marches, into the lowland areas of South Wales, and even into the mountains of Gwynedd. These invaders established themselves in powerful lordships and colonised them with settlers from England and France. The intrusion of these settlers appears to have placed phenomenal strains upon the structures of native Welsh society. These strains initially came to a head between 1094 and 1098, when there were several Welsh uprisings against their new territorial overlords.80 Furthermore, traditional slave raiding activities appear to have characterised these uprisings. For example, in around 1093, the Norman knight, Robert of Rhuddlan was slain whilst attempting to repel a Welsh slaving raid upon his newly established castle at Rhuddlan.81 In addition, the author of the Historia Gruffud vab Kenan remarked that after an assault on a Norman garrison on Anglesey the Welsh

78  This process of ‘acculturation’ was further facilitated by David’s settlement of English noblemen upon the lands of Southern Scotland, G. W. S. Barrow, The Anglo-Norman Era in Scottish History (Oxford: Clarendon, 1980), pp. 61–118. 79  The Anglo-Saxon elite had undoubtedly exerted political and economic influences over the Welsh territories for centuries. Nevertheless, the territorial incursions into Wales following the Norman Conquest significantly intensified such influences. 80  R. R. Davies, Conquest, Coexistence and Change, Wales 1063–1415 (Oxford: Clarendon Press, 1987), p. 35. 81  OV, vol. iv, p. 141.

Slavery And Cultural Antipathy

763

“travelled back with the booty, with French and English in fetters”.82 The author of the Brut y Tywysogyon noted that in 1094 the Welsh … destroyed all the castles of Ceredigion and Dyfed, except two, that is, Pembroke and Rhyd-y-gors. And they took with them the people and all the cattle of Dyfed, and they left Dyfed and Ceredigion waste.83 The increase in Welsh slave raiding activities during this period appears to have been symptomatic of the growing cultural antipathy between traditional elements of Welsh warrior class and the external cultural influences exerted by the new ‘chivalric’ elite. The dynamics of this cultural antipathy are perhaps best illustrated by the career of two men in particular; Cadwgan ap Bleddyn, the Welsh ruler of Powys, and his son, Owain ap Cadwgan. The native chronicler of the Brut y Twywsogyon had a great deal to say about these two individuals and clearly felt that their volatile relationship with Henry I had been extremely significant. Cadwgan ap Bleddyn (d. 1111) was one of the many sons of Bleddyn ap Cynfan, ruler of Powys, who was murdered by his rival Rhys ab Owain in 1075.84 Early in his career Cadwgan appears to have behaved in manner befitting a traditional Welsh prince attempting to assert his place within a patriarchal hierarchy Following the death of Rhys ap Tewdwr, king of Deheubarth, in 1093 we find Cadwgan staking a claim to a portion of Rhys’ territory by opportunistically ravaging there.85 Cadwgan appears to have been aggrieved when the external forces of the Norman lord William fitz Baldwin subsequently stepped into the political vacuum and assumed control of Deheubarth. As a result Cadwgan appears to have staunchly resisted the new intruders and he was heavily involved in the insurgency of the 1090s. In 1096 his warband was said to have attacked

82  “Ac val kynt y kerdasssant wynteu dracheuyn a’r anreith, ac a Freinc a Saesson en rwym ganthunt ac en garcharoryon.” HGK, pp. 42, 73. 83  “… gestyll Keredigyawn a Dyuet eithyr deu, nyt [amgen] Penuro a Ryt y Gors; a’r bobyl a’r holl any[ueileit] Dyuet a dugant gantunt ac adaw a oru[gant] Dyuet] a Cheredigyawn ynn diffeith.” ByT, pp. 34–35, see also Annales Cambriæ J. Williams ab Ithel (ed.) (London: Longman, Green, Longman, and Roberts, 1860), pp. 34–35. The events of 1094 are further supported by the later and slightly less reliable Brenhinedd y Saesson, pp. 86–87. 84   ByT, pp. 28–99. 85  Ibid., 32–33. Cadwgan’s actions appear to have been motivated by a traditional feud between the dynasties of Powys and Deheubarth for Rhys ap Tewdwr had killed two of Cadwgan’s brothers in 1088, see also Davies, Coexistence, p. 73.

764

Wyatt

the castle of Pembroke and “… plundered it of all its cattle and ravaged the whole land; and they returned home with vast spoil.”86 In 1098 Cadwgan fled to Ireland to escape a massive incursion into North Wales by forces under the command of Hugh of Avranches, earl of Shrewsbury.87 Yet, the chronicler of the Brut relates that in the following year Cadwgan returned to Wales and made peace with the invaders after which he received a portion of his patrimony in Powys.88 Following this submission Cadwgan’s behaviour appears to have become remarkably modified. It is true that he became involved in a rebellious plot against Henry I in 1102 yet, his co-conspirators in this instance were not Welsh warriors but the Norman nobles Robert, earl of Shrewsbury, and his brother Arnulf of Montgomery, earl of Pembroke.89 Cadwgan appears to have been forgiven by King Henry for this misdemeanour and he was subsequently allowed to receive control over Ceredigion and further lands in Powys.90 Indeed, Cadwgan prospered from his cooperative attitude towards the English monarch and in 1109 he is said to have prepared a great feast for all of the rulers of Wales. This may well have been a deliberate attempt to emulate the munificence of his ‘chivalric’ eastern neighbour.91 Yet, Cadwgan’s increasing connection and affiliation with the English and continental elite appears to have created tensions within his own household. This is evident from the actions of Cadwgan’s son, Owain, who reacted violently to the intrusion of external cultural influences into his father’s court. Owain expressed his displeasure in a manner analogous to that adopted by the mythical Welsh warrior Efnisien in the Welsh tale Branwen Uerch Llyr.92

86  “… a’e yspeilaw o’e holl anyueileit a diffeithaw yr holl wlat; a chyt a diruawr anreith yd ymhoelassant adref.” ByT, pp. 36–37. 87  Cadwgan escaped to Ireland along with another Welsh prince, Cruffudd ap Cynan of Cwynedd, ibid., pp. 37–39. 88  Ibid., pp. 38–39. 89  Ibid., pp. 41–42. 90  Ibid., p. 47. 91  Ibid., p. 55. In addition, Cadwgan later married a continental bride (the daughter of the Norman noble, Picot of Sai) named his son by her ‘Henry’, and built a Norman style castle at Trallwng Llywelyn near Welshpool, ibid., p. 63 and Davies, Coexistence, p. 67. 92  Like Owain, Efnisien resisted external cultural influences using violent and antisocial behaviour. In addition, he opposed the mixing of Welsh blood by murdering his own nephew who was the product of a union between his sister, Branwen, and the Irish king Matholwch. See chapter 2. The similarities in behaviour may be no coincidence, indeed, Katherine Millersdaughter has argued that the written reproduction of the Four Branches, during the late 11th century early twelfth century should be regarded as an “anti-colonialist

Slavery And Cultural Antipathy

765

The chronicler of the Brut y Tywysogyon relates that during Cadwgan’s great feast of 1109 Owain heard much talk of a beautiful Welsh noblewoman named Nest. Nest was the wife of the Norman castellan Gerald of Windsor and, even more significantly, she had also been a mistress of the English king, Henry I.93 During the night that followed the feast Owain besieged Gerald’s castle with a small war band and raped Nest and abducted her along with Gerald’s children and much of his treasure.94 Yet, this was no random act of youthful drunkenness and delinquency. It was a symbolic statement against the cultural intrusion of external mores into the circle of the Welsh elite and an expression of the traditional warrior values of power and virility. Through this action Owain was directly defying his father’s will. He was also protesting against Cadwgan’s apparent rejection of traditional cultural values. In addition, Owain was emphasising his power by dishonouring and symbolically emasculating both Henry I and his vassal, Gerald of Windsor, who had dared to occupy Welsh territory. As was often the case in traditional patriarchal societies Owain’s overtly political action was characterised by sexual violence and female abduction. Owain’s rapacious behaviour was rendered all the more symbolic, both culturally and politically, because his chosen victim was a Welsh noblewoman who had been sleeping with the enemy. His actions appear to have had the desired effect as the Brut relates that upon hearing of this outrage his father … was indignantly grieved thereat because of the rape that had been committed upon Nest … and also for fear lest king Henry should be enraged at the injury to his steward.95 Cadwgan subsequently ordered his son to return the castellan’s wife and property but to no avail. Following Owain’s refusal of compliance, Richard Bishop of London, King Henry’s steward at Shrewsbury, placed a price on Owain’s head and arranged for the war bands of Ithel and Madog ap Rhiddid, sworn

discourse” within in the geopolitical context of the Anglo-Norman expansion into Wales, see “Geopolitics of Incest”, p. 285. 93  Gerald of Wales notes that Henry fitz Henry was the illegitimate son of an adulterous union between Henry I and Nest, see JDW, lib. ii, cap. vii, p. 189. This contention is also supported by the Annales Cambriæ, p. 47. See also J. E. Lloyd, A History of Wales (London: Longmans, Green, 1912), vol. ii, p. 499. 94   ByT, pp. 56–57. 95  “… kymryt ynn drwc arnaw gann sorr a oruc ef hynny o achaws y treis gyt a wnathoedit a Nest uerch Rys, a heuyt rac ouyn llidyaw Henri urenhin am sarhaet y ystiwart.” ByT, pp. 56–57.

766

Wyatt

enemies of Owain’s kindred, to hunt down and capture the Welsh renegade.96 Owain appears to have learned of this plot and subsequently fled to Ireland as a fugitive. Moreover, Cadwgan appears to have felt that his paternal inaction had implicated him in Owain’s crime and he soon accompanied his son into exile.97 In spite of these initial fears the influence of Henry I’s charmed circle appears to have had a profound impact upon Cadwgan and he returned shortly afterwards and submitted to the king. For his part Henry appears to have followed the reconciliatory policy of his father. He received Cadwgan with favour and attempted to draw him even further into his orbit by rewarding the Welsh Prince with the return of his territory and a noble French bride.98 Yet, the Brut y Tywysogyon relates that Cadwgan achieved this favourable settlement upon one very strict condition: … that there was to be no association or friendship between him and Owain, his son, and he was not to allow him to come into the land, and that he was not to give him either counsel or advice, either support or help.99 In this short extract the Brut’s chronicler lays bare the dynamics and stark reality of Henry’s strategy of acculturation which could undermine and even sever the closest ties of Welsh kinship. As for Owain he continued to act in a manner befitting the leader of a traditional warrior fraternity. He returned from Ireland shortly after Cadwgan and allied himself with the war band of Madog ap Rhiddid. This alliance is particularly significant because, as already noted, Madog was a long standing enemy of Owain’s kindred. The Brut notes that shortly after Owain’s exile, Madog had ravaged and taken possession of Owain’s portion of Powys.100 Moreover, Richard bishop of London 96  Ibid., pp. 57–59. It is interesting that Bishop Richard should have become involved in such a quarrel. Moreover, it seems that he employed William the Conqueror’s tactics of using native forces to deal with traditionally minded enemies. Significantly, Richard secured the support of Ithel and Madog by promising that they would be accepted into Henry’s charmed circle, see ByT, pp. 56–57. 97  Ibid., pp. 58–59. Indeed, Cadwgan may have been expecting Henry I to react with traditional ferocity towards this act of sexual violence against one of his concubines. 98  See also notes 74 and 91, above. 99  “A rodi a oruc y brenhin y [kyuoeth] Cadwgawn trwy yr amot hwnn yma hyt na bei gytymeithas na chyfueillach rygtaw ac Owein, y vab, ac na add idaw dyuot y’r wlat, ac na rodei idaw na chygor na chussul na nerth na channhorthwy.” ByT, pp. 62–63. 100  Ibid., pp. 58–61.

Slavery And Cultural Antipathy

767

had previously employed Madog and his brother to hunt Owain down and neutralise him earlier that same year. So how had these two sworn enemies been drawn together? The Brut notes that, following Owain’s escape to Ireland, Madog had subsequently quarrelled with Richard bishop of London because he was harbouring and protecting a dissident contingent of traditional ‘Saxon’ warriors who were ravaging Marcher territories and then escaping back to Madog’s lands.101 Richard therefore ordered Madog to hand over these ‘Old English’ (Saesson) warriors however he refused to do so. It was at this point that Madog ap Rhiddid made an alliance with his bitter rival, Owain ap Cadwgan, indeed the author of the Brut was compelled to comment how … peace was made between those who before that were enemies. And they made a solemn pledge upon relics that neither would make peace with the king without the other, and neither of them would betray the other.102 The actions of these two Welsh warriors following this pledge is very interesting, Owain’s war band joined forces with both the warriors of Madog ap Rhiddid and the dissident ‘Saxons’ in an inter-ethnic pact of violence which is highly reminiscent of the one made between the fían band of the sons of Donn Désa and the British warrior leader Ingcél Gáech in the fictional Irish tale Togail Bruidne Da Derga.103 These three war bands further swelled their ranks with local ynfydyon warrior-youths and ravaged the coastal territories, plundering West Wales in a similar manner to the díberg raids undertaken in that Irish tale. The Brut’s description of their activities cited already in chapter 2 is worth repeating in this context:

101  “Saesson” Ibid., pp. 62–63. Hugh Thomas has argued that the presence of these English warriors provides evidence of the significant ethnic hostility that continued to exist between traditionally minded Anglo-Saxon elements and the French speaking elite in England. Interestingly, Thomas also relates the activities of these English warriors to the post-Conquest silvatici, who we have already seen, were a warrior fraternity that is comparable, in many senses, to the Welsh ynfydion, see English and Normans, p. 63. 102  “A hynny a gauas, a gwneuthur hedwch rwg y rei a oedynt ynn elynyon kyn no hynny. Ac ymaruoll uchbenn creireu a wnaethant hyt na hedychei un a’r brenhin hep y gilyd, ac na uredychei un onadunt y gylyd.” ByT, pp. 64–65. 103  See chapter 2.

768

Wyatt

… and for the second time they went and summoned ynfydyon from Ceredigion to add to their numbers, and they made for the land by night and burned it and slew all they found in it, and plundered others and took others with them in fetters and sold them to their folk or sent them bound to the ships.104 Provoked by external cultural forces which threatened to undermine traditional warrior norms (just as the traditional warrior order of the sons of Donn Désa had been threatened by the peaceful rule of king Conaire) the ynfydyon united with other like-minded war bands who would, under normal circumstances, be mortal enemies, and together they indulged in an orgy of violent behaviour and traditional slave-raiding activities. Cadwgan’s failure to control the disruptive behaviour of his wayward son appears to have exasperated Henry I who subsequently confiscated Cadwgan’s territory and held him hostage at the royal court.105 Cadwgan’s fall from grace did not last for long as following the death of his brother Iorwerth, ruler of Powys, in 1111 Henry granted him his patrimony. Yet, Cadwgan’s ascendancy in Powys was to be somewhat short-lived and he was murdered shortly afterwards by Owain’s treacherous ally Madog ap Rhiddid.106 At this juncture Henry I attempted to neutralise the disruptive behaviour of the traditional warriors Madog and Owain by, once again, pitting them against each other. He, therefore, forgave Owain ap Cadwgan for the rape of Nest and summoned him to the royal court. The prestigious magnetism of the English court proved too powerful for Owain and he subsequently made peace with Henry and was granted his father’s lands in Powys.107 Following Owain’s submission it appears that Henry attempted to integrate this violent Welsh warrior into his chivalric circle. The chronicler of the Brut y Tywysogon suggests that Henry took a personal interest in ‘rehabilitating’ Owain. Indeed, he placed the following words into the mouth of the English king during a meeting he held with Owain in 1114.

104  “Ac etwa y mae[nt] ynn trigyaw ynn teruyneu y wlat. A’r eilweith yd aethant a galw ynuydyon Keredigyawn y chwanecau y rif, a chyrchu dros nos y wlat a’e llorci, a llad pawb o’r a gawssant yndi, ac yspeilaw ereill a dwyn ereill gantunt yg karchar, ac eu gwerthuy dynyon neu y hanuon ynn rwym y’r llogeu, a gwedy llosci y tei a llad a gawssant o’r anyueileit.” ByT, pp. 68–69. 105   ByT, pp. 70–71. 106  Ibid., pp. 74–75. 107  Ibid., p. 77. Furthermore, Henry’s policy was successful as within just over a year Owain had captured the renegade Madog and personally removed his eyes.

Slavery And Cultural Antipathy

769

This I will tell you. I am going to Normandy, and if you will come with me I will make good for you everything that I have promised. And I will make you an ordained knight.108 The Brut subsequently reports that Owain accompanied Henry I to Normandy and he appears to have acquitted himself well there as the chronicler relates that following the expedition Owain received all that he was promised.109 It would seem that another traditionally-minded warrior had been tamed by the familiar strategy of acculturation and then been received into the ‘civilising’ bosom of royal favour. Yet, as we shall see this was not quite the end of Owain’s story. In the following year (1115) the disenfranchised prince of Deheubarth, Gruffudd ap Rhys, returned to Wales following a long exile in Ireland and began to reassert his claim to power by behaving with traditional warrior ferocity.110 Gruffudd appears to have been aware of the cultural antipathy that had been simmering within Welsh society. He, therefore, adopted a strategy that consistently targeted English strongholds such as Swansea, Llandovery, Tywi and Aberystwyth. Gruffudd clearly intended to capitalise upon the tensions created by the English acculturation of the Welsh elite. He appears to have been very successful in this respect as the Brut y Tywysogion remarks that following Gruffudd’s successful ravages Thereupon ynfydyon of the land gathered to him from all sides, thinking because of that incident that he had overcome everything.111 The chronicler’s further use of the term ynfydyon in this instance is extremely interesting. As we have seen this was a term that was synonymous with the traditional activities of the Welsh warrior fraternity.112 Babcock has noted that the term was used only five times in the whole chronicle and only in reference 108  “ ‘Hynn a dywedaf it. Mi a af y Normandi, ac o deny di gyt a mi, mi a gywiraf it pop peth o’r a edeweis it. A mi a’th wnaf yn varchawc urdawl.’ ” Translation modified from Jones, ibid., pp. 82–83. 109  Ibid. 110  Ibid. Gruffudd ap Rhys was brother of Nest and the son of Rhys ap Tewdwr, the deceased king of Deheubarth. He was, therefore, not particularly well disposed towards Owain ap Cadwgan! 111  “Odyna yd ymgynnullassant jeueinc ynuytyon y wlat o poptu attaw o tybygu goruot ohonaw ar pop peth o achaws y damwein hwnnw.” ByT, pp. 88–89. 112  See chapter 2.

770

Wyatt

to Welsh nobles who attacked the invaders and settlers from England.113 He has, therefore, hypothesised that Owain ap Cadwgan and Gruffudd ap Rhys’s followers were specifically referred to as ynfydyon because they contravened the traditional bonds of loyalty and deserted their lords, who were doing castle duty for the English.114 An alternative and perhaps more convincing interpretation might be that the ravages of the ynfydyon were an expression of deepseated cultural anxiety emanating from traditional warriors of middling rank. This cultural reaction was directed against their elite’s increasing adoption of external cultural values and ties of allegiance. It was the elite Welsh castellans standing guard over English strongholds that had been forced to defend against Gruffudd’s offensives. Furthermore, as the example of Cadwgan ap Bleddyn reveals, it was the native elite rather than the youthful warriors who were busy breaking the bonds and behavioural norms of traditional society. The negative attitude of the Brut y Tywysogyon towards the ynfydyon may be explained by the author’s ecclesiastical sensibilities that could never condone the actions of a group so closely associated with violent disorder. The inclusion of traditional Anglo-Saxon warriors in Owain and Madog’s orgy of violence in 1110 reveals that this was not simply a ‘nationalistic’ or ethno-centric movement.115 Indeed, Gruffudd ap Cynan, the king of Gwynedd, and his sons similarly allied themselves with like-minded Hiberno-Norse warriors and pursued traditional slave raiding activities in opposition to their chivalric counterparts.116 Such activities were not only uprisings against English domination. They were a defiant gesture of adherence to traditional warrior values in the face of the increasing infiltration of external cultural norms moulded by the reform movement. Henry I appears to have recognised the nature of the disruptions caused by Gruffudd ap Rhys and his ynfydyon and in 1116 he summoned the most suitable man for removing this troublesome element. The author of the Brut describes his actions in the following manner: In the meantime king Henry sent messengers to Owain ap Cadwgan to bid him come to him … And when he came, the king said to him, “Owain, my most beloved, do you know that petty thief Gruffudd ap Rhys who is rising up, as it were, against my magnates? And since I believe you are a

113  Babcock, “Imbeciles”, pp. 1–11. Indeed, the term is first applied in 1110, in reference to the slaving activities of Owain ap Cadwgan’s warrior band in Dyfed. 114  Ibid., pp. 6–8. 115  See also chapter 4. 116  Wyatt, “Gruffudd”, pp. 596–617.

Slavery And Cultural Antipathy

771

man most true to me, I would like you to lead a host, along with my son, to drive out Gruffudd ap Rhys.”117 Henry, therefore, continued to employ the tried and tested tactic, used so successfully by his father, of eradicating the threat posed by traditional warrior elements by employing a similar force to seek out and destroy them. Owain’s subsequent actions reveal that his newly acquired chivalric persona had particularly shallow roots. Indeed, he appears to have prosecuted this royal task with the vicious brutality of a traditional warrior. The author of the Brut relates that Owain sent his warriors into the woods where Gruffudd ap Rhys and his forces were thought to be taking refuge and ordered his men … not to spare the sword against man or woman, boy or girl, and whomsoever they caught they were not to let him go without killing him or hanging him or cutting off his members.118 It appears that Owain’s men went on to enslave a large group of local people whilst carrying out these orders.119 We can only guess how Henry I might have reacted to such behaviour being conducted in his name as Owain’s traditional savagery was to be his undoing. The Brut relates that following Owain’s ravages some of the fugitives managed to escape to the castle at Carmarthen where they found a champion in Owain’s longstanding enemy, Gerald of Windsor. These refugees subsequently complained to Gerald that that they had been “plundered and pillaged” by Owain ap Cadwgan.120 Gerald was probably aware of Owain’s royal mission; however, unlike King Henry he had not forgiven the Welsh prince for the rape of his wife. He, therefore, summoned a large force that surprised Owain’s war band and the Welsh prince was subsequently slain by arrow fire.121 Evidently traditional codes concerning sexual jealousy, power 117  “Yghyfrwg hynny yd anuones henri vrenhin kenhadeu at Ywein vab Cadwgawn y erchi idaw dyuot attaw … A phan doeth y dywat y brenhin wrthaw, ‘Vy gharedicaf i, Twein, a atwaenosti y lleidryn gan Ruffud ap Rhys yssyd megys yn kyuodi yn erbyn vy nhywyssogyon i? A chanys credaf i dy uoti yn gyw[i]raf gwr y mi, mi a uynhaf dy uot ti yn tywyssawc llu gyt a’m mab i y wrthlad Gruffud vab Rys.’ ” Translation modified from Jones, ByT, pp. 96–97. 118  “… hyt not arbettei y cledyf nac y wr nac y wreic nac y vab nac y verch, a phwy bynhac a delhynt nas gollyghynt heb y lad neu y grogi neu try chu y aelodeu.” Ibid., pp. 96–97. 119  The Brut notes that Owain’s men followed the tracks of a multitude of fleeing common folk and seized them before they were able to reach the safety of Carmarthen castle, ibid., pp. 96–99. 120  “… ac yn menegi y ry yspeilaw o Ywein ap Cadwgawn …”, ibid., pp. 98–99. 121  Ibid., pp. 98–99.

772

Wyatt

and honour also continued to simmer beneath the chivalric surface of Owain’s old enemy. Following Owain’s death Gruffudd ap Rhys received the kingdom of Deheubarth from Henry I, yet, this did not entirely restrain his traditional sensibilities. During the final year of his life (1136) he attacked the English forces at Cardigan together with the sons of Gruffudd ap Cynan, king of Gwynedd.122 Once again this native assault was characterised by slave raiding.123 Moreover, Gruffudd’s allies in this conflict, Owain and Cadwaldr ap Cynan of Gwynedd, were to remain heavily involved in the Hiberno-Norse slave supply network and they continued to carry out slave raiding activities during the central years of the twelfth century.124 Nevertheless, the English acculturation of Welsh princes continued at a pace. This is, perhaps, best exemplified by the career of Rhys, the lord of Deheubarth, son of the recalcitrant Gruffudd ap Rhys. Rhys, it seems, had not inherited his father’s traditional sensibilities. He became one of Henry II’s most trusted magnates and acted as the king’s justiciar in South Wales during the final quarter of the twelfth century.125 Rhys adopted the manners and customs of his English counterparts and became an enthusiastic patron of the reformed religious orders in Wales.126 Yet, in one significant respect Rhys did continue to adhere to more traditional norms of behaviour for he maintained a prolific number of concubines (one of whom was his own niece).127 Rhys’ continuing adherence to resource polygyny had predictable consequences and following his death in 1197 his kingdom lapsed into a bitter internecine conflict. The chronicler of the Brut followed these events with some concern. After the death of the Lord Rhys, Gruffudd, his son, succeeded after him in the rule of his territory; whom Maelgwn, his brother, seized when the said Maelgwyn, after having been banished before from his territory, and 122  Ibid., pp. 114–115. 123  Ibid. The chronicler notes that Owain and Cadwaladr obtained “… an exceedingly great number of captives and spoils.” 124  For examples see Wyatt, “Gruffudd”, p. 615. Indeed, it appears that slave raiding continued to be an important activity for the dynasty of Gruffudd ap Cynan. The ByT relates that in 1165 Gruffudd’s grandson, Dafydd ab Owain, ravaged Tegeingl “… ac y mudawd y dynyon a’e haniueileit y gyt ac ef hyt yn Dyfryn Clwyt.” (“… removed the people and their cattle with him to Dyffryn Clwyd.”) ByT, pp. 144–145. See also below. 125  Davies, Coexistence, pp. 222–223. 126  Ibid., p. 221. 127  Ibid., p. 224. Similarly, Cadwgan ap Bleddyn of Powys sired children by at least five separate partners, ibid., p. 71.

Slavery And Cultural Antipathy

773

his men along with him and the war-band of Gwenwynwyn along with them, came to Aberystwyth and gained possession of the town and of the castle and slew many of his people and carried off others into slavery and gained possession of all Ceredigion and its castles.128 It appears that whilst traditional patterns of patriarchal power and illegitimate inheritance were maintained then the ravages of warrior fraternities would continue to be a prevalent feature of Welsh society. As we shall, see this phenomenon was by no means restricted to the Welsh communities.

Acculturation and Antipathy in Twelfth-Century Scotland

During the final quarter of the eleventh century intensive slave raiding activities were conducted by Scottish warrior fraternities against the North of England. These activities were largely undertaken by forces under the control of the Scottish king Malcolm Canmore. The Durham chronicler commented that during Malcolm’s reign he had harried Northumbria five times “… with savage devastation, and carried off the wretched inhabitants as captives, to reduce them to slavery.”129 Malcolm’s raids upon England appear to have been characterised by female abduction and violence directed against Church institutions.130 Furthermore, the court of the Scottish king appears to have been 128  “Gwedy marw yr Arglwyd Rys y dynessaawd Gruffud, y vab, yn y ol yn llywodraeth y kyfoeth; yr hum a delis Maelgwyn, y vrawt, pan doeth y dywededic Vaelgwyn, wedy ry alldudaw kyn [n]o hyny o’e gyuoeth, a’e wyr y gyt ac eff, a theulu Gwenwynwyn y gyt ac wynt, hyt yn Aber Ystwyth a gweresgyn y dref a’r castell a llad llawer o’e bobyl a dwyn ereill yg keithiwet a gweresgyn holl Geredigyawn a’e Chestyll.” ByT, pp. 178–179. 129  “Quinquies namque illam atroci depopulations altrivit, et miseros indigenas in servitutem redigendos abduxit captivos.” HR, SMOO, vol. ii, p. 221, translation from Anderson, SAEC, pp. 112–113. 130  The Durham chronicler notes that in 1070 Malcolm’s forces ravaged Northumbria and burned the Church of St. Peter in Wearmouth and many other Churches besides. Following this attack he is said to have led many “youths and girls” (“juvenes vero et juvenculæ”) away into slavery so that Scotland was filled with the “slaves and handmaids of the English race” (“Repleta est ergo Scotia servis et ancillis Anglici generis.”) HR, SMOO, vol. ii, pp. 190–192, translation from Anderson, SAEC, pp. 91–93. Furthermore, in 1079 Malcolm once again raided Northumbria and his army is said to have assaulted the Priory at Hexham. Again a multitude of men and women were said to have been led away into slavery, HR, SMOO, vol. ii, pp. 36–38. This account is supported by the ASC E, 1079 (Swanton), pp. 213–214 and Ælred of Rievaulx, De Sanctis Ecclesiæ Haugustaldensis, in The Priory of Hexham, its Chroniclers, Endowments and Annals, J. Raine (ed.), Ssoc, 44,

774

Wyatt

a refuge for disruptive English warriors and noblemen who had been expelled from England following the Norman Conquest.131 Malcolm was a willing patron for these disenfranchised individuals and, as such, would seem to have been something of a figurehead for traditional warrior elements within the British Isles. The Durham chronicler certainly seems to have regarded him as such and describes Malcolm to be a man “… to wit of the greatest ferocity and with a bestial deposition …”132 Yet, other twelfth-century English chroniclers take a slightly more positive view of the Scottish king, predominantly because of his marriage to the Anglo-Saxon princess Margaret, the grand-niece of Edward the Confessor.133 Although this marriage had not been in the direct political interest of the English elite it was regarded by many ecclesiastical commentators to have been a fortuitous union. William of Malmesbury felt that Margaret had been a pious and civilising influence upon the Scottish king and his people.134 Even the critical Durham chronicler conceded that “… by her zeal and industry the king himself laid aside his barbarity of manners, and became more honourable and more refined.”135 Margaret’s panegyrist, Turgot, prior of Durham, also praised the queen for her devotion to religion and her adherence to the promotion of ecclesiastical reform.136 During Malcolm’s reign Margaret 1863, pp. 173–203, 177–180, see also p. 287 ns. 44 and 45, above. The Scottish Chronicle of Melrose also notes that Malcolm’s forces ravaged the North of England in these years, however, it does not mention his slave raiding activities, see Anderson, ESSH, vol. ii, pp. 23 and 45. Nevertheless, the English accounts concerning Malcolm’s slave raids are supported by the late medieval Scottish historians John of Fordun and Walter Bower, see CSJV, vol. i, pp. 203–204 and Scotichron., lib. iv, vol. iii, pp. 57, 65. 131  William of Malmesbury notes that Malcolm “… had a warm welcome for all runaways on the English side …” (“… Malcolmus omnes Anglorum perfugas libenter recipiebat …”) GRA, vol. i, pp. 462–463. See also HR, SMOO, vol. ii, pp. 190–192 and ASC D (Swanton), p. 201. These disaffected warriors and nobles included Edgar the Ætheling, Earl Gospatric of Northumbria and a certain Mærleswein, see also Duncan Scotland, pp. 118–119. 132  “… homo scilicet ferocissimus mentemque bestialem gerens …”, HR, SMOO, vol. ii, pp. 36–37. 133  Duncan, Scotland, p. 119. 134  GRA, pp. 726–727. 135  “… cujus studio et industria rex ipse, deposita morum barbarie, factus est honestior atque civilior.” HR, SMOO, vol. ii, p. 192. 136  Turgot had been Margaret’s personal chaplain and he appears to have known her well. His biography/hagiography of Margaret was written down sometime during the first decade of the twelfth century see Vita S. Margaretæ Scotorum Reginæ, Symeonis Dunelmensis Opera et Collectanea, J. H. Hinde (ed.), SSoc, 51, 1868, pp. 234–254, translation in Anderson, ESSH, vol. ii, pp. 59–88. Interestingly, Turgot credited Margaret with freeing many English individuals that had been enslaved by her husband’s warriors, ibid., p. 76. There has been some debate over the authorship of this source for a summary see L. L. Honeycutt, “The

Slavery And Cultural Antipathy

775

appears to have acted as a conduit through which reform values and chivalric sensibilities were transmitted into the upper echelons of Scottish society. One significant consequence of Margaret’s English connections was that her sons were raised and educated at the English court. As we have already seen this upbringing appears to have imbued these Scottish princes with the chivalric and reform-minded sensibilities of the English elite. However, such cultural intrusions at the highest level of Scottish society also created significant internal tensions and, as in Wales, they encouraged increasing levels of cultural antipathy towards this English infiltration. The vehement nature of this cultural antipathy was to become apparent in the years following Malcolm Canmore’s death. Malcolm had six sons by Margaret and at least two by his other wife Ingibiorg, daughter of the Earl Thorfinn the mighty of Orkney.137 Upon Malcolm’s death it must have seemed that a violent succession dispute was inevitable. However, it appears that the education of Malcolm’s sons at the English court had tempered their tendencies towards such fraternal competition. Indeed, it was Malcolm’s more traditionally minded brother, Donald Bán, who seized the throne of Scotland.138 Donald appears to have acted as a figurehead of a movement that wished to resist the increasing influence being exerted by the English court upon the Scottish royal circle. This is evident from the accounts given by the AngloSaxon Chronicler who comments that following Malcom’s death “… the Scots elected Donald, brother of Malcolm, king, and drove out all the English who had been with King Malcolm before.”139 The seizure of the Scottish throne by a traditional warrior leader was clearly not welcomed at the English court. The English king William Rufus subsequently supported Duncan, Malcolm’s eldest son by Ingibiorg of Orkney, in a bid to regain his patrimony and oust Donald Idea of the Perfect Princess: The Life of St. Margaret in the reign of Matilda II”, ANS, xii, 1989, pp. 81–98. 137  The fate of Ingibiorg is unknown; however, Duncan feels that Malcolm did not put her away. Nevertheless, he suggests that Ingibiorg must have died before Malcolm married Margaret. Similarly, Barrow mentions Malcolm’s marriage to Ingibiorg but does not discuss her fate, Kingship and Unity, p. 27. Yet, given what we know about the significance of polygyny for the traditional ‘Celtic’ elite it seems perfectly feasible that Malcolm maintained both wives simultaneously, see Duncan, Scotland, pp. 118–119. 138  Donald’s succession, thereby, adhered to traditional Scottish custom, which dictated that able bodied adult males within one generation of the king (i.e. brothers or cousins) should succeed to the throne rather than sons or nephews, see Barrow, Kingship and Unity, pp. 24, 31. 139  “Da Scottas pa Dufenal to cynge gecuron Melcolmes broðer. 7 ealle þa Englisce út adræfon.” TSCP, p. 228 translation from Swanton, ASC E, p. 228.

776

Wyatt

from power.140 Duncan subsequently marched north with an army of “English and French” (“Engliscra 7 Frenciscra”) supporters and succeeded in removing Donald from the throne of Scotland.141 However, Duncan’s newfound hegemony was to be extremely short-lived; his ascendance to the Scottish throne soon resulted in dramatic traditional backlash against his importation of English and continental warriors into the Scottish fold. The author of the Anglo-Saxon Chronicle once again comments upon the events noting that Donald was expelled by Duncan who was received as king. But afterwards some of the Scots gathered themselves together, and slew almost all his (Duncan’s) followers; and he himself escaped with a few Thereafter they were reconciled, on the condition that he should never again introduce English or French into the land.142 Yet, it appears to have been extremely difficult for Duncan to allay the concerns of Scottish warriors regarding his chivalric affiliations and sensibilities. In spite of an apparent reconciliation with the more traditional Gaelic factions in his kingdom he was slain within a year by the mormaer of Mearns, one of Donald Bán’s supporters.143 Furthermore, William of Malmesbury remarks that Duncan’s half brother, Edmund, had assisted in this killing. It, therefore, appears that not all of Margaret’s sons had adopted the mores of their English hosts.144 Following Duncan’s assassination Donald once again seized the Scottish throne. As a result, William Rufus sponsored a further expedition which was to end in Donald’s subsequent capture and the removal of his eyes.145 This expedition was headed by Edgar, Malcolm’s eldest son by Margaret. Edgar seized the reins of power following Donald’s departure but he appears to have pragmatically resisted the temptation to introduce a wave of southern colonists into Scotland. Nevertheless, his ascent to the throne was to draw the Scottish elite further into the cultural orbit of their southern 140  Ibid. Duncan had been at the English court since 1072 when he was given as a hostage by Malcolm to William the Conqueror. 141  Ibid. 142  “Ac þa Scottas hi eft sume ge gaderoden. 7 for neah | ealle his mænu ofslogan. 7 he sylf mid feawum æt bærst. Syððan hi wurdon sehte, on pa gerád þ he nafre eft Englisce ne Frencisce into þam lande ne ge logige.” Ibid. 143  Ibid., p. 230, these events are confirmed by The Chronicle of Melrose, see Anderson, ESSH, vol. ii, p. 89, see also Duncan, Scotland, p. 125. 144  GRA, pp. 724–727. 145  Perhaps, fittingly, Donald was the final king to be buried at the traditional resting place of Scottish-Gaelic royalty at the monastery of Iona. See Barrow, Kingship and Unity, p. 31.

Slavery And Cultural Antipathy

777

neighbour. Following his death in 1107, Edgar was succeeded by his younger brothers Alexander I (1107–1124) and then David I (1124–1153). Both men had been raised at the English court and tutored in the values of continental warfare and reform. Both encouraged their followers to adopt revised warrior mores and both established and patronised continental-style monastic institutions in Scotland.146 As we have seen, David I also endowed many of English acquaintances with lands in the south of his kingdom.147 He also became a highly praised champion for reform within the Scottish Church.148 Yet, such cultural intrusions at the highest level undoubtedly created tensions amongst traditional Gaelic elements within his kingdom. David appears to have recognised that he required the support of these traditional factions and he pragmatically indulged and tolerated their unchivalrous mores.149 His tolerance in this respect has resulted in the slightly disconcerted and ambivalent portrayal of his career extant within the twelfth-century English chronicles.150 These twelfth-century accounts highlight the tensions and difficulties created by the ‘top down’ infiltration of reforming ideals into the warrior communities of Gaelic Scotland. Indeed, such tensions are readily apparent when we examine the events that surround the battle of the Standard of 1138.

A Clash of Cultures? The Battle of the Standard, 1138

The Battle of the Standard was triggered by events in England and by David’s close connection with the English court. In 1136 civil war broke out in England as a result of the succession dispute between Stephen of Blois and the Empress Matilda. David who had been a vassal of Matilda’s father, Henry I, appears to have used this dispute as an opportunity to press his territorial claims upon the north of England.151 In 1136 his forces invaded Northumbria and seized Carlisle 146  Barrow, Kingship and Unity, p. 81, Davies, Domination, 17 and Duncan, Scotland, pp. 131– 132, 144–148. 147  Duncan, Scotland, pp. 135–7, 139–41. 148  For examples see SMOO, vol. ii, pp. 330–331, see also above. 149  See below. 150  Barrow, David I of Scotland (1124–1153) The Balance of the Old and the New, pp. 6–7. 151  See G. W. S. Barrow, “The Scots and the North of England” in E. King (ed.), The Anarchy of King Stephen’s Reign (Oxford; New York: Clarendon Press, 1994), pp. 231–253, 245–246. See also K. J. Stringer, “State Building in Twelfth Century Britain: David I, King of Scots and Northern England.” in J. C. Appleby and P. Dalton (eds.), Government, Religion and Society in Northern England 1000–1700 (Stroud: Sutton, 1997), pp. 40–62 and W. Croft-Dickinson, Scotland from the Earliest Times to 1603 (Oxford: Clarendon Press, 1977), p. 66.

778

Wyatt

and Newcastle.152 The Scottish king’s warriors periodically ravaged the northern province for the following two years in spite of a peace settlement that David had reached with King Stephen in 1136. David appears to have deployed these raids as a means to consolidate his authority and to redirect the cultural antipathy of his more traditional warriors away from his own chivalric contingent. By focussing the attentions of traditional factions such as the Galwegians upon an external conflict with their old enemy, David may have hoped to temporarily alleviate the tensions that were simmering within his own kingdom.153 Stephen’s panegyrist makes it clear that David deliberately granted his more traditional followers a licence to conduct pillaging and slave raiding activities.154 So King David … sent out a decree through Scotland and summoned all to arms, and giving them free license he commanded them to commit against the English, without pity, the most savage and cruel deeds they could invent.155

152  See, Stringer, “State building”, pp. 40–50 and Barrow, “Scots and the North”, pp. 245–247. 153  The fragile nature of David’s political hegemony over the more traditional Gaelic regions is highlighted by R. Andrew Macdonald who has suggested that David had minimal control over the warriors of Galloway who looked to their own traditional leaders, Ulgric, Donald and Fergus in this period. See, “Rebels without a cause? The relations of Fergus of Galloway and Somerled of Argyll with the Scottish Kings, 1153–1164” in E. J. Cowan and R. Andrew MacDonald (eds.), Alba, Celtic Scotland in the Middle Ages (East Linton: Tuckwell Press, 2000), pp. 166–186, 171. 154  In early medieval Irish and Scottish society kings had punctuated their reigns with crech ríg or ‘royal prey’. These were plundering expeditions conducted against neighbouring territories that were intended to assert leadership, promote group solidarity and provide rewards for followers. Warriors, thereby, sold their loyalty for the symbols of prestige and power attainable by plunder. The acquisition of slaves and booty were the primary motivation for such raids. Malcolm I was said to have made such a raid on England in the second quarter of the tenth century during which his army seized a multitude of people and cattle, see Chronicle of the Kings of Scotland, Chronicles of the Picts, Chronicles of the Scots, W. E Skene (ed.) (Edinburgh: H. M. General Register House, 1867), p. 10, see also Anderson, ESSH, vol. i, p. 452. The activities of David’s army in 1138 should be regarded within this context. In addition, Stringer has argued that David may have deployed his more traditional factions as a means to instil terror into the population of Northern England over whom he was attempting to secure dominion. See Stringer, “State Building in Twelfth Century Britain”, p. 44, Barrow, “Scots and the North”, pp. 245–247, Duncan, Scotland, p. 114 and Ó Croínín, Medieval Ireland, p. 74. 155  “Rex igitur David, sic enim uocabatur, perpenso per Scotiam edicto omnes in arma commouit, laxatisque permissionis suæ frenis, quicquid in Anglos truculentius, quicquid excogitare possent inhumanius, remota pietate peragere præcepit.” GS, pp. 54–55.

Slavery And Cultural Antipathy

779

Similarly, Henry, archdeacon of Huntingdon, remarked with horror that … the king of Scots, under cover of piety, on account of the oath he had sworn to King Henry’s daughter, commanded his men in barbarous deeds. For they ripped open pregnant women, and tore out the unborn foetuses. They tossed children on the points of their lances. They dismembered priests on their altars. They put on to the bodies of the slain the heads cut off crucifixes, and changing them round, they put back on the crucifixes the heads of the dead. Everywhere that the Scots attacked would be filled with horror and barbarity, accompanied by the cries of women, the wailing of the aged, the groans of the dying, the despair of the living.156 Hysterical accounts such as Henry’s would appear to be both sensationalist and propagandist in nature. Yet, we must ask ourselves who were these twelfth-century authors intending to shock? When we examine the various accounts that bitterly condemned the savage behaviour of the Scottish warriors prior to the battle of the Standard, then we see that they had two things in common. First, they all constitute a clear expression of the newly constructed ‘superior’ identity discussed earlier.157 Secondly, they almost all emanate from writers within the milieu of the ecclesiastical reform movement. Ælred of Rievaulx (1109–1167) was a Cistercian monk, based at Hexham, who had spent time at the court of King David I. Richard of Hexham ( fl. 1141) was prior at the Augustinian house there; John of Hexham ( fl. 1180) was his successor. John of Worcester was a monk writing during the first half of the twelfth century at the monastery reformed by Bishop Wulfstan of Worcester. A monk residing at St. Cuthbert’s in Durham composed the Durham chronicle, attributed to Symeon of Durham, during the first half of the twelfth century.158 Robert, Bishop of Bath, most probably compiled the Gesta Stephani during the central years of the twelfth century. It should be noted that several of these writers hailed from houses that had been directly affected by the ravages of Scottish warriors. Furthermore, an acknowledgement of the reforming sensibilities 156  “Rex namque Scotorum, quia sacramentum fecerat filie regis Henrici, per suos execrabiliter egit. Mulieres enim grauidas findebant, et fetus anticipatos abstrahebant. Pueros super acumina lancearum iactabant. Presbiteros super altaria detruncabant. Crucifixorum capita abscisa super cesorum corpora ponebant. Mortuorum uero capita mutuantes super crucifixa reponebant. Quecumque igitur Scoti attingebant, omnia erant plena honoris, plena immanitatis, aderat clamor mulierum, eiulatus senum, morientium gemitus, uiuentium desperatio.” HA, pp. 710–711. 157  See Gillingham, “Conquering”, pp. 66–84. 158  See D. Rollason (ed.), Symeon, Historian of Durham and the North (Stamford: Shaun Tyas, 1998) for the debates over the authorship of this chronicle.

780

Wyatt

of these writers is extremely important for understanding their vehement antipathy towards the activities of the Scottish forces. Indeed, their attitude towards the Scots must be regarded within the context of wider reform concerns about the conduct of the warrior elite in general. These concerns are perhaps best exemplified by the complaints of John of Hexham concerning the conduct of the David’s traditional warrior contingent. … They slew all the men, and bound together with cords the maidens and widows, naked, in troops, and drove them away into Scotland under the yoke of slavery … The Scots also broke into the sanctuaries of the Lord and in the consecrated places irreverently committed acts violent, lewd and execrable.159 John of Hexham’s concerns regarding the activities of the Scottish warriors were fourfold. Firstly, he complained that these warriors had carried out ‘uncivilised’ slave raiding activities of an overtly sexual nature.160 Secondly, and as a corollary of this, John implied that these warriors were debauched and lacking in moral fibre. Thirdly, he suggests that they were unrestrained, lawless and indiscriminately violent. Indeed, many of the English accounts regarding these warriors imply that they lacked centralised control and note their failure to adhere to revised codes of warrior behaviour.161 Finally, John clearly felt that these warriors were imbued with an almost pagan like ethos that motivated them to violate the Truce of God by targeting religious institutions and personnel.162 It has been argued that the reaction of English chroniclers, like John of Hexham, towards the Scottish activities in 1138 can be explained by the Scottish warriors’ employment of ‘total war’ tactics in order to accumulate slaves.163 The alleged annihilation of infants and the aged by the Scottish forces can be seen to have strong anthropological analogies with the activities of slave raiding

159  “Non pepercit seui, ætati, aut ordini, conditioni cujusquam vel professioni. Prægnantes cum parvulis dissecuerunt, virgines et viduas, occisis cunctis maribus, funiculis colligatas nudas catervatim sub jugo servitutis in Scotiam abegerunt … sanctuaria quoque Domini Scotti confregerunt, et in sacratis locis violenta, obscena, et abhominanda irreverenter perpetraverunt.” SMOO, vol. ii, p. 290, translation from Anderson, SAEC, pp. 181–182. 160  See also note 130, above. 161  See chapter 2. 162  See also notes 130 and 159, above. 163  Gillingham, “Conquering”, pp. 70–73 and Strickland, “Slaughter”, pp. 41–61.

Slavery And Cultural Antipathy

781

groups in other societies.164 Yet, this does not explain the English chroniclers’ hysterical accounts regarding the anti-ecclesiastical activities conducted by supposedly Christianised Scottish warriors. It was certainly the case that these writers were attempting to dehumanise and ‘de-Christianize’ their Scottish foes by portraying them as sub-human heathens. Nevertheless, the proliferation of these accounts does suggest that there may well have been an element of truth in them. Indeed, as in Welsh society, the reform norms and revised codes of behaviour that were infiltrating into the circle of the Scottish elite would have posed a direct threat to traditional warrior practices and expressions of power and virility. The slave raiding activities and anti-ecclesiastical attacks that characterised the Scottish raids upon England may well have constituted a deliberate and symbolic reaction towards this cultural infiltration.165 Indeed, as the English Church assumed an increasingly hard line against the overtly violent and sexual behaviour of the warrior elite, it seems that traditional elements within Scottish society expressed their antipathy towards these revised values and norms. They did so by reasserting their affiliation to traditional warrior practices and expressions of identity and power: through slave taking, female abduction and excessive violence. The forces of cultural antipathy were also creating significant internal tensions within the Scottish communities. These tensions are highlighted by the English chroniclers’ accounts of dissent and conflict within the Scottish ranks prior to the battle of the Standard. The traditional Galwegian contingents in David’s forces were at the very heart of this internal discord. Indeed, they appear to have had little, if any, respect for his royal authority.166 These were the same warriors that had been consistently singled out by the English chroniclers as being heavily involved in slave raiding and anti-clerical activities. Moreover, the traditional sensibilities of these warriors seem to have clashed dramatically with the chivalric values and norms of behaviour of the knights of King David’s retinue. Ælred of Rievaulx remarked that during David’s excursion into Northumbria in 1136 that his ‘civilised’ knights had attempted to

164  Patterson, Social Death, pp. 120–121. The twelfth-century Icelandic source Landnámabók notes that similar tactics had been employed by Norse warriors, Landnámabók, vol. ii, pp. 379–380. 165  It is acknowledged that Scottish concerns regarding Canterbury’s highly politicized claims to primacy over the Scottish Church also may have provided a significant contributory edge to this antipathy, see M. Brett, “Canterbury’s Perspective on Church Reform and Ireland, 1070–1115” in Bracken and Ó Riain-Raedel, Ireland and Europe, pp. 13–35, 19. 166  See above.

782

Wyatt

prevent the enslavement of English individuals at the hands of the more barbarous Galwegian forces. Ælred relates that Hence it was that, when the most cruel nation of the Galwegians raged with unheard brutality, and spared not sex nor age, our countrymen (nostrates) who were with the king (David) were moved by compassion and sent many rescued from their hands, to Hexham, as a sure defence of their safety.167 Such apparently noble behaviour may well have been regarded as a sign of cultural affiliation and indeed intrusion by David’s more traditional warrior contingent. The Scottish king appears to have found it extremely difficult to control the antipathy of these traditional factions towards his own chivalric retinue. This is best exemplified by Ælred of Rievaulx’s account of events within the Scottish camp prior to the battle of the Standard. Ælred relates that an argument broke out between David’s knights and the Galwegian leaders concerning who would lead the initial attack against the English. He felt that this dispute had become extremely heated and as a result King David had been forced to intervene and restrain both factions “… lest a disturbance should suddenly arise out of this altercation.”168 Ælred also recounts that one of David’s knights, Robert de Bruce, attempted to dissuade the king from fighting alongside these traditional warriors and against his English brethren.169 The words that the Ælred places into Robert’s mouth are very illuminating.

167  “Hinc est quod cum impiissima gens Galwensium inaudita crudelitate sævirent, nec sexui parcerent, nec ætati; nostrates, qui erant cum rege, pietate commoti, plures de eorum ereptos manibus ad Hagustaldunum quasi ad certum suæ salutis auxilium transposuerunt.” De Sanctis Ecclesiæ Haugustaldensis, p. 183 translation from Anderson, SAEC, p. 171. Similarly, John of Hexham noted that David I emancipated some of the maidens and widows that had been enslaved by his forces and restored them into the care of Robert, prior of Hexham in 1138, see SMOO, vol. ii, p. 290. 168  “Tunc rex utrosque compescens, ne tumultus hac altercatione subito nasceretur, Galwensium cessit voluntati.” Ælred noted with concern that David “(subsequently) yielded to the will of the Galwegians.” Relatio de Standardo, CRSHR, vol. iii, pp. 190–191, translation from Anderson, SAEC, p. 199. 169  David I had granted the lordship of Annandale (a district of some 200,000 acres adjacent to the English border) to Robert of Brus. Moreover, Robert was an ancestor of the later Scottish royal line, see Barrow, Kingship and Unity, p. 32.

Slavery And Cultural Antipathy

783

Against who do you bear arms today and lead this huge army? Against the English, truly and the Normans. O king, are not these they with whom you have ever found useful counsel and ready help, and willing obedience besides? … New to you is this confidence in Galwegians, attacking with arms today those by whose aid hitherto you have ruled the Scots with affection, the Galwegians with terror.170 Interestingly, Ælred suggests that when David refused to step back from the brink of conflict Robert de Bruce broke his oath of fealty to the king and deserted his forces.171 Evidently, Robert preferred to invalidate his most important social relationship rather than be associated with the ‘uncivilised’ warriors of David’s army. His actions in this instance may be juxtaposed neatly with those of the Welsh ynvydyon who preferred to side with more traditional forces (with whom they were normally bitter enemies) and break their oaths of fealty to their ‘civilised’ and ‘chivalric’ Welsh lords.172 The battle of the Standard is portrayed within the English chronicles using the language of holy war. Thurstan the reform-minded Archbishop of York personally assembled and commanded the English army.173 Prior to the battle it 170  “ ‘Adversum quos hodie levas arma, et imensium hunc ducis exercitum? Adversum Anglos certe et Normannos. O rex, nonne isti sunt quorum semper et utile consilium et auxilium promptum, gratum insuper obsequium, invenisti? … Nova tibi est in Galwensibus ista securitas, qui eos hodie armis petis per quos hactenus amabilis Scottis, terribilis Galwensibus, imperasti.’ ” Ibid., pp. 192–193, translation modified from Anderson, SAEC, pp. 192–193. Thomas comments that “this speech was a tricky piece of rhetoric, trying to persuade the Scottish king that his interests lay with the English and Normans, not with the Scots and Galwegians, but the distinction between the two opposed pairs of peoples is strongly emphasized in it.” English and Normans, p. 313. 171  Ibid., p. 195. Robert’s rejection of fealty to David is also noted by Richard of Hexham, see DGRS, CRSHR, vol. iii, pp. 161–162. Robert was in a difficult position because he held lands both north and south of the English border. The conflict at the Standard, therefore, presented him with a dilemma of conflicting interests. Nevertheless, iElred clearly felt that it was David’s association with traditional Gaelic factions that had prompted Robert into choosing sides. 172  See above. 173  The Vita Thurstani relates that the Archbishop actually took part in the battle and claims that he and his knights, slew several thousand Scots, “Vita Thurstani Archiepiscopi” in Historians of the Church of York and its Archbishops, J. Raine (ed.) (London: Longman, 1886), vol. ii, p. 266. Henry of Huntingdon, however, feels that the Archbishop merely coordinated the English forces and sent Ralph, bishop of the Orkneys to the battlefield as his representative, HA, p. 713. See also D. Nicholl, Thurstan Archbishop of York (1114–1140)

784

Wyatt

appears that Thurstan gave the English troops papal sanction for their enterprise and absolved them of their sins.174 Moreover, the chroniclers’ description of the English army is characterised by crusading language. The English forces were portrayed as model warriors of the reform movement, as the milites dei (soldiers of God). This is best exemplified by the account of Ælred of Rievaulx who described these troops in the following manner: Shield was joined to shield, side pressed to side, lances were rained with pennons unfurled, hauberks glittered in the brilliance of the sun; priests white clad in their sacred robes, went round the army with crosses and relics of the saints, and most becomingly fortified the people with speech as well as prayer.175 The Scottish forces on the other hand, were depicted in terms akin to those that might be applied to a pagan horde. Richard of Hexham described them like this: … that execrable army, savager than any race of heathen, yielding honour to neither God nor man … In front of the battle were the Picts (i.e. Galwegians), in the middle the king (David) with his knights and his English; the rest of the barbarians extended round them on all sides, roaring.176 (York: Stonegate Press, 1964), pp. 227–228. Thurstan certainly had a vested interest in the conflict as York had claimed primacy over the Scottish Church and had secured papal backing over this issue. See Hugh the Chanter, The History of the Church of York 1066– 1127, C. Johnson (ed. and trans.) (Oxford: Clarendon Press, 1990), pp. xlv–liv and Duncan, Scotland, pp. 131, 219. 174  “…  ac deinde absolutionem et benedictionem Dei et suam eis sollempniter tribuit … Tunc crucem suam et sancti Petri vexillum, ac suos homines eis tradidit.” “… then he (Archbishop Thurstan) solemnly bestowed upon them absolution, and God’s blessing and his own … Then he gave them his cross, and the banner of St. Peter (i.e., the banner of St. Peter’s, York) and his vassals.” DGRS, CRSHR, vol. iii, pp. 161–162. See also Relatio de Standardo, p. 182. 175  “Scutis scuta junguntur, lateribuus latera conseruntur, laxatis vexillis eriguntur lanceæ, ad solis splendorem loricæ candescunt; sacerdotes sacris vestibus candidat, cum crucibus et reliquiis Sanctorum, exercitum ambiebant, et sermone simul et oratione populum decentissime roborabant.” Relatio de Standardo, pp. 191–192, translation from Anderson, SAEC, pp. 201–202. 176  “Igitur ille detestandus exercitus, omni paganorum genere atrocior, nec Deo nec hominibus reverentiam deferens … In fronte belli erant Picti, in medio rex cum militibus et Anglis suis; cetera barbaries undique circumfusa fremebat.” DGRS, CRSHR, vol. iii, pp. 151–152, 163,

Slavery And Cultural Antipathy

785

Accounts such as these have prompted Gillingham into arguing that contemporary writers viewed the Standard … not just as a battle between Scots and English but as a titanic and ferocious struggle between two opposing cultures, the civilised and the savage.177 Gillingham somewhat undermines this perceptive comment by moving on to portray the battle as, essentially, a struggle of ‘nascent nationalism’ between the English and the ‘Celtic’ Scots.178 Yet, the words which Ælred of Rievaulx placed into the mouth of Robert de Bruce reveals that he, at least, understood the ethnic-political situation to be far more complex than this. Indeed, when we examine the sources closely then we see that a myriad of contingent forces made up the Scottish host at the Standard. Although the Scottish army was led by David’s chivalric contingent it seems clear that his troops were very much in a minority. Contemporary writers reveal that the Scottish forces were made up of a conglomerate of traditional warrior groups from many different backgrounds which reflected the hybrid entity of the Scottish kingdom at this time. For example, Richard of Hexham noted that David’s “impious army” (“nefandus exercitus”) was composed of Normans (probably David’s knights), Germans (presumably Norseman from the Isles), English, of Northumbrians and Cumbrians (traditional Saxon warriors from the north), of (men of) Teviotdale and Lothian, of Picts (who are commonly called Galwegians) and of Scots (traditional Gaelic warrior elements).179 Ælred of Rievaulx, similarly, commented that the Scottish king’s army contained men from Cumbria, Teviotdale, Galloway, Lothian, Scotland, England and men translation from Anderson, SAEC, pp. 180, 202. The noise made by traditional warriors in the Scottish army might constitute an example of the war howl employed by fían warriors and discussed in chapter 2. 177  Gillingham, “Conquering”, p. 74. 178  See Gillingham, “Conquering”, pp. 74–75 and “Henry of Huntingdon and the Twelfth Century Revival of the English Nation” in S. Forde et al. (eds.), Concepts of National Identity in the Middle Ages (Leeds: University of Leeds, School of English, 1995), pp. 75–101, 79–80. 179  “Normannis, Germanis, Anglis, de Northymbranis et Cumbri, de Teswetadala et Lodonea, de Pictis, qui vulgo Galleweienses dicuntur, et Scottis.” DGRS, CRSHR, vol. iii, p. 152.

786

Wyatt

(probably of Scandinavian extraction) from the Western Isles.180 Furthermore, it appears that unlike Robert de Bruce certain other ‘chivalric’ warriors had decided to ‘go native’ and adopt more traditionally minded allies and modes of behaviour.181 Accounts reveal that there was also a significant contingent of traditional Northumbrian warriors in David’s army. These included Edgar, the illegitimate son of earl Gospatric of Northumbria, and Robert and Utred, the sons of Maldred.182 Richard of Hexham described the activities of these traditional English warriors in some detail remarking that At that time certain pestilent men, whose whole aim and joy was to plan and perpetrate crimes, combined together in detestable union, the more effectually to attain the desires of their malevolence … Urged by greed, therefore, and encouraged by impunity, and spurred by madness, they raided through Northumbria like wolves seeking a prey to devour.183 Richard’s account of this ‘unholy’ alliance of Anglo-Saxon warriors in a raiding frenzy clearly bears a striking resemblance to the accounts of the Welsh ynfydyon extant in the Brut y Tywysogyon.184 Therefore, if one were to remove 180   Relatio de Standardo, pp. 189–191. Similarly John of Worcester notes that David’s army was made up of “enemies of many different regions” (“pluribus diuerse gentis hostibus fere” McGurk translates this as “many hostile peoples”), CJW, vol. iii, pp. 236–237. The author of the Gesta Stephani remarks that David had assembled from his own people “and from the nearer parts of Scotland … a mass of rebels into an incredible army.” (“Huius igitur gentis rebellem multitudinem a citerioribus Scotiæ partibus in exercitum inopinabilem adunans …”) GS, pp. 54–55. 181  For example, the Galwegians who committed atrocities at Clitheroe were said to be under the command of King David’s nephew, William Fitz Duncan. William was the son of David’s half brother Duncan II. John of Hexham directly implicates fitz Duncan in the Galwegian slave raiding activities, see SMOO, p. 291. See also, Duncan, Scotland, p. 219. 182  Robert and Utred would appear to have been descendants of the wealthy Anglo-Saxon Northumbrian nobleman, Maldred (son of Crinan and husband of Uchtred’s daughter). Maldred was also the father of earl Gospatric of Northumbria and so these three individuals were probably related. Moreover, Edgar’s illegitimate status made him a likely candidate for membership in a traditional warrior fraternity. For the debates over Maldred’s possible connections to the Scottish royal family see Aird, St. Cuthbert, p. 69, note 41. 183  “Ea tempestate quidam pestilentes, quitus omne Studium et gaudium erat scelera excogitare ac perpetrare, ut malignitatis suæ desideria efficacius consummarent detestabili concordia in unum convenerunt … Igitur rapacitate stimulante, et impunitate patrocinante, ac furore exagitante, more luporum quæritantes prædam quam devorarent, per Northymbriam discurrerunt.” DGRS, CRSHR, vol. iii, pp. 166–167, translation from Anderson, SAEC, p. 209. 184  See chapter 2.

Slavery And Cultural Antipathy

787

David’s chivalric contingent from the Scottish battle line (which Ælred clearly wished David had done) then it is perhaps possible to view the battle of the Standard “as a ferocious struggle between two opposing cultures.” As such it might be regarded as an array of forces from northern and western Britain that adhered, by and large, to the cultural values of traditional warrior society and an opposing army from the south and the continent that represented an alternate set of norms and codes of warfare.185 The chroniclers’ descriptions of the brave ‘English’ and ‘Norman’ warriors who faced this formidable army unequivocally depict the battle of the Standard in this manner.186 Some sense of nascent ‘nationhood’ and economic superiority certainly contributed to their construction of this prestigious ‘other’ identity for the English based forces. Yet, its key components were a sense of chivalric unity and shared religious identity. The crusading language employed to describe the English army clearly reveals that this pious self-image was underpinned by the ideology of reform. This ideology had moulded a revised set of the behavioural norms and warrior values that were in direct opposition to more traditional sensibilities and expressions of power. The reason that the battle of the Standard attracted so much attention from contemporary ecclesiastical commentators was because it was regarded as an integral part of their wider struggle to disseminate these same values throughout the recalcitrant warrior-centred communities of the British Isles. Indeed, the reforming papacy showed a keen intense interest in the outcome of this battle, which is indicated by the prompt arrival on the scene of the papal legate, Alberic, bishop of Ostia. It was this continental bishop, rather than King Stephen’s nobles,

185  Henry of Huntingdon certainly felt that this had been a conflict between the forces of reform and an army of more traditional values. He comments that prior to the battle the English forces were absolved by their priests and then cried out “‘Amen! Amen!.’” Meanwhile the Scottish forces issued forth the “warcry of their fathers” (“Exclamauitque simul; exercitus Scotorum insigne patrium”), HA, pp. 716–717. Barrow has noted that Norse and Gaelic factions within Scotland were forming powerful alliances during this period. He argues that in the north and west there was “… an intensification of the old links with Ireland, a process contrasting sharply with what was happening in the east and lowland Scotland.” See Barrow, Kingship and Unity, p. 107, this view is supported by Andrew MacDonald, “Rebels”, pp. 166–186, 170. See also above. 186  See Davies, English Empire, pp. 127–128 and Gillingham, “Henry and the English”, pp. 79– 80, “Conquering”, pp. 73–75. For a more modified view see Dauvit Broun, “Anglo-French Acculturation and the Irish Element in Scottish Identity” in B. Smith (ed.), Britain and Ireland 900–1300, Insular Responses to Medieval European Change (Cambridge: Cambridge University Press, 1999), pp. 135–153, 152, 153.

788

Wyatt

who dictated the ensuing peace treaty with David I.187 The detail of this peace agreement (related by Richard of Hexham) are extremely revealing. Now (the Scots) had long differed from the cisalpine, indeed from almost the universal church … But at this time inspired by divine grace they all with one accord received pope Innocent’s commands and his legate (Alberic) with great honour … (Alberic) also summoned king (David) for the re-establishment of peace between him and the king of England; and for the sake of this fell at his feet, entreating him to take pity upon holy church and upon himself and his subjects, to whom he (David) had caused so many and so great evils … This also he obtained of the Galwegians, that they should bring back to Carlisle before the same time limit all captive girls and women whom they might have and restore them to liberty there. (The Galwegians) also, and all the others, promised him most faithfully that they would by no means violate churches thenceforth; and that they would spare children and woman-kind, and (men) who were disabled by weakness or age; and that they would thenceforth slay no one unless he opposed them.188 We may assume that this is a relatively accurate account of the treaty of Durham because Robert Biset, prior of Hexham, was present at the peace council.189 It is particularly interesting that the treaty stressed the significance of the need to return those English women who had been enslaved. This would suggest 187  See DGRS, CRSHR, vol. iii, pp. 169–171 and SMOO, vol. ii, p. 298. See also Duncan, Scotland, pp. 220, 260–261. 188  “Illi vero diu a Cisalpina, immo fere ab universa ecclesia discordantes … Tuc vero divina gratia inspirati, mandata Innocentii papæ et legatum ejus omnes unanimiter cum magna veneratione susceperunt … Convenit quoque regem de reformanda pace inter eum et regem Angliœ, et hujus rei gratia ad ejus pedes cecidit, scilicet ut sanctæ ecclesiæ et sui ipsius et suorum misereretur, quitus tot et tanta mala fecerat … Hoc etiam apud Pictos impetravit, quod omnes puellas ac mulieres captivas, quas habere possent, ante eundem terminum ad Cartel reducerent, et eas ibi libertati redderent. Ipsi quoque et omnes alii firmissime promiserunt, quod nullo modo ecclesias amplius violarent; et quod parvulis et fœmineo sexui et ex infirmitate et ætate debilibus parcerent; et omnino neminem nisi sibi resistentem amplius occiderent.” DGRS, CRSHR, vol. iii, pp. 170–171, translation from Anderson (my italics), SAEC, pp. 211–212. Given Richard of Hexham’s account of Scottish warriors enslaving English women cited earlier (supported by John of Hexham and Henry of Huntingdon) it seems safe to assume that these women were being liberated from slavery. 189  Robert Biset was Richard’s predecessor at the priory of Hexham. The author was, therefore, both a friend and colleague of this first hand witness.

Slavery And Cultural Antipathy

789

that the traditional forces within David’s army had been partially successful in their attempt to dishonour their chivalric counterparts through the practice of female abduction. Just as the fictional Ulster warriors of the Irish Táin Bó Cualnge emphasised the recovery of their strength through the liberation of their women, so too, the honour and prowess of the English elite could only be regained when the women under their guardianship had been returned to freedom.190 The emancipation of these women might also have been regarded as a tactic that would help to limit the sexual sins committed by their captors. Furthermore, the underlying motivations behind the treaty negotiated by Alberic are clear. Both the papacy and the English monarchy wished to impose chivalric and reforming ideals onto the traditional elements within the British societies.191 Whether this could ever be achieved by force alone appears unlikely. The cultural antipathy between the English elite and traditional warriors based in Scotland, Wales, Ireland and the Isles was to continue for much of the twelfth century. William of Newburgh described the actions of William the Lion’s Scottish army of 1174 in a, by now, familiar manner. … and while they sought out their prey it was the delight of that inhuman nation, more savage than wild beasts, to cut the throats of old men, to slaughter little children, to disembowel women and to commit other atrocities of a kind too horrible to mention. So, while this army of monstrous bandits was let loose upon this wretched province and the barbarians were revelling in their inhumanity, the king of Scots, himself attended and guarded a more honourable and civilised body of soldiery (honestiori mitiorique stipatus militia) …192 190  See chapter 2. 191  In this respect the treaty may have been partially successful, indeed Aird has commented that Fergus Lord of Galloway responded to the defeat at the Standard by recognising “the value of imitating the socio-cultural values of the opposition, values which had already taken root in the heartlands of the Scots monarchy among the gens maritime.” See W. Aird, “‘Sweet Civility and Barbarous Rudeness”: A View from the Frontier. Abbot Ailred of Rievaulx and the Scots” in Steven G. Ellis and Lud’a Klusáková (eds.), Imagining Frontiers, Contesting Identities (Pisa: Edizioni Plus, Pisa University Press, 2007) pp. 59–75. For an interesting discussion of pragmatic the adoption of reform sensibilities by more traditional ‘Celtic’ leaders in the late twelfth century see Smith, “Frontiers”, p. 245. 192  “… et dum prædæ insisterent, jugulare senes, trucidare parvulos, eviscerare feminas, et hujusmodi, quæ horrendum est et dicere, genti inhumanæ et feris plus efferæ voluptas fuit. Immisso exercitu, barbarisque inhumane debacchantibus, rex ipse, excubante circa se honestiori mitiorique stipatus militia …” Historia Rerum Anglicarum, lib. ii, CRSHR, vol. i, pp. 182–183, translation from EHD, vol. ii, pp. 348–349. Ralph de Diceto (d. 1202)

790

Wyatt

William of Newburgh’s account in this instance reveals once again the ambivalence with which the English chroniclers regarded the chivalric monarchs of Scotland. The chronicler clearly felt that William the Lion was a product of his own more ‘civilised’ cultural milieu. Yet, William of Newburgh also recognised that there was an irreconcilable incompatibility between this ‘civilised’ Scottish monarch and his more ‘barbarous’ and traditional subjects. This cultural incompatibility clearly continued to foster deep-seated tensions within Scottish society. These tensions were to resurface, once again, following William the Lion’s capture by English forces at Alnwick in 1174. Indeed, William of Newburgh relates that when William’s warriors learned of his fate they turned upon the chivalric Englishmen within their own ranks. So seizing this opportunity, the Scots disclosed their inborn hatred against them (i.e. William’s English knights), which they had previously concealed through fear of the King, and slew as many as they lighted upon, while those able to escape took refuge in the royal fortress.193 William’s description of these events is one of many that illustrate how twelfthcentury English perceptions regarding the other populations of British Isles were founded as much upon cultural and behavioural differences as they were upon nascent ‘national’ identities or political affiliations. This cultural and behavioural differentiation was constructed around the English elite’s adoption of the revised norms and sentiments of the ecclesiastical reform movement. Moreover, the vehement condemnations that the English chroniclers directed against the violent slave raiding activities of traditional warriors suggest that their activities were more than just acts of cultural conservatism. They were deliberate and symbolic statements of identity and cultural defiance against the external values and norms that were infiltrating into their communities.194

also commented that during this raid the Scots led “… away young women captive …” (“… captivas abducere mulierculas …”). Tmagines Historiarum Radulfi de Diceto Decani Lundoniensis Opera Historica, vol. i, W. Stubbs (ed.) (London: Longman, 1876), p. 376. 193  “Occasione ergo temporis Scotti innatum, sed metu regio dissimulatum, in illos odium declarantes, quotquot incidebant peremerunt, refugientibus in munitiones regias ceteris qui evadere potuerunt.” Ibid., translation from EHD, vol. ii, p. 350. 194  This contention is supported by the work of R. Andrew MacDonald who has argued that Somerled, ruler of Argyll, and Fergus of Galloway deliberately affiliated themselves with more traditional Norse-Gaelic culture and made alliances with Irish and Scandinavian rulers in direct opposition to the “new world order” being introduced by the Scottish monarchy during the twelfth century, see “Rebels”, pp. 166–186, 170.

Slavery And Cultural Antipathy



791

Invasion, Antipathy and Slavery in Twelfth-Century Ireland

Due to its geographical positioning Ireland was somewhat less susceptible to the English cultural infiltration that so dramatically affected both Wales and Scotland during the late eleventh and early twelfth centuries. Yet, political and trading contacts had existed between England and Ireland for centuries. The island’s maritime links meant that it was by no means an isolated and culturally homogeneous region.195 Nevertheless, the vibrant Scandinavian settlements on Ireland’s coastlines had resulted in a continuing Irish cultural affiliation with the more traditional sensibilities of the Northern world. Ireland had long been a refuge for political exiles and refugees from England and following the Norman Conquest many disaffected Anglo-Saxon warriors appear to have fled to the courts of Irish kings such as Diarmait mac Maél na mBó of Leinster (1047–1072).196 That many of these traditional English warriors chose to flee North to Scotland and West to Ireland rather than to the Continent is highly indicative of the cultural alignment of these regions.197 However, during the twelfth century there appears to have been increasing economic, social and cultural contact between Ireland and her more southerly neighbours.198 William of Malmesbury reveals the increasing influence that the English crown was exerting upon the Irish elite in this period. He reports that when Muirchertach Ó Briain, king of Dublin and Munster (1086–1119), defied Henry I the English king punished him by placing a trade embargo upon Ireland. William notes that following the imposition of this economic blockade the Irish king soon capitulated and from thenceforth he and his successors “… were so devoted to our King Henry that they wrote nothing except what would please him and did nothing except what he told them.”199 An increasing political and cultural affiliation between the English and Irish elites is also illustrated by the actions of Diarmait Mac Murchada, king of Leinster (1132–66, 1170–71). In 1166 Tigernán Ua Ruairc, prince of Bréfne, and Diarmait Ua Máel Sechlainn of Meath expelled Diarmait from his kingdom. Yet, rather than choosing a Norse, Scottish or Welsh ally Diarmait looked to

195  Davies, Domination, p. 7. 196  Ibid., p. 5. Diarmait gave asylum to the sons of Harold Godwineson following 1066 just as he had provided a refuge for their father during his exile in 1051. 197  In addition, this reflects contemporary political alignments. 198  Ibid., p. 7. 199  “… ita deuotos habuit noster Henricus ut nichil nisi quod eum palparet scriberent …”, GRA, pp. 738–739.

792

Wyatt

Henry II and Richard fitz Gilbert, earl of Pembroke for support.200 Diarmait’s alliance with Fitz Gilbert was to have far reaching implications as it heralded an era of English invasion and settlement in Ireland. We have already seen that this resulted in expressions of cultural antipathy on both sides. Indeed, of all the twelfth-century English condemnations regarding the ‘barbaric’ mores of their ‘Celtic’ neighbours those directed against the Irish population would appear to have been the most venomous. This, in turn, suggests that the English elite who settled in Ireland experienced a more acute sense of cultural and behavioural differentiation than had their counterparts who settled in Wales or Scotland. This may in part be due to the physical separateness of Ireland and also to the comparatively low level of cultural interaction between the two populations prior to 1169.201 Nevertheless, it was the Irish population’s failure to adopt the norms of the reform movement (together with the closely associated revised codes in warfare and values of chivalry) that was by far the most important factor in the construction of this sense of cultural difference. By the twelfth century many individuals within the Irish Church had readily embraced the revised norms and codes of conduct expounded by the continental reform movement.202 Reforming synods were held and prelates from Ireland travelled to France and Italy and studied at the prominent reform schools of learning.203 However, the extreme significance of traditional warrior norms and expressions of patriarchal power in Ireland meant that Irish ecclesiastics found it extremely difficult to impose their revised codes of behaviour onto their secular elite. We have already seen that Archbishop Lanfranc of Canterbury had complained about the polygamous behaviour of Irish kings in 1073–1074.204 Furthermore, the late twelfth-century accounts of Gerald of Wales and Bernard of Clairvaux suggest that the Church’s attempts to impose its strict codes concerning monogamy, exogamy and legitimacy upon the Irish 200  F. X. Martin has noted that Diarmait had been a vigorous supporter of the reform movement and he established a Cistercian abbey at Baltinglass, County Wicklow in 1148. Yet, despite his strong English connections it appears that Diarmait had continued to express his power through traditional practices. Indeed, the antipathy between Diarmait and his chief protagonist Tigernán Ua Ruairc had, in part, been stimulated by Diarmait’s abduction of Tigernán’s wife, Derbforgaill, in 1152, see “Diarmait”, pp. 49–50. In addition, Diarmait had ordered that Mór, the Abbess at Cell Dara, be abducted and raped shortly after his accession in 1132, see chapter 3. 201  This may also be due to the nature and attitudes of contemporary commentators such as Gerald of Wales. 202  Smith, “Frontiers”, p. 251 and Brett, “Canterbury’s perspective”, pp. 13–35. 203  Davies, Domination, pp. 16–18. 204  See above.

Slavery And Cultural Antipathy

793

population had been largely unsuccessful.205 The Irish ecclesiastics’ failure in modifying such traditional behaviour prompted Gerald to complain that … they are too slack and negligent in the correction of a people that is guilty of such enormities. Because of their not preaching to and reproving their people, I preach that they should be reproved themselves.206 Yet, Gerald’s attitude towards his Irish counterparts may have been unnecessarily harsh. Expressions of warrior power through virility and violence were of fundamental social importance in Irish society. Any attempt to modify these expressions and alter traditional behavioural norms would have undoubtedly been met with entrenched opposition.207 As we have already seen the AngloSaxon ecclesiastical hierarchy achieved little or no success when faced with a very similar challenge despite the best efforts of powerful individuals like Archbishop Wulfstan of York.208 The papacy clearly felt that the Irish population’s recalcitrance in the face of reform warranted external intervention. Pope Adrian IV’s Laudabiliter of 1159 must, therefore, be viewed in a similar context to Archbishop Thurstan’s absolution of English troops prior to the battle of the Standard.209 The Irish ecclesiastical hierarchy certainly welcomed Henry’s arrival in 1171. Indeed, they gladly received this new and powerful ally in their struggle to impose reform upon the secular population of Ireland.210 Moreover, Henry’s invasion of Ireland appears 205  See notes 36 and 53, above. See also J. Gillingham, “The Beginnings of English Imperialism”, Journal of Historical Sociology, 5, 1992, pp. 392–409, 403–405 and “Foundations”, pp. 61–62. See also Davies, English Empire, pp. 129–130. 206  “In episcopis vero et prælatis hoc fere solum reprehensione dignum invenio, quod in populi tam enormiter delinquentis correctione desides nimis sunt et negligentes. Quod igitur nec prædicant nec corripiunt, hinc ipsos prædico corripiendos; quod non arguunt, hinc argu; quod reprehendere negligunt, hinc reprehendo” TH/EH, pp. 173–174, translation from O’Meara, HTI, pp. 112–113. Paradoxically, Gerald praised the Irish clergy for their adherence to the ascetic lifestyle, which suggests that reform norms were firmly established within the ecclesiastical population of Ireland, ibid., p. 172. 207  Just as they had been in pre-Conquest England, see chapter 4. 208  Ibid. 209  See above. 210  This appears to be borne out by first decree enacted by the Synod of Cashel in 1172 which ordered “… quod universi fideles per Hiberniam constituti, repudiato cognatarum et affinium contubernio, legitima contrahant matrimonia et observent.” (“… that all the faithful throughout Ireland should repudiate cohabitation between those related by kinship or marriage, and should enter into and abide by lawful marriage contracts.”) EH, pp. 98–99. See also Davies, Domination, p. 17 and Smith, “Frontiers”, p. 251.

794

Wyatt

to have been regarded by the English ecclesiastical chroniclers as a further victory in their continuing struggle to impose a uniform international Christian code of ethics upon their neighbouring communities. The English elite almost certainly shared the attitudes of their chroniclers who regarded slave-raiding/ trading activities to be synonymous with sexual sin, immorality and ‘barbarous’ behaviour. They were undoubtedly aware that the continuing existence of the slave markets in Ireland, and in particularly Dublin, were perpetuating the trade in slaves. In addition, they recognised that these markets were being supplied with English, Norman and Flemish victims from Northumbria, Cumbria, South Wales and the Welsh Marches. The intense desire to impose reform norms upon the Irish population was, therefore, supplemented by a desire to extinguish the Irish Sea slave trade. This was made clear in Gerald of Wales’s account of the Council of Armagh in 1170 when the English conquest of Ireland was said to have been justified … because of the sins of their own people, and in particular because it had formerly been their habit to purchase Englishmen indiscriminately from merchants as well as from robbers and pirates, and to make slaves of them, this disaster had befallen them by the stern judgement of the divine vengeance, to the end that they in turn should now be enslaved by that same race. For the English, in the days when the government of England remained fully in their hands, used to put their children up for sale—a vicious piracy in which the whole race had a part—and would sell their own sons and relations into Ireland rather than endure any want or hunger. So there are good grounds for believing that, just as formerly those who sold slaves, so now also those who bought them, have, by committing such a monstrous crime, deserved the yoke of slavery. The aforesaid council therefore decided that throughout the island Englishmen should be freed from the bonds of slavery and restored to their former freedom.211 211  “… propter peccata scilicet populi sui, eoque precipue quod Anglos olim tam a mercatoribus quam predonibus atque piratis emere passim ei in servitutem redigere consueverant, divine censura vindicte hoc eis incommodum accidisse, ut et ipsi quoque ab eadem gente in servitutem vice reciproca iam redigantur. Anglorum namque populus, adhuc integro eorundem regno, communi gentis vicio liberos suos venales exponere, et priusquam inopiam ullam aut inediam sustinerent, filios proprios et cognatos in Hiberniam vendere consueverant. Unde et probabiliter credi potest, sicut venditores dim ita et emptores tarn enormi delicto iuga servitutis iam meruisse. Decretum est itaque predicto concili, et cum universitatis assensu publice statutum, ut Angli ubique per insulam servitutis vinculo mancipati in pristinam revocentur libertatem.” EH, pp. 69–71. Gerald provides the only account of this Council.

Slavery And Cultural Antipathy

795

This is a particularly fascinating tract as it illustrates how the dynamics of English cultural antipathy towards slave raiding activity was intimately entwined with the objectives of the reform movement. Gerald’s account reveals that the Council of Armagh’s settlement was extremely similar in objectives and tone to the peace treaty enacted by Alberic, bishop of Ostia following the battle of the Standard. It implies that because Anglo-Saxon society had participated in the sin of slave trading/raiding, it had been conquered by the Normans and then integrated into the ‘civilised’ and religiously uniform sphere of continental Europe. By the same token, Irish society had now been punished by conquest for its participation in these illicit practices and should now be similarly integrated. The Council’s statement constituted an enforced prohibition against such unchivalric activities in Ireland and demanded the emancipation of all those Englishmen who had been alienated from their origins through enslavement. The motivations for such measures were inextricably bound up with the ideology of the reform movement and this was made clear at the subsequent Synod of Cashel in 1172. This Synod proclaimed that … it is right and just that, as by divine providence Ireland has received her lord and king from England, she should also submit to a reformation from the same source. Indeed, both the realm and the church of Ireland are indebted to this mighty king for whatever they enjoy of the blessings of peace and the growth of religion; as before his coming to Ireland, all sorts of wickedness had prevailed among this people for a long series of years, which now, by his authority and careful administration are abolished.212

212  “Itaque omnia divina, ad inster sacrosancte ecclesie, iuxta quod Anglicana observat ecclesia, in omnibus partibus ecclesie amodo tractentur. Dignum etenim et iustissimum est ut, sicut dominum et regem ex Anglia sortita divinitus est Hibernia, sic etiam exinde vivendi formam accipiat meliorem. Ipsi namque regi magnifico tam ecclesia quam regnum Hibernie debent quicquid de bono pacis et incremento religionis hactenus est assecuta. Nam ante ipsius adventum in Hiberniam, multimoda malorum genera a multis retro temporibus ibidem emerserant, que ipsius potencia et munere in desuetudinem abiere.” EH, p. 100, translation from IHD 1172–1922, p. 19. Again, Gerald is our only source for this document. Scott and Martin translate the first sentence as “… it is proper and most fitting that, just as by God’s grace Ireland has received her lord and king from England, so too should she receive a better pattern of living from that quarter.” I have, therefore, chosen to use Curtis and McDowell’s translation, which correctly highlights the reform minded sentiments behind the decrees of the document. Moreover, the “multimoda malorum” that were of primary concern to the reforming hierarchy appear to have been the practice of illicit marriage and concubinage.

796

Wyatt

The rich Irish commercial centres, such as Dublin, appear to have been the primary targets of Henry II who quickly stamped his authority upon Richard fitz Gilbert’s enterprise and reserved control of these centres for the English crown.213 The reasons for Henry’s interest in these commercially vibrant settlements have, therefore, appeared all too obvious to modern historians writing from politically specific and economically based perspectives. Duffy has noted that the over-lordship of Dublin gave Henry full claim to the high kingship of Ireland, and Ó Croínín argues that the king’s real purpose was to assert his authority over the territories that had been conquered by his vassals.214 Wallace has argued that the incredible wealth generated by the port of Dublin acted as a magnet for royal intervention.215 Such considerations undoubtedly motivated Henry II in his actions. However, the emotive forces of cultural antipathy should not be underestimated. Henry wished to gain control over the lucrative Irish marketplace of Dublin yet he must also have been acutely aware that the wealth of this city had been primarily founded upon the Irish Sea slave trade. In spite of the immense profits that this trade had generated Henry appears to have followed the advice of his ecclesiastical advisers and placed a prohibition upon slave trading there. Like his predecessor William the Conqueror, then, Henry clearly felt that his chivalric and pious image as a supporter of reform was far more valuable than any economic gain that could be made from allowing this trade to continue. His prohibition of slave trading was further facilitated by the slaying and/or evacuation of many key figures within the Hiberno-Norse community of Dublin, a situation supplemented by a subsequent influx of English settlers from Bristol and elsewhere.216 By 1283 only 40 Ostmen of relatively low status remained in the cantreds of the kingdom of Dublin.217 Moreover, Henry and his leading churchmen clearly recognised that by assuming control of Dublin and the other Hiberno-Norse towns they could

213  S. Duffy, “Ireland’s Hastings: The Anglo-Norman Conquest of Dublin”, ANS, xx, 1998, pp. 69–87, 80. 214  Ibid., pp. 80–85 and Ó Cróinín, Medieval Ireland, p. 288. 215  Wallace, “Archaeology”, p. 159. Duffy agrees with Wallace and remarks that Henry wasted no time in establishing his authority over the merchants of the city, see “Ireland’s Hastings”, pp. 80–82, 85. 216  Duffy, “Ireland’s Hastings”, p. 85. For a full discussion of this process see E. Purcell, “The Expulsion of the Ostmen, 1169–1171: The Documentary Evidence”, Peritia, 17–18, 2004, pp. 276–294. 217  J. Bradley, “The Interpretation of Scandinavian Settlement in Ireland” in Settlement and Society, pp. 48–78, 62.

Slavery And Cultural Antipathy

797

deny disruptive warrior factions within the British Isles access to their traditional refuges, bases of operations and slave market places. The Irish and Hiberno-Norse were fully aware of the cultural and political ramifications of the English conquest of Dublin and the city was not relinquished without a struggle. In 1171 a huge Irish-Scandinavian fleet of 160 ships assaulted the city but was heavily defeated.218 Following this Ruaidrí Ua Conchobair of Connacht, the high king of Ireland, besieged the city for two months with a hosting of men from all over Ireland, but was finally tricked by an English counter attack into a humiliating defeat.219 Interestingly, Ruaidrí’s forces, like king David’s army at the Battle of the Standard, included a conglomerate of traditional warriors. Amongst the Irish contingents there was Tigernán Ua Ruairc, prince of Bréifne and Murchad Ua Cerbaill, king of Airgialla along with Mac Duinn Shléibe, king of Ulaid, king Domnall Ua Briain of Thomond, Murchad Mac Murchada, king of Uí Chennselaig and a huge fleet made up of Norsemen from the Isle of Man and the earldom of Orkney.220 The unprecedented nature of this gathering suggests that these warriors were all too aware of the implications of this chivalric intrusion into their society and culture; their fears would appear to have been justified. When Henry II assumed political control over his Irish dominions in 1172 he asserted English supremacy through his support for the key objectives of the reform movement in Ireland.221 By adopting this reforming stance Henry was following a tried and tested English royal strategy that had initially been instigated by William I. This royal adoption of reforming values was, at times, used as a cynical pretext for territorial, political or economic expansionism. Nevertheless, it clearly provided rulers like Henry II with a powerful ideological justification for their actions. In the eyes of the contemporary English elite and their chroniclers Henry had struck a further blow against the sinful forces of the traditional warrior order in Britain. The English imperial expansionism into Ireland during the second half of the twelfth century was to trigger centuries of cultural antipathy between traditional elements within Irish society and their chivalric counterparts. Indeed, Henry II’s newly acquired role as the champion of the Irish Church appears to have denied the native laity of much of their authority. One consequence of this appears to have been a significant revival in secular paganism during 218  Duffy, “Ireland’s Hastings”, p. 79. 219  Ibid., pp. 79–80. 220  EH, note 115, p. 306. 221  See notes 211 and 212, above.

798

Wyatt

the early years of the thirteenth century.222 This pagan revival was focussed around the birth of the prophesised warrior-king, Aodh Eanghach, who, it was foretold, would rid Ireland of her invaders and whose fertility was said to be intimately connected with the soil of Ireland.223 The pagan cult of Aodh appears to have dissipated following the arrival of the Dominican and Franciscan friars in the mid-thirteenth century.224 Yet, its existence provides evidence of the power and potency of pre-Christian value systems in Ireland. Moreover, the violent ravages of traditional style warrior fraternities were to plague Irish society for many centuries.225 It appears that the disruptive and dangerous lifestyle of these groups also attracted certain English settlers. As late as 1297 the Anglo-Irish parliament complained that certain degenerate Englishmen had taken to wearing fían style hairstyles and assuming a life of brigandage.226 In addition, English warriors were not averse to adopting traditional strategies of warfare. The Annals of Connacht relate that in 1236 a host of the Goill Erenn (the foreigners or English settlers) were searching for an outlawed Irish nobleman named Fedlim mac Cathail Chrobdeirg. When Fedlim eluded them they chose instead to raid the locality and “… captured many noble women and carried them off into captivity and bondage.”227 Their English mainland contemporaries would have undoubtedly regarded such behaviour as a sign that these colonists had been polluted by their ‘barbarous’ social surroundings.228 Indeed, Gerald of Wales warned of the dangers that faced the English settlers living in Ireland because

222  K. Simms, From Kings to Warlords (Woodbridge: Boydell, 1987), pp. 15–16, 26–28 and K. Simms, “Gaelic warfare in the Middle Ages” in T. Bartlett and K. Jeffrey (eds.), A Military History of Ireland (Cambridge/New York: Cambridge University Press, 1996), pp. 99–115, 104–106 and M. Mhaonaigh, “Pagans and Holy Men: Literary Manifestations of Twelfth Century Reform” in Bracken and Ó Riain-Raedel, Ireland and Europe, pp. 143–161, 154. 223  Simms, Kings to Warlords, p. 27. 224  Ibid., p. 15, and Simms, “Gaelic Warfare”, p. 106 and J. Watt, The Church in Medieval Ireland (Dublin: Gill and Macmillan, 1972), pp. 76–77. 225  Simms, “Gaelic Warfare”, pp. 101–102, 104–106. 226   Statutes and Ordinances, and Acts of Parliament of Ireland, King John to Henry V, H. F. Berry (ed.) (Dublin: HMSO, 1907), vol. i, pp. 210–211. See also Simms, “Gaelic Warfare”, p. 101. 227  “… 7 cur gabatur moran do mnaib maithi ann sin 7 co rucsat leo iat a mbroitt 7 i ndairi …” Annála Connacht, The Annals of Connacht, A. Martin Freeman (ed. and trans.) (Dublin: Dublin Institute for Advanced Studies, 1944), pp. 58–61. 228  See Davies English Empire, pp. 136–140 and Gillingham, “Invasion of Ireland”, pp. 28–29.

Slavery And Cultural Antipathy

799

… he who touches pitch will be defiled by it; that foreigners coming to this country (i.e. Ireland) almost inevitably are contaminated by this (treachery), as it were, inborn vice of the country—a vice that is most contagious.229 Yet, in spite of the activities of the thirteenth-century Goill Erenn it appears that English hegemony over Dublin effectively removed the nerve centre of the slave supply network and eventually brought large-scale slave raiding and trading in the British Isles to an end. The demise of this trade in the ‘Celtic’ regions can, therefore, be viewed as militaristic enforcement of chivalric values and reforming norms. The removal of slave markets and traders, however, did not end smaller scale raiding activities or bring a halt to the cultural antipathy between the English elite and more traditional elements within the communities of the British Isles. Nevertheless, the increasing prestige attributed to the ideals of the reform movement and the accompanying revised warrior norms were to have a profound impact upon the traditional systems of patriarchy within these British societies. The powerful expansion of English cultural and political influence during the twelfth century acted as a catalyst for the dissemination and adoption of reform values throughout the British Isles. As these revised norms came to be adopted by the neighbouring elites their behavioural conduct especially in warfare became radically modified. In particular, the practice of slave raiding that had once been a power affirming symbol of cultural identity came to be regarded as an uncivilised activity associated with lascivious behaviour and immorality. By the thirteenth century the secular elite within the British societies had, generally, come to regard slave raiding and trading activities to be illicit exercises associated with common brigandage and thievery rather than with power, masculinity and warrior prowess.

229  “… et qui picem tangit coinquinabitur ab ea … adeo inquam, bonus mores corrumpunt colloquia prava, ut hoc vitio patriæ tanquam innato et contagiosissimo etiam alienigenæ huc advecti fere inevitabiliter involvantur.” TH/EH, p. 168 translation from O’Meara, HTI, p. 109.

An Explanation of Military Slavery Daniel Pipes This chapter seeks to account for the two basic facts of military slavery: that it existed at all and that it occurred only in Islam-dom. Why would anyone choose to recruit soldiers as slaves? Why did Muslims alone in fact do so? These questions are intertwined; the raison d’être of military slavery is bound up with its purely Islamicate existence; its causes cannot be understood apart from Islam.

A Connection to Islam?

The distribution pattern of military slavery sketched in chapter 2 strongly suggests that Islam lay behind the existence of this institution. Yet before inquiring into the role of Islam, let us consider other possible factors for the existence of military slavery. Of the alternatives, three stand out: climate, Turks, and the stirrup.1 Muslim lands from Spain to northeast India share a basic climatic feature—dry heat; perhaps this or some other environmental feature could explain military slavery. Without attempting to elaborate on this connection, we can reject it as a cause by noting that the lands from Spain to India enjoyed the same climate before 622 ce. Were military slavery related to climate, we would expect it to have existed in those same regions before the appearance of Islam; but it did not.2 Furthermore, military slavery occurred also in wetter areas such as southern and eastern India. Turks are often associated with military slavery, with good reason. The Abbasid use of slave soldiers coincided with their importation of Turks; Turks served in almost every military slave corps in the Middle East and northern India; and they dominated the best-known and most spectacular group, the Source: Pipes, Daniel, “An Explanation of Military Slavery,” in Daniel Pipes, Slave Soldiers and Islam: The Genesis of a Military System, New Haven: Yale University Press, 1981, 54–102. 1  The notion that military slaves were merely mercenaries in Muslim clothing has been treated in chapter 1. The assumption here is that they were a distinct phenomenon. Some scholars grant that military slavery occurred only in Islam-dom but resist the idea that it is connected to Islam. See M. A. Cook, introduction to V. J. Parry et al., A History of the Ottoman Empire to 1730 (Cambridge, Eng., 1976), p. 7; Frederick Cooper, Plantation Slavery on the East Coast of Africa (New Haven, 1977), p. 194 n. 183. 2  Pp. 161–66 discusses some of these possibilities.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_026

An Explanation Of Military Slavery

801

Mamluks of Egypt. Their close connection to the institution of military slavery both as soldiers and as rulers makes it reasonable to consider the Turks its originators and propagators. Two considerations show that this was not so, however. The first resembles the argument against a climatic explanation; if Turks were the cause, why did military slavery not exist among Turks outside of Islamdom? Turks in the preIslamic Middle East and in non-Muslim areas did not institute military slavery. Most notably, for centuries they faced Byzantium and numerous empires in China and pre-Islamic Iran; they also fought for the Muslims for 150 years before becoming military slaves.3 Non-Muslims made efforts to incorporate Turks into their service, yet none devised the system of military slavery.4 Nor did the system exist in the Turks’ Inner Asian homelands. A second argument against the Turkic explanation comes from the Umayyad dynasty in Spain. It appears that simultaneously with the development of military slavery in Iraq under the Abbasids, this Spanish dynasty was independently making systematic use of slaves as soldiers. No Turks were present in early 3d/9th-century Spain.5 Later, too, sub-Saharan dynasties made extensive use of slave soldiers, and almost none of them were Turks. Thus, Turks correlate much less clearly with military slavery than does Islam. The stirrup offers a much better reason for military slavery. To understand how, I shall summarize Lynn White, Jr.’s ideas on the stirrup in Europe6 and then explore its possible effects on Islamdom. Placing stirrups on a horse vastly enhances the power of its rider. When he sits in the saddle without any grip for his feet, a soldier wields a lance with the strength only of his arms and legs. The stirrup welds horse and rider into a single unit; now the soldier can attack with his lance carrying the entire force of his and the horse’s combined weight. Heavy cavalry outfitted with stirrups enjoys a great preponderance over infantry (it does not affect light cavalry in most cases). But it was exceedingly expensive; equipment for one horseman in medieval Europe cost about the same as plough-teams for ten families—not even counting the price of remounts and retainers. White approaches the stirrup’s role in medieval Europe by accepting Heinrich Brunner’s classic statement of 1887 on the origins of feudalism. 3  D. Pipes, “Turks in Early Muslim Service,” documents this. 4  On the Turks and Crusaders, see note 95 below. 5  Pp. 192–93. 6  Masterfully presented in chapter 1 of his Medieval Technology and Social Change (London, 1962). For one of many critiques, see R. H. Hilton and P. H. Sawyer, “Technical Determinism: The Stirrup and the Plough,” Past and Present 24 (April 1963): 90–100.

802

Pipes

Brunner connects two events which occurred closely in time: the increased role of cavalry in the Franks’ armies between 732 and 775 ce and the confiscation of church lands by Charles Martel (r. 714–41). He suggests that the Franks had to augment their cavalry and that the church lands helped to pay for the vast new expenses of horses, equipment, and training; but nowhere does Brunner explain what it was that compelled the Franks to increase their cavalry at that time. Here, where prior theories were lacking, White introduced the stirrup as an explanation. The metal stirrup originated in fifth-century China, whence it passed to Central Asia and the Middle East. The Muslims had it by 74/694 and surely used it after 92/711 in their conquest of Spain and invasions of Gaul. Although the Muslims lost to the Franks at Poitiers in 115/733, the latter presumably noted the stirrup’s effectiveness and took steps to adopt it, enlarging the cavalry and designating church lands to pay for it. According to White’s thesis, the stirrup brought about a military transformation which contributed to a far-reaching social upheaval, the feudal order. The soldiers on horses who benefited from the new military order acquired political and economic power too and evolved into an aristocracy. Thus, the stirrup led to a series of adjustments which created asocial division in western Europe: it produced a military aristocracy and charged the peasantry with the expense. Did the stirrup have a comparable effect in Islamdom?7 Perhaps it caused a shift along geographic rather than social lines. In most of the heavily populated parts of the Eastern Hemisphere (notably those in North Africa, the Middle East, India, and China), war-horses could not be raised.8 The introduction of the stirrup enhanced the power of the peoples living where horses could be raised—primarily in steppe lands and in deserts—and reduced the strength of peoples living in densely inhabited areas, especially cities. By making horses more important to warfare, the stirrup redistributed power from civilized to barbarian peoples. The non-horse-breeding areas in Islamdom, like the peasantry in western Europe, became a nearly passive source of funds for the horse aristocracy, in this case from the steppes and deserts. The centers of civilization came under assault from horse-breeding barbarians during the roughly seven centuries when cavalry reigned supreme (700–1400 ce). If one accepts this reasoning—and I propose it here without being convinced of it myself—then military slavery appears to be an answer by the civilized centers to the predations of the horse barbarians. The centers adopted various strategies to deal with the horsemen: all made efforts to acquire horses 7  Its possible effects in sub-Saharan Africa are discussed in J. Goody, Technology, Tradition, and the State in Africa (London, 1971), pp. 34–37. 8  Here, as so often, Japan resembles Western Europe.

An Explanation Of Military Slavery

803

of their own; and the Muslims also made the imaginative and successful effort to bring in the riders too. Thus, military slavery can be understood as a response to the shift in military balance caused by the stirrup. To explain military slavery in this way also neatly accounts for its appearance not long after the stirrup appeared on the scene, without any reference to Islam. Unlike the climatic or Turkic reasons, this one answers why military slavery did not exist before Islam. It also explains why so many of the slave soldiers came from steppe and desert regions. Yet even the stirrup theory correlates far less well than Islam does. If it was the decisive factor, why, then, did the other non-horse-breeding areas (Hindu India and China in 700–1400 ce) not also develop military slavery? Furthermore, many military slaves were foot soldiers, like the black troops in Egypt and the Janissaries, the renowned elite corps of the Ottoman Empire. Something besides the stirrup must have been involved in the role of slave soldiers in Islamdom. The striking correlation between Islam and military slavery shown in chapter 2 raises the possibility of a causal connection between this religion and this pattern of military recruitment. However, correlation alone, as statisticians and logicians never tire of pointing out, does not imply causation. Some correlations are merely unrelated; no one claims significance for the “nearperfect correlations … between the death rate in Hyderabad, India from 1911 to 1919, and variations in the membership of the International Association of Machinists during the same period.”9 Others are counterlogical: statistics indicate that the presence of fire engines at fires correlates with more destructive fires; can one infer that fires would be less destructive if no fire engines were dispatched?10 High correlation between Islam and military slavery “can serve only as the starting point for further investigation and analysis.”11 In general, “to establish the regularistic causal proposition that X caused Y, three things must be demonstrated. First, there must be a correlation between X and Y. Second, there must be a proper temporal relationship in their occurrence, X i must occur before Y i. Third, there must be at least a presumptive agency which connects them.”12 In the case of Islam and military slavery, correlation does exist and the temporal sequence is proper (Islam preceded military slavery); to prove a causal link, “a presumptive agency which connects them” must be established. 9  D. H. Fischer, Historians’ Fallacies: Toward the Logic of Historical Thought (New York, 1970), pp. 168–69. 10  R. M. MacIver, Social Causation (Boston, 1942), p. 92. 11  Ibid., p. 93. 12  Fischer, Historians’ Fallacies, p. 169.

804

Pipes

Islamic or Islamicate? Military slavery may have been connected either to the religion or the civilization of Islam. If it is considered a part of the religion or its legal system, we call it “Islamic”; if an aspect of the more diffuse civilization which accompanies it, then “Islamicate.” Islamic elements are myriad and indisputable: the five “pillars of the religion”; the lesser ritual prescriptions; the entire body of the Sharīʿa, which deals with almost every aspect of human existence; and numerous features not required by Islam but characteristic of Muslim life, such as turban wearing, use of the Arabic script, and Sufi orders. All these elements come with the religion as part of an Islamic package; they are nonfunctional and can only be explained in the light of Islamic ideals and traditions. Clearly, military slavery is not Islamic; it has no religious sanction and it is not even unambiguously legal.13 Nothing whatever argues for its being part of the Islamic religion or its legal system.14 If not Islamic, is it Islamicate? Islamicate elements are not an outgrowth of Islamic religion and law, yet are integral to Muslim life. Such an element need not be by nature Islamic: The specifically Islamic quality of a cultural element might well … owe nothing to its origin, but simply express the fact that Islam, by taking it up, put its mark on it or tended to appropriate it. Islam … tended to call forth a total social pattern in the name of religion itself…. In many spheres, not only public worship but such spheres as civil law, historical teaching, or social etiquette, Muslims succeeded quite early in establishing distinctive patterns identifiable with Islam as religion.15

13  See pp. 94–95. 14  This conclusion does not imply that there are no Islamic military patterns, but only that military slavery is not one of them, H. J. Fisher has argued for the significance of Islamic prayer to the conduct of Muslims in battle (“Prayer and Military Activity in the History of Muslim Africa South of the Sahara,” Journal of African History 12 (1971): 391–406). P. Moraes Farias has drawn a connection between prayers and closed-line formation (“The Almoravids: Some Questions concerning the Character of the Movement during its Period of Closest Contact with the Western Sudan,” Bulletin de l’Institut fondamental d’Afrique noire 29, ser. B (1967): 794–878, especially pp. 811–20). See also W. H. McNeill, The Rise of the West (Chicago, 1963), p. 428 n. 11. 15  Brunschvig, p. 54; Hodgson, 1:89.

An Explanation Of Military Slavery

805

Even something so utterly unrelated to Islam as the lateen sail can be characteristic of Muslims.16 Regardless of origin, such an Islamicate pattern must fit somehow into the structures established by Islam: [It is] legitimate to consider Muslim doctrine as a factor, not only when it happens to introduce a new solution from its own resources or brings about a new solution directly or indirectly, but also each time that, having integrated an interior or foreign solution into its system, and colored it its own way, it contributed to getting it adopted or maintained. How many practices, which have nothing Islamic about them in principle, have been naturalized as Muslim to the point of becoming characteristic of Islām…?17 By nature, Islamicate patterns are less distinct than Islamic ones, for they lack the clear impress of Islam. Some have been observed,18 but very few have been systematically tied to Islam. Military slavery lends itself particularly well to this relation because it correlates to Islam so visibly and has no comparable existence outside Islamdom. How does one make a connection between slave soldiers and Islam? 16  J. H. Parry, The Establishment of the European Hegemony: 1415–1715: Trade and Exploration in the Age of the Renaissance (New York, 1966), p. 21. 17  Brunschvig, p. 54. 18  For some studies, see note 3 to chapter 1. The following list presents my casual compilation of some features which appear to be Islamicate (omitting those which are discussed in this chapter): Political: no legislation (only ad hoc decisions); severe problems of succession; huge palace complexes; very few women in political life; absence of municipal organizations; nonterritorial loyalties; discomfort living under non-Muslim rule. Military: heavy use of cavalry. Social: fluid class structures and social mobility; clothing differentiates social and ethnic statuses; sharp distinction between in (Muslim) and out (nonMuslim) groups; kinship ties paramount. Economic: commerce prestigious, agriculture scorned; silver and gold coinage both present; wealth an attribute of power, rarely the reverse; little scope for risk money; slaves abundant. Intellectual: religious authorities control education; memorization emphasized; religious orientation of nearly all cultural life. Artistic: representations discouraged; calligraphy emphasized; geometric and vegetal forms prevalent. Geographic: uncohesive structure of cities; concentric arrangement and hierarchical division of quarters; houses built to insure privacy; deforestation (due to sheep and goats replacing pigs); urban domination of surrounding districts; religious segregation in cities and countryside. Sexual and psychological: separation of the sexes: the veil and harem; women’s honor emphasized; four interpersonal relations present, four absent (according to Halpern). Other: men encouraged to marry father’s brother’s daughter; Arabic personal names; few organized athletics (except in Iran).

806

Pipes

A number of writers, both social thinkers and historians of Islam,19 have linked military slavery with the patrimonial nature of Islamicate governments. (A patrimonial government is one based ultimately on relationships of paternal authority and filial dependence in contrast to a feudal one, which is marked by contractually fixed fealty based on knightly militarism.)20 Patrimonial rulers always needed trusted agents to execute their will; outsiders (aliens and unprivileged subjects) have the fewest conflicting interests, so they provide the most reliable agents. From this vantage point, military slavery could be said to be a means of acquiring such agents. Muslims alone developed a system to bring in agents, because their governments were the most patrimonial. This simple explanation is unfortunately neither adequate nor accurate. It is not enough to explain military slavery as a consequence of patrimonialism; the basic question still remains: why were Islamicate governments patrimonial? Furthermore, it is not true that Islamicate dynasties were predominantly patrimonial; many of the most important ones (for instance, the Seljuks, Ayyubids, Mamluks, Ottomans, Safavids, and Mughals) had important feudal elements. Moreover, it is a mistake to view military slaves as political agents; over and over, they were recruited primarily as soldiers. They were warriors brought in to staff armies; although they also served their masters as agents, military duties came first. The reason why only Muslims established military slavery lies deeper, in the nature of Islamic political ideals and their effect on the actual conducting of politics.

Why Muslim Subjects Relinquished Power

We have established that military slavery must have been connected to the civilization of Islam. Now we face the central question: what Islamicate reasons caused Muslims alone regularly to recruit their soldiers as slaves? What uniquely Islamicate pattern caused military slavery to come into existence? This section identifies a fundamental pattern of public life in Islamdom: withdrawal by Muslim subjects from the governments and armies which ruled them; the subsequent sections show who took their place, connect this pattern to military slavery, and demonstrate how military slaves served Muslim rulers 19  On them, see D. Pipes, “The Strategic Rationale for Military Slavery,” Journal of Strategic Studies 2 (1979): 34–35. My explanations suggested in this article are superceded by the arguments presented here. 20  R. Bendix, Max Weber: An Intellectual Portrait (London, 1966), p. 360.

An Explanation Of Military Slavery

807

better than their alternates. Military slavery recedes temporarily into the background in the following pages; the discussion here concerns a fundamental characteristic with numerous and profound consequences for Islamicate public life; military slavery was but one of those, though an especially visible one. The argument concerning Islamicate public life is presented as a hypothesis, and a springboard for discussion, not as a firm conviction about the “presumptive agency” connecting slave soldiers and Islam. Islamic Public Ideals While all religions postulate ideals that human beings cannot consistently maintain, Islam alone of the universalist religions makes detailed political ideals part of its basic code, the Sharīʿa.21 To understand the nature of the statecraft in Islamdom, one must appreciate the special place of political and military ideals in the Islamic tradition. It is most vividly seen in contrast to the Jewish and Christian traditions. In their formative periods, both Judaism and Christianity evolved outside a state structure and without government support; in contrast, Islam from the first and almost always thereafter developed in conjunction with political authority. After the Jewish kingdom was destroyed in 586 bce, Jews had to adopt their religion to exile, dispersal, and political disenfranchisement; as a result, Judaism grew independent of constituted authority and had no need for it. Christianity from its very inception drew a clear line between the realms of religion and politics, disassociating itself from government. Intermittent Roman persecution during its first three centuries accentuated this division, although it diminished when Christianity became the Roman state religion. The relationship of Christianity to the state then became a burning issue for the next millennium and a half in the Catholic church (less so for other churches), but was not clearly resolved. After 1500 ce, Protestant reaffirmation of the early ideal of separation found increasing acceptance, so that by now Christianity has again turned away from politics. Even when the Catholic church was most powerful West European political institution in the High Middle Ages, a pious Christian could fulfill his religious duties without reference to the government. Jews and Christians are inclined to view as government a necessary evil.

21  I am not considering Confucianism a religion. The following argument connecting military slavery to Islam was inspired by a reading of C. S. Kessler, “Islam, Society and Political Behaviour: Some Comparative Implications of the Malay Case.” This insightful article testifies to the original and unorthodox ideas which those who study Islam in the peripheries can provide.

808

Pipes

In contrast, Islam from its first years had close relations with the state. Muhammad, as is well known, founded not only a religion but the community of its believers, a political unit: he served both as prophet and statesman for the infant community. From that time, each Muslim ruler has not only a mundane role but also an Islamic one; his foremost duty is to carry out the Sharīʿa. Good Islamic government means enforcing its regulations and creating an environment where Muslims can live a correct life; the sources of Islamic inspiration have thus a great deal to say about proper political conduct. A bad government only partially enforces the Sharīʿa; anarchy means no enforcement at all. Islam requires a government run by Muslims because the fulfillment of a proper religious life depends on this enforcement; no other agency can maintain the Sharīʿa, represent the umma, or wage jihād. Thus, while Jews and Christians in the practice of their religions can ignore the authorities above them, Muslims cannot. Jews formed closed minority communities for thousands of years without attaining sovereign control; on the contrary, the Church has been so powerful that it has on occasion threatened to take over the state. More dependent on government, Islam is vulnerable to its vicissitudes; a failure by Muslims to attain the public ideals required by their religion would have serious consequences for their attitudes toward government and involvement with it. The contrast between ideals and the actualities of public life has thus been especially significant for Muslims. Umma, Caliphate, Jihād: The Ideal The political and military ideals of Islam are well known: they are the subject of introductory courses in Islam and of full-length studies,22 so they need not detain us long. Three Arabic words may be used to sum up these ideals: umma, caliphate (= khilāfa), and jihād, referring to the community of Islam, its political leadership, and its warfare. The umma of Islam emerged in Medina under Muhammad. It developed in opposition to the communities of the other monotheistic religions and to the Arabian tribes. The Islamic umma aimed to make the Muslims into both a religious community like those of the Jews and Christians and into a supratribal unit in which Islam replaced kinship-based affiliations. The convert to Islam put prior affiliations aside when he joined the umma. As Islam became 22  Some of these studies include: L. Gardet, La Cite musulmane: vie sociale et politique (Paris, 1954); Khadduri; E. I. J. Rosenthal, Political Thought in Medieval Islam: An Introductory Outline (Cambridge, England, 1958); Tyan.  The following discussion ignores financial and judicial ideals, although the same argument applies equally to them (e.g. zahāh, Haarmann, pp. 10–14).

An Explanation Of Military Slavery

809

universalistic, when non-Arabians joined, the umma became universalistic too. From humble beginnings in Medina, it spread through Arabia and then (very sparsely, for there were few Muslims at first) to the lands between Spain and India. In subsequent centuries, the umma filled out as conversions to Islam increased, and it spread yet more widely into new areas. Whatever its extent, the umma was a single unit, ideally under the leadership of a single man, the caliph. Loyalty to the caliph was the tangible expression of devotion to the unity of the umma. The position of caliph emerged in an unplanned way after the death of Muhammad in 11/632; the fledgling umma needed a leader, so a successor (khalīfa) to Muhammad emerged. He had no prophetic function but succeeded Muhammad only as leader of the umma. The caliph symbolized the unity and power of the Muslim community. The unity of the umma and the rule of the caliph both point to a third characteristic Islamic ideal, the jihād, military action intended either to defend or expand the boundaries of lands ruled by Muslims.23 This ideal has two important implications; first, it encourages, under proper circumstances, aggression against non-Muslims as fulfillment of a religious precept. Second, it prohibits warfare among Muslims, for violence must only be employed to spread the rule of Islam—and how can it be spread to regions that Muslims already control? The ideals represented by the umma, caliphate, and jihād complement each other; Islamic doctrines call for political unity and peace among Muslims. Differences between believers must not lead to political divisions, much less to war. Umma, Caliphate, Jihād: The Reality In contrast to the theoretical statements of Islamic ideals, their actual role in history has gone almost unnoticed. Premodern Muslim kingdoms, rulers, and warfare varied widely over 1,200 years and several continents. Kingdoms ranged from local dynasties to far-flung empires, rulers from military conquerors to religious revivers, warfare from tribal skirmishes to sieges of many years. Within this large diversity, one fundamental pattern prevailed: political and military realities virtually never met the ideals established by Islam. The actual course of Islamicate history compares sadly with the ideal of a unified umma under a caliph waging jihād against non-Muslims only. Both of these ideals were transgressed against early and permanently. The unity of the umma did last, admittedly, for about thirty years (until the murder of ͨUthmān in 35/656), 23  It is always important to remember that jihād spreads the rule of Islam, not the religion. Muslims may use violence to gain political control but not to coerce belief; and since nonMuslims pay higher taxes, there is usually no incentive to convert them anyway.

810

Pipes

but major intra-Muslim warfare began in 37/657 with the civil war between ͨAlï and Muͨ āwīya; a century later, in 138/756, the umma was formally split when the Umayyad ruler of Spain refused to recognize the Abbasid caliph as sovereign; from that time on for about a century, a new region became independent every five years.24 The unity of the umma was never reestablished. Perceived common enemies (for example, the Crusaders, Mongols, Zionists) could produce local feelings of solidarity thereafter, but no positive bonds effectively united Muslims. Furthermore, the size of the umma and its diversity made unity impossible; the ever expanding area of Dār al-Islām was simply too large to remain politically unified. The caliphate endured a slower decline. It withstood the vicissitudes of the first three centuries, remaining powerful and virtually unchallenged. Shīʿīs, Khārijīs, and rebellious governors had only limited impact, and even when the caliphs fell under the control of military slaves in the mid-3d/9th century, their prestige survived. Then the first rival caliph emerged in 297/ 909, when the Shīʿī Fatimids took power in Tunisia; not long after, in 316/929, the ruler of Umayyad Spain also declared himself amīr al-muʾminīn (“Commander of the Faithful,” another term for “Caliph”). More serious yet, the Shīʿī Buyids conquered Iraq in 334/945 and held the Abbasid caliphs near-captive for a century; from this date on the caliphs became mere figureheads. Seljuk substitution for Buyid overlordship in 447/1055 did not change this, though the Seljuks were Sunnīs. Aside from a purely local revival at the end of the 6th/12th century, the caliphate remained politically obscure until its demise at the hands of the Mongols in 656/1258. After that only a very attenuated version of the office continued to exist for several centuries before disappearing altogether. Its revival by the Ottoman rulers in the late 12th/18th century was a doomed undertaking. The caliphs did fill the important task of representing the political power of Islamdom until the Buyid conquest; from that time on, Islamdom has been politically fragmented. The jihād was partially maintained; from the Arabian conquests onward, Muslims have eagerly defended or expanded the borders of Dār al-Islām. The Byzantine, Spanish, and Indian fronts saw sporadic fighting over eight centuries, the Balkan over six; Muslims responded actively to the Crusader and Mongol invasions; they took up arms sporadically against pagans in sub-Saharan Africa and on the Inner Asian steppe; and in modern times, they resisted encroachments by seaborne Western Europeans and land armies advancing from Russia and China. Yet, though Muslims carried out the injunction to fight non-Muslims, they ignored the prohibition against warfare among believers. 24  A list of the breakaway dynasties may be found on pp. 178.

An Explanation Of Military Slavery

811

Indeed, Muslims fought one another far more often than they did the infidels; true jihād constituted a pitifully small percentage of their total warfare, for in contrast to the distant boundaries of Dār al-Islām, those of neighboring Muslim governments were close by. Disputes with fellow-Muslims arose much more often than with infidels. To be sure, when Muslims fought Muslims, each side armed itself with doctrinal justifications; but the fact that both parties considered themselves Muslim could never be ignored or forgotten. Umma, caliphate, and jihād represent only the summits of Islamic public ideals; the Sharīʿa also concerns itself with the details of much else, generating equally deep gulfs between expectations and reality in such matters as taxation and legal justice. Islamic public ideals have remained permanently unattainable since shortly after the time of Muhammad: the political order envisaged in the Qurʾān, Sunna, Sharīʿa, in the legal handbooks and political treaties, has never existed.25 But Islamic public ideals were none the less important for being unattainable. Though beyond reach, they constantly influenced attitudes toward politics and warfare, causing Muslim subjects to look askance at the governments that divided the umma and the armies that wasted Muslim lives. The presence of these ideals caused Muslim subjects to reject their rulers, and in this way, Muslim rulers failed. They failed not by external criteria (for example, capacity to dominate, duration of rule) but by falling short of Islamic standards.26 A government did not embody the aspirations of its subjects, its boundaries had little meaning for them, and its warfare had virtually none. Yet Muslim rulers were no worse than others; their subjects were dissatisfied not because of any inherent insufficiencies but because of heightened Islamic expectations. Muslim subjects looked for successors to Muhammad but instead found 25  Kessler, pp. 38–39. 26  I must reemphasize that this is not my judgment on the Muslim rulers but an understanding of how their subjects felt. The reader may note two modern parallels: (1) In the twenty years following World War II, Americans heard incessantly about the virtues of their country and the benevolent role it was playing in the world. This created a political ideal which the US government could not in reality attain; as a result, each stumble seemed like a calamitous failure. The Vietnam War and the Watergate Affair (the latter a peccadillo by any other standard) led to severe disaffection because they contrasted darkly with the ideals. In the United States this was a passing theme, a result of the ideological excesses of the Cold War; the gap between ideal and real in premodern Islamicate civilization was a permanent part of the religiously-based culture, constantly generating disaffection. (2) By parading glamorous, unattainable images, movies and television (especially advertisements) often convince the viewer that his own life is deficient and leaves him dissatisfied with reality.

812

Pipes

business-as-usual politicians. The existence of Islamic public ideals served more to denigrate real governments than to guide them; those ideals undermined more than they sanctioned. This may be seen most clearly with regard to the umma. The political importance of the umma lay not in its positive ability to mobilize Muslims but in its negative effect on real governments. The umma ideal made reality look transient and miserable; Muslims preferred the vision of their glorious community. The existence of an ideal of the umma thus worked to restrict the appeal of existing kingdoms. Even after the unified umma had become a pipe dream, it remained a vital emotional anchor. Regardless of the divisions into kingdoms and the warfare among them, the community of Islam did not lose its allure for Muslim peoples; they saw devotion to the umma as an expression of allegiance to Islam and its principles. Concern for the totality of Muslims, often a fuzzy notion, took the place of local patriotism. Muslim subjects felt more of a bond to the umma than to their governments; the transitory, arbitrary, local rule of some king paled in the face of the permanent, grand, universal umma.27 As the unity of the umma, rule by the caliph, and warfare against only unbelievers turned into pious fictions, Islamic ideals became ever more isolated from life. The government of Allah and the government of the sultan grew apart. Social and political life was lived on two planes, on one of which happenings would be spiritually valid but actually unreal, while on the other no validity could ever be aspired to.28 But, unreal as they were, Islamic public ideals continued to dominate Muslim thinking. Any number of examples show this; military slavery itself makes the point well. The scholars argued fine points of Sharīʿa but did not recognize— or at least acknowledge—this preeminent institution. With the single exception of Niẓām al Mulk’s brief description (Appendix 2), no writer on politics deigned to discuss military slavery! And it never appears in the Sharīʿa. Having no place in theory, slave soldiers had no place in the Muslim consciousness, which remained attuned to Islamic ideals and pulled away from discordant realities. But the world was filled with harsh facts; how did the Muslims deal with them?

27  Goitein, p. 40; Lapidus, p. 30; Hodgson, 2:53, 57. 28  Grunebaum, Medieval Islam, p. 143; also p. 153.

An Explanation Of Military Slavery

813

The Withdrawal from Power Muslim subjects responded to troublesome realities by withdrawing from politics and warfare. They avoided the blatant nonimplementation of Islamic goals in the public sphere by retreating into other domains, where Islam had much greater success. Not all Muslim subjects withdrew, of course, but insofar as an Islamicate pattern existed, Islam caused its believers not to participate in public affairs. As a result of the unattainable nature of Islamic public ideals, Muslim subjects in premodern times relinquished their political and military power.29 In practice this meant that Muslim subjects turned away from the dominant political and military institutions. “The prevailing attitude toward power is skepticism”;30 the populace stayed aloof from the decision makers. This “widespread disinclination to collaborate in government” was particularly true of those who took the Islamic message most seriously.31 The ʿulamāʾ did their best to avoid serving as qādīs; some Sufis refused “to touch funds coming from an amīr, on the ground that they represented illicit gains”; and in the 13th century courageous jurists in Egypt declared prayer in a cemetery chapel, which the Sultan had erected, as not permitted, on account of the inhumane methods used during its construction. A tacit, boycottlike opposition to the government existed in certain pious circles of the early Islamic Middle Ages, when any money coming from the government was considered religiously forbidden property. It is even reported that some pious people considered it forbidden to drink water from a canal dug by the … government, or to fasten their bootlaces at the light of a lamp belonging to the government.32 The reputations of Muslim authorities who did serve the rulers suffered as a result; the trustworthiness of Abū Yūsuf (d. 182/798), author of Kitāb alKharāj, came into doubt because he worked for the government.33 A number 29  Ideas for this argument came initially from a discussion with Richard W. Bulliet on 21 December 1977. 30  Grunebaum, Islam, p. 25. 31  Ibid., p. 132. 32   Ulama’: N. J. Coulson, “Doctrine and Practice in Islamic Law: One Aspect of the Problem,” Bulletin of the School of Oriental and African Studies 18 (1956): 211–26. On Sufis see Hodgson, 2:96. On jurists see H. Ritter, “Irrational Solidarity-groups. A Socio-psychological Study in Connection with Ibn Khaldūn,” Oriens 1 (1948): 32–33; Grunebaum, Medieval Islam, p. 143. 33  Ibn Khallikān, 6:379.

814

Pipes

of prominent Muslim ethical and religious writers, such as Abū’l-Layth asSamarqandī (d. 373/983) and Abū’l-Ḥamīd al-Ghazālī (d. 505/1111) discussed the problems entailed in dealing with rulers.34 Islamicate writings widely reflected this attitude to government. Further, Muslim subjects and rulers alike internalized this attitude, making it a normative pattern of behavior; all agreed that Muslim subjects should not become involved in public affairs.35 Muslim subjects avoided armies (less so navies) even more than governments and administrators. Hearing repeatedly about the unity of the umma and the sinfulness of fighting against other Muslims, they stayed out when, as was usually the case, the enemy was Muslim. (Exceptions here are those cases where Muslims fought to save their own lives.)36 As a result, almost everywhere 34  Goitein, pp. 205–06. 35  An intriguing alternative explanation for the alienation of Muslim subjects from their rulers can be traced back to an argument Hodgson makes about the ecology of the Middle East creating a culture which had few ties to the land:  “The unusual access to interregional trade in the mid-Arid Zone, combined with its aridity and its openness to overland conquests and to imperial formations, all having increasing effect in the course of the millennia, could have resulted, over the whole region rather than just locally, in an unusual degree of legitimation of culture oriented to the market, and this is a form favouring cosmopolitan mobility rather than civic solidarity (2:73–74).”  The aridity of its agricultural lands, its central location in the Oikoumen, and its accessibility to conquest by land combined to make Middle Eastern civilization little oriented to geographical loyalties. Hence, Middle Eastern Muslims naturally directed their loyalty more to the umma. This then became part of the high culture which spread with Islam.  Another explanation for the uninvolvement of Muslim subjects comes from a different Middle Eastern tradition. Autonomous cities had existed since the most ancient times, before bureaucratic empires functioned. Cities were accustomed to pay off imperial foreigners with quitrents in return for local freedom of action in the hands of the religious and commercial elites. Although this procedure was outdated by about 600 bce, it remained in force through Parthian and even Sasanian times; perhaps its legacy extended into the Islamic period as well and expanded from the Middle Eastern base to all Islamdom. 36  In 389/999, for example, the populace of Transoxiana “followed the advice of its teachers and decided that ‘when the struggle is for the goods of this world’ Muslims are not obliged to ‘lay themselves out to be murdered’ ” (quoted in W. Barthold, Turkestan Down to the Mongol Invasion, trans. V. Minorsky et al., 3d ed. (London, 1968), pp. 267–68). This implies that Muslim populaces should normally stay out of warfare. The Sharīʿa reflects this view, too: “Abstain and desist from civil strife [between Muslims]” al-kaff wa’l quʿūd fīl-fitna. Ibn Baṭṭa al-ʿUkbarī (d. 387/997) ash-Sharḥ wa’l-Ibāna ʿalā Uṣūl as-Sunna wa’d-Diyāna, ed. and trans. H. Laoust (Damascus, 1958), 67/126, with further references. (R. Stephen Humphreys provided me with this reference.)

An Explanation Of Military Slavery

815

in Islamdom and throughout premodern times, Muslim subjects rarely fought for the armies that ruled over and protected them. To be sure, Muslims did agitate against the government that ruled them: disloyal governors, oppressed peasants, unruly desert tribes, and angry city mobs sporadically rebelled, but except when non-Muslims threatened, indigenous subjects of a ruler almost never made a sustained attempt to control their own government. Notables with local roots, landlords or merchants, hardly ever led the subjects to overthrow a regime and put themselves in power. (Rare attempts to do this were, significantly, led by religious figures.) A common pattern found in other premodern civilizations almost never exists in Islam-dom: indigenous peoples in charge of their own governments and staffing their own armies. When non-Muslims threatened everything changed; danger to the Sharīʿa or to Dār al-Islām meant that Islam had to be defended, and in these circumstances Muslin subjects often became more active. In early Islamic times, before many of the conquered peoples had converted, the Muslims felt endangered and participated more actively in war. So, too, an offensive jihād stirred Muslims to fight. The main marches of Islamdom (Spain, Anatolia and the Balkans, India) demonstrate this point vividly, as do areas such as West Africa and Central Asia.37 Faced with Christian, Hindu, Buddhist, or animist enemies, Muslim subjects responded favorably to their rulers’ need for military manpower. Similarly, European, Chinese, and Japanese imperialism mobilized them in modern times; and most recently, the US presence in Iran aroused an atavistic Muslim response. When Muslim subjects fought, they most probably perceived a threat from non-Muslims; but when a government reasonably maintained the Sharīʿa and kept the infidels at bay, they stayed away from armies and tended to their private gardens. So, instead of public life, Muslim subjects in normal times concentrated on personal matters. They were principally interested in leading the good life and much less in who administered it.38 Intense family, communal, and religious involvements, where Islamic precepts and ideals were often attained, took the place of power politics and warfare. Persons interested in righteous living did best to restrict their activities to private affairs.

37  A few examples are: the Samanid and Saffarids in early Islam (when Muslims were still a tiny minority in most places), Syrian cities 340–550/950–1150 (when the Byzantines threatened), and the Sarbadarids (in response to a Mongol threat). Haarmann, pp. 18–19, points out the unusual nature of the Muslim response to the Crusader challenge. 38  Grunebaum, Islam, p. 132.

816

Pipes

Affiliations tended to be either small-scale or Islamic. (Small-scale groupings are face-to-face societies in which everyone virtually knows everyone else; they are typically based either on proximity or on kinship relations.) In rural areas of Islamdom, villages and tribes predominated; in urban areas, quarters and fraternal associations (such as youth clubs, trade guilds, even criminal gangs) had the most importance.39 Everywhere the family came first, though, as the supreme locus of Muslim life. Larger affiliations (in which a person does not know everyone) derived from Islam. These were primarily two, the madhhab and the ṭarīqa. The madhhabs were systems of jurisprudence (often translated as “law schools”) which developed into social institutions. Each madhhab undertook to translate the regulations of the Qurʾān and ḥadīth into a complete legal structure. They evolved out of early study groups of scholars; by the early 5th/11th century, nearly all Sunnī Muslims held allegiance to a madhhab (and Shīʿī and Kharījī groups each had their own madhhabs, too). Surprisingly, these informally organized legal affiliations became vehicles for popular expression in the period between al-Ma‌ʾmūn’s reign and the Buyid Conquest (ca. 200–330/820–950), largely as a result of the practice of one madhhab, the Ḥanbalī, of organizing common people “into cadres dedicated to defending the principles of the school.”40 The other madhhabs appear to have followed the Ḥanbalīs and to have similarly transformed their systems of jurisprudence into genuinely popular movements.41 At the same time, they spread geographically and each of them became predominant in some regions.42 The ṭarīqas (Sufi orders or brotherhoods) developed from the 6th/12th century and paralleled the madhhab organization on a more emotional and mystical level. Although Sufi thought had existed long before, the orders were first organized in the 6th/12th century. The ṭarīqas differed widely one from another and between regions; always, however, they supplied an emotionally and socially conducive context for intense religious feelings; strong bonds between members of the same lodge made an important contribution to the social order. They offered “a sense of spiritual unity” which the failed political institutions were unable to provide.43 For most Muslims, ṭarīqas played a much larger social role than did any of the government agencies. 39  For discussions of these social questions, see the writings of Claude Cahen, Marshall G. S. Hodgson, and Ira M. Lapidus. 40  Lapidus, p. 36, relying largely on the work of Henri Laoust, George Makdisi, and Dominique Sourdel. 41  Ibid., p. 41. 42  Ibid., pp. 42–43. 43  Hodgson, 2:221.

An Explanation Of Military Slavery

817

Similarly, the religious elites, the Sufi masters (pīrs) and the ʿulamāʾ usually replaced the ruler as the source of authority and prestige for the average urban Muslim. Because the political structure lay beyond the usual concerns of Muslim subjects, religious status had more bearing than political or military power. These elites provided daily guidance for the populace and also served as intermediaries between Muslim subjects and their rulers. They were indigenous. These disparate affiliations, small-scale and Islamic, shared one key feature; none of them had coercive powers—they could not tax, raise armies, nor claim a monopoly on violence. Some of them might occasionally mobilize on a local scale, but none could challenge the rulers. They absorbed energies which otherwise would have gone into politics. The commonplace assertion that Islam does not distinguish between religion and politics is true; but, paradoxically, by embracing politics and warfare, by making them central to Islamic life, Islam removes them from the lives of most Muslims. “While Islam is in one sense the political community par excellence it has tended to make the pious Muslim more and more nonpolitical.”44 The mixing of religion with politics and warfare leads to a sharp division between public and private domains; instead of government and warfare, Muslim subjects devoted themselves far more to religious, social, and family concerns. As a result, “the true central thread of Islamic history lies not in the political realm of the caliphs and sultans but in the social realm where the ulama served as the functioning heart of the historic Muslim community.”45 Politics and warfare have played a smaller role in the lives of Muslims than in those of other peoples; only when non-Muslims threatened did they engage in those areas themselves. The ruling structure stood in striking isolation from the peoples’ lives;46 in particular, it could not draw them in as soldiers. Who, then, staffed Islamicate armies? Withdrawal by Muslim subjects created a power vacuum which opened Islamicate public life to domination by others. Armies became the playthings of nonsubjects; one succeeded another with hardly any reference to the subject populations.47

44  Grunebaum, Islam, p. 136. For another view, see R. Bendix, Kings or Peoples: Power and the Mandate to Rule (Berkeley, 1978), pp. 47–49. 45  Bulliet, p. 138. 46  On this, see R. W. Bulliet, The Patricians of Nishapur (Cambridge, Mass., 1972) and I. M. Lapidus, Muslim Cities in the Later Middle Ages (Cambridge, Mass., 1967). 47  Nonsubjects dominated politics almost as much, but the following discussion takes up only their military role.

818

Pipes

Marginal Area Soldiers

Two geographic terms, “marginal area” and “government area” sharpen the analysis of Islamicate military patterns; soldiers in Islamicate armies nearly all came from marginal areas, the steppes, deserts, mountains, and forests being thus defined. Throughout premodern times, these hinterlands supported only simple forms of social and cultural life, for they contained few cities or large concentrations of population; their inhabitants usually lived far apart from each other or did not settle at all. They more often had contact with animals than with fellow humans outside the family. The economy was simple and nearly self-sufficient, though articles from the cities were highly prized and formed the basis of an important trade. Social life was rudimentary, with few divisions or distinctions between individuals. Culture remained at a popular level, for marginal areas could not support specialists; the minstrel or folk artist had to devote his days to providing for himself. Intellectual life beyond traditional wisdom hardly existed. Political organization in marginal areas followed predominantly tribal lines, though not always (the Balkan population lived in mountains without being tribal, while even city-dwellers in sub-Saharan Africa remained tribal). On the whole, institutions stayed simple, rarely involving more persons than knew each other. Ibn Khaldūn calls these peoples savage (wahṣhī) and enumerates them: “The Arabians and the Zanāta are such peoples, as are similar groups, for instance the Kurds, Turkomans, and the Veiled Ṣanhāja.”48 The “government areas” included what was not marginal, the permanently settled lands where city-dwellers and farmers lived more closely together. Unlike marginal areas, which permitted only simple economies and cultures, government areas supported more complex forms of human activity; in a word, they housed civilization. With regard to political organization, they developed large-scale structures—governments—in contrast to the tribes in marginal areas.49 The distinction between outsiders and insiders parallels that of government and marginal areas. “Insiders” are the ethnically dominant, economically stable, socially accepted, religiously conforming population of a settled region. They populate both cities and countryside and have a stake in preserving the established order because they prosper in it. “Outsiders,”50 by contrast, are a diverse lot, either foreigners or outcast indigenous groups such as minorities 48  Muq, 1:295 (my translation from the Arabic, 1:263). 49  The blād as-sibā and blād al-makhzan in Morocco reflect this distinction most precisely. 50  Ibn Khaldūn uses the term khārijīya to express “outsider” (Muq 1:279 n. 84, 376).

819

An Explanation Of Military Slavery

(persons differing in language, religion, color, or some other significant trait), the indigent, slaves, and any other disadvantaged group.51 The outsiders who filled empty spaces in Muslim armies came overwhelmingly from marginal areas. They enjoyed vital military and political advantages over soldiers from government areas, even though they were culturally less sophisticated. Ibn Khaldūn52 expresses this contrast symbolically: “Sieves were altogether non-existent among (the Arabs) and they ate wheat (kernels) with the bran. Yet the [military] gains they made were greater than any other ever made by human beings.”53 Several elements contributed to the recurrent military superiority of marginal area soldiers: the hardships of their way of life, their healthiness, and their social organization; the fact that no government controlled them was of key importance. Whether outsiders lived in the steppes, deserts, or mountains, merely staying alive was a steady challenge. Agriculture was erratic and animals lived in a fragile climatic balance as the vagaries of weather hit with particular severity. Life was hard: each person had to maintain excellent physical strength in order to carry out his daily functions of farming, herding, or hunting. The marginal areas could not afford soft jobs (for example, as domestic servants, aristocrats, or philosophers) or nonproductive institutions (such as opulent courts). Pleasures were simple and comforts primitive; their environment compelled these peoples to stay lean and hardy. Marginal area peoples were also healthier; spending most of their lives away from densely populated areas, far from

51  A person coming from a marginal area is automatically an outsider, for an insider must participate in constituted society and therefore live in a government area. On the other hand, an outsider may come from either a marginal or a government area so long as he is from outside a given society. Area Person

Marginal

Government

Outsider Insider

× ×

× ×

52  Ibn Khaldūn’s writings accord so well with the following argument that I shall combine quotes from al-Muqaddima and al-ʿIbar with my views. In the process, I shall admittedly be taking his statements out of context, for Ibn Khaldūn stresses groups whereas I am looking at individuals. Still, he often comments on the qualities of individuals, too, and it is these I shall quote. 53  Muq, 1:419.

820

Pipes

insects and rodents and away from standing water, they escaped most of the endemic diseases found in settled government areas. The inability to support a government implied an absence of public authority in marginal areas; this had important social consequences. It forced marginal area peoples to protect themselves by grouping together and reenforcing the bonds of mutual trust. A single individual could not protect himself; he had to be guaranteed by his groups, usually the family and tribe. Elaborate codes of honor and vigilante tactics developed to ensure order. The total effect was to sharpen each person’s wits and military capacity. Raiding for booty and feuding for honor were endemic; for reasons of both defense and attack, every male practiced the martial arts from infancy, was trained as a soldier, and stayed in practice at all times. Belonging to a mutual assistance group with military capabilities gave the social structure of marginal area men an inherent potential for warfare. They belonged to ready-made military organizations—the tribes. Besides these, they had few groupings: the urban quarters, fraternal groupings, madhhabs, and ṭarīqas found in government areas barely existed in the marginal lands. The tribe and village were usually the only forums. Lack of government fostered an independent ethos; each person had to fend for himself against the elements and his fellow men. Ibn Khaldūn describes the spirit this engendered as follows: The Bedouins … live separate from the community. They are alone in the country and remote from militias. They have no walls and gates. Therefore, they provide their own defense and do not entrust it to, or rely upon others for it. They always carry weapons. They watch carefully all sides of the road. They take hurried naps only when they are together in company or when they are in the saddle. They pay attention to every faint barking and noise. They go alone into the desert, guided by their fortitude, putting their trust in themselves. Fortitude has become a character quality of theirs, and courage their nature. They use it whenever they are called upon or an alarm stirs them.54 In sum, their environment compelled these peoples to develop the physical and mental qualities ideal for military prowess. The weak perished and the survivors made powerful soldiers. Again Ibn Khaldūn: Since … desert life no doubt is the reason for bravery, savage groups are braver than others. They are, therefore, better able to achieve superiority 54  Muq, 1:257–58.

An Explanation Of Military Slavery

821

and to take away the things that are in the hands of other nations…. The more firmly rooted in desert habits and the wilder a group is, the closer does it come to achieving [military] superiority over others.55 Besides these martial advantages which marginal area soldiers held over all government area peoples, they had an additional strength over Muslims from government areas; by definition, there was no government in marginal areas, so the peoples who lived there were not ruled by a constituted political authority which fell short of the Islamic ideals. They had nothing, therefore, from which to turn inward. Muslims who were not subjects did not relinquish power because no government existed to alienate them. Their small-scale and Islamic groupings involved them in the exercise of power; as a result, they consistently ruled and fought. The combination of not being alienated and innate military superiority gave marginal area soldiers an enormous advantage over Muslim peoples. It explains why they were the main Muslim political actors; they dominated Islamicate public life, founding the dynasties and staffing the armies. The relinquishment of power by insiders and the strength of marginal area soldiers resulted in a general pattern; soldiers from steppes, deserts, and mountains had an almost total hold over Islamicate armies.56 To a degree unmatched in any other civilization57 they dominated the armies that brought new governments to power and their subsequent replacements. Here we show the extent to which they brought the new dynasties in; the following pages explain why the troops continued to come mostly from marginal areas even after a government had been established. It is easy to demonstrate the role of marginal area soldiers in founding new governments; a survey of armies similar to that done in chapter 2 (which looked at military slaves) makes this clear, for the major dynasties mentioned

55  Muq, 1:282–83. 56  This role has received but scant attention. Some hints may be found in: Ashtor, p. 18; Bosworth, “Recruitment,” p. 64; J. C. Hurewitz, “Military Politics in the Muslim Dynastic States 1400–1750,” Journal of the American Oriental Society 88 (1968): 97; J. Celerier, “Islam et geographie,” Hesperis 39 (1952): 347; Gellner, p. 3, has noted this fact most clearly: “It seems to be a striking feature of the history of Muslim countries that they frequently, indeed generally, have … penumbra of marginal tribalism.” 57  Hodgson makes this point, too, but with different terminology and a different explanation. He calls it “militarization” (2:64) and attributes it to a “stalemate between agrarian and mercantile power” (2:65).

822

Pipes

there, as well as most others, depended primarily on marginal area soldiers to bring them to power.58 Such a pattern existed nowhere else to the same extent as it did in Islamdom, not even in the same lands before they became Muslim. In pre-Islamic Egypt, for example, while marginal area soldiers had some role from the time of the Hyksos onward, native Egyptians almost always fought in the armies of Egypt, They predominated when Egypt was independent and served as auxiliaries when foreigners (Libyans, Cushites, Persians, Greeks, and Romans) ruled.59 But with the coming of the Muslims, native Egyptians (except for the Bedouin, who were not insiders) largely disappeared from the armies of the government. It was not until modern times, 1238/1823, when Muḥammad ʿAlī conscripted Egyptian peasants, that they again reentered the army. Muhammad ʿAlī undertook this novel measure partly because his Nubian recruits died off; but, more importantly, he had seen the success the French had with a national army. Recruitment While steppe, desert, and mountain soldiers provided a source of great power, they had their own particular drawbacks. However mightily they began, they rapidly became unreliable after conquering a government area, in either of two ways: some settled down and lost their martial strength; others retained that strength but became unruly. In either case, they became undependable and had to be replaced with fresh soldiers. Marginal area men could, of course, deteriorate and become unruly simultaneously, but for the sake of clarity, I shall analyze these processes separately. The striking contrast between the warriors of one generation and their effete grandsons has provoked much speculation. Though the rapid degeneration of marginal area soldiers is a conspicuous pattern, its causes remain vague. Originally, courage and hardiness characterized the marginal area soldier; these were not innate qualities but were acquired by living in a harsh environment, which he left on entering a polity. The milieu which had forged those qualities was necessary to maintaining them too. Once they undertook the softer life as rulers, marginal area soldiers began to lose the very qualities 58  The short accounts of dynastic origins provided in C. E. Bosworth, The Islamic Dynasties (Edinburgh, 1967) make this clear. 59  J. Lesquier, Les Institutions militaires de l’Egypte sous les Lagides (Paris, 1911), pp. 134–35; R. O. Faulkner, “Egyptian Military Organization,” Journal of Egyptian Archeology 39 (1953): 32–33, 37, 46; A. R. Schulman, Military Rank, Title and Organization in the Egyptian New Kingdom (Berlin, 1964), pp. 28 ff.; but see L. A. Christophe “L’Organisation de l’armée égyptienne a l’époque ramesside,” Revue du Caire 20 (1957): 399–400, for a different view.

An Explanation Of Military Slavery

823

which had brought them military success. “Whenever people settle in fertile plains and amass luxuries and become accustomed to a life of abundance and luxury, their bravery decreases to the degree that their wildness and desert habits decrease.”60 As a new world of amusements, affluence, and culture opened up, marginal area soldiers often willingly and joyfully reveled in it. This easy life so badly undermined their courage and hardiness that already their own offspring lacked their power; their sons fought no better than other men born in the polity and their grandsons might have fought even less well. The loss in martial virtues was both rapid and permanent. Eventually, “the warrior distinguishes himself from the artisan only by his inability to work; the men of marginal areas differ from civilized men only in their clothing.”61 Ibn Khaldūn devotes much attention to the reasons for the decay of dynasties. He specifies three elements in this process: the love of glory (which destroys ʿaṣabīya, group-feeling), indulgence in luxuries (which destroys the economy), and tranquillity (which undermines the soldiers’ martial qualities). He makes the following points about the loss of martial qualities: When people become accustomed to tranquillity and rest and adopt them as character traits, they become part of their nature…. The new generations grow up in comfort and the ease of luxury and tranquillity. The trait of savagery (which former generations had possessed) undergoes transformation. They forget the customs of desert life that enabled them to achieve royal authority, such as great energy, the habit of rapacity, and the ability to travel in the wilderness and find one’s way in waste regions. No difference remains between them and ordinary city dwellers, except for their (fighting) skill and emblems.62 Besides lapsing into comfortable indolence, grandsons of warriors lost the other advantages of the marginal areas, their good health, strict codes, independent spirit, and lack of alienation from the government. As they became identical to government area men, the army suffered from a loss in soldierly qualities. This degeneration of marginal area soldiers forced the ruler and military leadership to seek out new sources of soldiers. 60  Muq, 1:282. John of Salisbury (d. 1180) put it as follows: “If in war men’s bodies are wounded with swords, in peace they are no less wounded with pleasures.” The Statesman’s Book, translated and selected by J. Dickinson (New York, 1963), p. 14. 61  ʿIbar, 6:3. 62  Muq, 1:341–42.

824

Pipes

However, it was possible for marginal area soldiers to keep their martial skills and avoid deterioration by maintaining their old lifestyle. Cattle nomads who continued herding most often retained their old abilities; although their locale changed, their activities remained the same, so their descendents stayed hardy and fought well. The unreliability of these soldiers lay not in their military aptitude but in their political unruliness. As victorious troops, the marginal area soldiers who brought a dynasty to power felt privileged; they and their descendents placed great demands on the ruler and considered themselves entitled to whatever he gave them. These soldiers considered the ruler in their debt and tolerated him only as an arbiter. They pressed him for concessions, squabbled among themselves, and obeyed him only when it suited them. In short, the ruler could not impose his will on them but had to tolerate them as an independent force. With time, however, the ruler would find their attitudes and power intolerable as he increasingly took on the responsibilities and challenges of a sovereign. Especially when the marginal area soldiers were the ruler’s kinsmen did they grow fat and clamor loudly for privileges, so he was forced to sever his dependence on them: “A ruler can achieve power only with the help of his own people…. [Eventually, however, ] the ruler shows himself independent of his people, claims all the glory for himself, and pushes his people away from it with the palms (of his hands). As a result, his own people become, in fact, his enemies.”63 Therefore, about three generations after the founding of a dynasty, some of the grandsons of warriors lapsed into comfortable indolence and others made unacceptable claims on the ruler. As the descendents of marginal area soldiers became unreliable, the army had to lessen its dependence on them and replace them with new troops. In order to prevent them [his people] from seizing power, and in order to keep them away from participation (in power), the ruler needs other friends, not of his own skin, whom he can use against (his own people) and who will be his friends in their place. These (new friends) become closer to him than anyone else.64 Who were these new friends, and where did they come from? Overwhelmingly, they were outsiders, fresh marginal area soldiers. This comes as no surprise; they had the same strengths as the marginal area soldiers who had originally conquered the government area and founded the dynasty: hardiness and good 63  Muq, 1:372. 64  Ibid.

An Explanation Of Military Slavery

825

health, martial skills and inclination. As marginal area men, the recruits had not experienced government and so had not withdrawn from it; and their way of life required that they be hardy and martial. When they joined a government army (or administration) they had no interests of their own in the kingdom, little sympathy for the subject peoples, and they were readily convinced to associate themselves with the existing outsider ruling elite. In a key passage, Ibn Khaldūn notes that the ruler must seek out fresh marginal area soldiers when his own become unreliable: In a dynasty affected by senility as the result of luxury and rest, it sometimes happens that the ruler chooses helpers and partisans from groups not related to (the ruling dynasty but) used to toughness. He uses (these people) as an army which will be better able to suffer the hardships of war, hunger, and privation. This could prove a cure for the senility of the dynasty….65 While marginal area soldiers dominated these recruited troops, insiders also participated (in contrast to the act of founding a dynasty, when insiders were usually entirely absent). Government area soldiers had drawbacks (they were inclined to be alienated and they lacked the best martial skills) but they also had two advantages, their abundance and low cost.66 The rulers made occasional use of them as auxiliary forces, as emergency reinforcements, as a counterbalance to the marginal area soldiers, or even as a source of revenue (for sometimes they paid to join the armed forces).67 In general, however, insiders played a small role. All rulers, even those who lacked easy access, had to acquire marginal area soldiers, sometimes at great cost, for no army was complete without them.68 The recruitment process did not end there, however; as fresh marginal area troops spent time in the government area, they too became unreliable. Once again, some lost their martial skills, others lost their loyalty. As fresh troops went stale, new ones had to be recruited in an unending cycle which lasted as long as the dynasty did. Ibn Khaldūn describes how the new friends become as demanding as the old relatives: 65  Muq, 1:342. 66  For example, C. E. Bosworth, “The Armies of the Ṣaffārids,” Bulletin of the School of Oriental and African Studies 31 (1968): 554; H. Inalcik, “The Sociopolitical Effects of the Diffusion of Fire-arms in the Middle East,” WTS, pp. 196–97. 67  D. Ayalon, “The Muslim City,” p. 326. 68  Ayalon, “Aspects,” p. 206.

826

Pipes

In taking [new friends] on as followers and [in] replacing his old clients and original followers by them, the ruler is motivated by the fact that (his old clients and followers) have become overbearing. They show little obedience to him. They look at him in the same way his own tribe and relatives do.69 Thus, martial decay and political unruliness rendered a perpetual search for new sources of military manpower obligatory. Rulers relying on marginal area soldiers found themselves constantly seeking out new soldiers; they had to find these or else watch their armies grow debilitated and fractious. Each pre-modern Islamicate government which depended largely on marginal area soldiers—and this included nearly every one—had to seek out alternate sources of manpower within two (or so) generations of its establishment and then continue to do so until the time of its demise. The recruitment of marginal area soldiers completed their hold over Islamicate armies. They both brought dynasties to power and staffed the armies thereafter. As the founding soldiers became unreliable, the government imported new marginal area soldiers to replace them. Thus, unreliability did not signal the exit of marginal area soldiers but only changed the cast. After the first spasm of conquest, importation might go on for generations. The influxes of marginal area soldiers marked the great shifts of power in politics: the first created the dynasty, the second maintained it, and the third destroyed it, beginning the cycle over again. The first forced its way in, violently imposing its will from the outside, while the second was imported and acquired power peacefully from within. Marginal area soldiers entered once as conquering warriors, the second time as paid soldiers. This explanation for outsider domination of Islamicate public life reverses the usual understanding. Other analyses, which assume that Muslim subjects sought power but failed to attain it because outsiders took over,70 focus attention on the wrong actors. I argue that the key lies in this fact: insiders relinquished power and allowed outsiders to take it up. Outsider soldiers and rulers were always in the wings in premodern times; the striking feature in Islamdom was not the presence of marginal area soldiers but the absence of insiders. The peculiarity lay with the insiders who did not rule more often. Had Muslim insiders wished to participate, no one could have prevented them; the fact that they did not points to their relinquishment of power, which then made rule by outsiders possible. Adherence to the ideals of Islam did not cause an 69  Muq, 1:376. 70  E.g. Goitein, p. 218; Grunebaum, Islam, p. 132.

An Explanation Of Military Slavery

827

abundance of outsiders in authority but a dearth of insiders who wished to assume it.

The Benefits of Military Slavery

When a ruler decided to recruit soldiers from marginal areas to replenish his armed forces, he had three means through which to acquire them: alliance, pay, or enslavement. Each of these methods had its advantages and drawbacks, yet I shall argue that in two crucial respects, rulers preferred slaves to either mercenaries or allies because they were acquired more easily and more thoroughly controlled. In these benefits lay the raison d’être of military slavery. Ibn Khaldūn refers to them in his eulogy of military slavery: This status of slavery is indeed a blessing … [the slaves] embrace Islam with the determination of true believers, while retaining their nomadic virtues which are undefiled by vile nature, unmixed with the filth of lustful pleasures, unmarred by the habits of civilisation, with their youthful strength unshattered by excesses of luxury.71 The advantages of enslavement will be illustrated by comparing slaves with the two alternate types of marginal area soldiers fighting for governments: mercenaries and allies. Acquisition A government could procure slaves more easily than it could either mercenaries or allies. It might purchase, capture, abduct, or steal a slave, but obviously not a free man. A slave could be compelled to join the army; mercenaries had to be enticed to serve, and allies had to find it expedient. The slave was subject to more active and flexible means of persuasion. By recruiting him through enslavement, the ruler did not have to wait until cooperative marginal area soldiers appeared on the scene,72 a common predicament of governments that did not enslave soldiers (such as Byzantium and China). In contrast to the limited conditions under which mercenaries or allies agreed to fight, slaves came according to circumstance: some arrived as tribute; others as merchandise, booty, contraband, or stolen property.

71  ʿIbar 5:371 (adopted from Ayalon, “Yāsa” C1. 119; see Appendix 2). 72  Hrbek, p. 545.

828

Pipes

Military slaves were usually procured as children and this, too, facilitated their acquisition. While mercenaries and allies could only be found among friendly peoples, children could be abducted or captured from enemies in wartime and, through training made into faithful soldiers. The pool of potential slaves could be many times larger than that of free recruits. Enslavement gave access to a wide variety of nationalities and this provided the army with a beneficial diversity of troops, as they often brought with them the special skills of their own peoples:73 This multiplicity of ethnic backgrounds and skills contributed directly to the flexibility and tactical power of Islamicate armies. Though mercenaries and allies, too, could have varied origins, the ruler had much less control over where they came from. Further, by enslaving his recruits, the Muslim ruler could choose his soldiers man for man. Mercenaries and allies arrived in corps or tribes and fought as a group; slaves came singly. The government could select its slaves carefully, which was not possible with free marginal area soldiers. This selectivity made possible a higher standard of quality for each soldier in slave armies. Along with these benefits, the procurement of military slaves also involve some special problems. As a dynasty declined in strength, it could no longer acquire its slaves through force (raiding, warfare, and so forth) but had to purchase them. Yet, as the dynasty weakened its resources diminished, so this expense grew ever more burdensome. The Mamluks of Egypt could neither reduce their dependence on new recruits nor acquire them inexpensively; the need to buy slaves contributed significantly to the economic troubles of the government.74 The distance slaves usually had to travel from their homelands to their country of service and the fragility of the supply lines could also cause problems.75 As slaves usually came from remote regions, enemy forces could easily block access to them. Abbasid dependence on the Tahirids to send them slave children reduced Abbasid control in northern Iran and added to the Tahirids’ strength. On the other hand, why the Ottomans did not cut off the supply of

73  Bosworth, Ghaznavids, p. 108. 74  Ayalon, “Aspects,” p. 208; E. Ashtor, “Recent Research on Levantine Trade,” Journal of European Economic History 1 (1973): 201; idem., Les Metaux precieux (Paris, 1971), pp. 99– 108; R. Lopez, H. Miskimin, and A. Udovitch, “England to Egypt, 1350–1500: Long-term Trends and Long-distance Trade,” Studies in the Economic History of the Middle East, ed. M. A. Cook (London, 1970), p. 127. (Boaz Shoshon gave me the references to Ashtor’s works.) 75  Ayalon, “Aspects,” pp. 207–08; Hrbek, pp. 552–53.

An Explanation Of Military Slavery

829

recruits from the Black Sea and the Caucasus areas to the Mamluk kingdom, once those two powers had become antagonistic, continues to baffle historians. The expense and the distance over which military slaves traveled presented two drawbacks peculiar to slave soldiers, but only in times of decline; these problems were not envisioned when a ruler founded a military slave corps in the second generation or so of the dynasty. Control Newly recruited soldiers from marginal areas entered as total aliens and outsiders, without affiliation either to the ruling powers or to the members of the polity. How could their master bind them to himself and his dynasty? Mercenaries or allies retained their own loyalties, culture, and methods of warfare, but slaves were subjected to reorientation; the government could secure their loyalty, impose cultural changes, and fit their military skills to the needs of the army. Mercenaries and allies imposed their fickle loyalties on the ruler. They could always desert and they constantly threatened to mutiny: “an ally was always a potential threat to independence”76 and a mercenary even more so. Since these troops often constituted the most powerful force in the kingdom, little could prevent them from becoming an unmanageable and destructive element, indifferent to any allegiance that blocked the way to booty. If dissatisfied with their plunder from warfare, they readily attacked their own employer or ally. Military slavery provided a means by which to control marginal area soldiers. Unlike mercenaries and allies, slaves could be compelled to undergo changes in identity; these changes were effected through the complementary processes of deracination, isolation, and indoctrination. Deracination exposed slaves to loneliness and new relationships; isolation furthered their susceptibility; and indoctrination transformed their personalities. Unlike mercenaries and allies, who usually arrived in tribal units and stayed in them, retaining their old loyalties, slaves came as individuals and had to build new attachments. Deprived of their own people, these soldiers had to accept the new affiliations offered them. The military slave corps developed into a substitute tribe and replaced the true kinship group in many instances. The adoption of a master’s nisba (kinship name) reflected the need for a new, albeit spurious filiation.77

76  Smail, p. 70. 77  Forand, “Relation,” pp. 62–63; D. Ayalon, “Names, Titles and ‘Nisbas’ of the Mamluks,” Israel Oriental Studies 5 (1975): 213–19; Lévi-Provençal, p. 106 n. 10.

830

Pipes

The master also isolated his slaves. He took them from their homelands to a strange country and cut them off from the rest of the society. They had no choice but to accept the ties provided them and to become loyal to him. They developed close relations with their comrades, all of whom shared the same predicament. Geographic isolation also reduced the possibility that a marginal area soldier would ever have to fight his own people by taking him far away from them. Combat against conationals strained the loyalty even of a military slave, though many examples of their loyalty to their masters in such situations can be found.78 Military slavery made indoctrination possible. Whereas mercenaries and allies arrived fully developed and resisted changes in their personalities and loyalties, military slaves came as children, unformed and susceptible to reorientation. Years of careful schooling imbued them with lifelong attachments to the Islamic religion, their master, his dynasty, and their comrades-in-arms. The master exerted continuous pressure on the slave recruits to relinquish their prior allegiances in favor of himself. Enslavement made possible the extended period of gestation which changed identities. Ibn Khaldūn explains: “When a people with group feeling train a people of another descent or enslave slaves and mawlas, they enter into close contact with them…. These mawlas and trained persons share in [their patrons’] group feeling and take it on as if it were their own.79 Mercenaries and allies invariably had concerns outside of their military service. They had family, kinsmen, herds, farms, and so forth, to which they devoted attention and from which they were loath to be long separated. These interests required time and conflicted with their service to the ruler. Slaves, on the contrary, could be made to live in isolation from the rest of society. Not only could they be prevented from earning outside income, but they could also be kept unmarried; surely the ruler could not compel anyone but his own slave not to marry. In return for receiving their entire income in salary from the ruler, the slaves served him all year round as a standing army. Military slaves fell far more completely under the cultural influence of the polity than their free rivals. In training they learned the customs, religion, culture, and language of the dynasty; this proved to be of great importance, for unless they were made to feel part of the dynasty, they could always turn against it. Military slaves never did this; they had become assimilated to the dynasty itself. They were part of the ruling elite, not its lackeys. When they 78  CHI, 4:162; P. Wittek, “Türkentum und Islam, I” Archiv für Sozialwis-senschaftund Sozialpolitik 59 (1928): 517; and see the many examples in chapter 6. 79  Muq, 1:276 (my translation from the Arabic, 1:245).

An Explanation Of Military Slavery

831

revolted, they did not attack the polity as such but the individuals in charge; if successful, they usurped the government from within. This acculturation did not prevent them, however, from preying on the populace of the polity; they engaged in this pursuit, as did all members of the ruling elite. But acculturation made them part of the government, so they could not attack the polity itself, though its populace remained their victims. The training process was the linchpin in the whole institution of military slavery. It established a slave’s character by instilling military skills, discipline, and an understanding of command structures. The years of training distinguished the military slave and determined his future career. He entered training a young and isolated boy and emerged a highly skilled, disciplined, and well-connected soldier. The mercenary or ally, not compelled to undergo training, usually lacked these important qualities. Military slaves received training first in the martial arts. Whereas mercenaries and allies showed impatience with the introduction of new techniques, slaves would learn new methods of fighting.80 Their servile status and their youth combined to force them to accept these changes. Marginal area soldiers often arrived in the polity brimming with independent spirit and unfamiliar with chains of command, yet governments could not tolerate the chaotic nature of tribal warfare, so they forced their slaves to learn discipline. Through military training, the natural courage and hardiness of marginal area soldiers was combined with the organization, techniques, and discipline of government armies. The slaves emerged superbly accomplished in the martial arts and fully integrated into an organized army. The main drawback of the training program lay in the time it required; while mercenaries and allies came fully prepared for battle, military slaves had to be acquired and trained far in advance of their employment. They could be properly used only in the context of long-range planning.81 Besides bringing military power to the dynasty as a whole, military slaves provided the ruler with political henchmen. While serving the army against external enemies, they also supported the ruler against internal rivals. Although complementary, these two functions were not identical. As agents, they were totally beholden to the ruler, devoted to him, and lacking any trace of envy; no better agents could be found. Mercenaries and allies could not reliably provide this personal service. Muslim leaders could choose to recruit alien marginal area soldiers in other ways, but these entailed more difficulties. For example, the Mughals did not 80  For some details on this, see H. Rabie, “The Training of the Mamlūk Färis.” 81  Ayalon, “Aspects,” p. 208.

832

Pipes

have many military slaves; instead, they employed Hindus as palace guards, used lower-class men as infantry, and recruited cavalry from Iran and Central Asia by offering especially high salaries.82 However, the Mughals often had problems acquiring these troops and keeping their loyalty. Given the Muslims’ need for alien marginal area soldiers, military slavery brought with it several advantages over other methods of organization; the slaves’ numbers, quality, and youth assured the best material to work with; their isolation, training, and indoctrination assured fine and loyal soldiers. Noting the advantages of military slaves, we should not find their military role in the millennium 820–1850 ce. So puzzling. The institution of military slavery was not an accident, a legalism, or a fluke, but a successful adaptation to the specific Islamicate need to acquire and control alien soldiers from marginal areas. However odd it may seem to our eyes, Muslim rulers reaped real military benefits from the enslavement of recruits.

Nonmilitary Factors

Besides providing Muslims with a mechanism for acquiring and controlling soldiers from outside marginal areas, military slavery had to fit into the general patterns of Islamicate life. Military slavery had a military rationale, but nonmilitary factors also contributed to its success and proliferation. These nonmilitary factors did not explain the purpose of the institution, but they did help to form an environment that was conducive to military slavery. Had the military needs been unchanged but other factors unsuitable, the system might never have come into being. Beyond fulfilling a function, military slavery also fitted into Islamicate society. 1. Slaves fought in battle from the first moments of Islam. They were already participating in Muhammad’s battles in sizable numbers.83 This fact must have made their later use in warfare more acceptable, though no explicit mention of the Muhammadan precedent has come to my attention. Since every act of Muhammad’s has attracted close scrutiny, it seems probable that the slaves who fought with him remained ever after in the Islamic consciousness. This may have given sanction to the use of slaves in warfare: “If the Prophet did so, we may too.” I cannot account for the striking absence of this justification from the sources, however. 82  On military slaves; see W. Irvine, The Army of the Indian Moghuls (London, 1903), p. 11; on others, Qureshi, pp. 131–33, 124. 83  Details are in chapter 4.

An Explanation Of Military Slavery

833

2. Both Islamic law and the events of the early period combined to give slaves in Islamdom an exceptionally high status. The Qurʾān and subsequent Sharīʿa regulations guaranteed the human dignity of the slave, especially one who was a Muslim. Although inferior on this earth, his equal value in relation to God found frequent emphasis. Few slaves have enjoyed as high a legal and social status as those in Islamdom. The fact of being a slave in itself meant less among Muslims than among other peoples, for the slave’s potential was greater than elsewhere. This dignified treatment of slaves made it more likely that they would hold important positions. In reality, too, the first experiences of slaves in Islam were extremely good. For a variety of reasons (to be discussed in chapter 6), they filled important positions from the first years of Islam and participated extensively in its early civilization. Where else have “slaves made important contributions and exerted strong influences in the realms of politics and public administration, warfare, religion, arts and crafts, music, poetry, grammar, and learning in general?”84 Slaves pervaded every area of activity, and their names ranked among the most celebrated of early Islamicate history and culture. 3. The Sharīʿa does not clearly permit or prohibit slaves from fighting. AlʿAdawī, a 7th/13th-century author of a mirror for princes, points to universal agreement among the jurists that a soldier must meet four qualifications (he must be adult, Muslim, whole of body, and mentally sound) and disagreement on a fifth: whether or not he must also be free. Al-ʿAdawī reports that Abū Ḥanīfa requires him to be free and ash-Shāfiʿī does not.85 Ash-Shāfiʿī makes slave participation optional in offensive jihād warfare while requiring it in defensive warfare.86 Ash-Shaybānī asserts that a slave may not fight without his master’s permission except in an emergency; as-Sarakhsī adds that the master may not force the slave to fight except in an emergency.87 A consensus seems to exist against using slaves in normal circumstances but in favor of their assistance in times of trouble. It is noteworthy that there was no clear condemnation of slaves fighting in normal circumstances; further, the restraints on slaves fighting appear to reflect a concern with property value more than with honor, 84  Haas, p. i. Haas considers every mawla a slave, despite his disclaimer (p. 7, n. 1). 85  al-ʿAdawi, p. 156. 86   Umm, 6:85. 87  as-Sarakhsī, pp. 182–83, 199–200; Khadduri, pp. 84–87 lists seven major qualifications, one of which—economic dependence—excludes slaves; Thābit, pp. 93–99, lists freedom as one of the five prerequisites for soldiers but recognizes the controversy on this point. The argument over the validity of a slave’s amān (see note 52 to chap. 2) tacitly acknowledges the presence of slaves in the Islamicate armies.

834

Pipes

the discussions discourage slaves from exposing themselves to needless danger and thus jeopardizing their master’s property.88 4. The strong and continuing relationship encouraged by the Sharīʿa between a freed slave and his patron (the walāʾ of manumission),89 created favorable conditions for military slavery. Slaves owned by a Muslim remained by their master long after emancipation, providing him with allegiance and service in return for protection and patronage. Few slaves returned to their countries of origin after becoming free; the majority stood by their patron.90 This continuing voluntary relationship increased the likelihood of men of slave origins serving their masters in important positions. 5. “Le Monde musulman est une civilisation esclavagiste.”91 The Muslims never seem to have suffered from a shortage of slaves; through many centuries, over many areas, they owned them in abundant numbers. The Muslims imported slaves from all the surrounding areas—especially sub-Saharan Africa, Europe, Central Asia, and India—while hardly ever falling victim to slavery themselves.92 The presence of slaves was, of course, a fundamental precondition of military slavery; and their existence in massive numbers was conducive to the system, since owners sought useful ways in which to employ the slaves at hand. Certainly this appears to have been the case for those rulers who received vast numbers of slaves as war captives (see p. 97). 6. An abundance of servile women in courts may have made rulers more predisposed to military slavery. A Muslim man could keep any number of concubines; this practice filled the rulers’ palaces with legions of slavewomen. Many a Muslim ruler found himself surrounded by females of slave origins, including his mother and wives; why not, then, include male slaves in his entourage too? 7. Islam summons an unparalleled allegiance from its adherents. It has the power to change loyalties; it routinely transforms a person’s whole orientation. This made it easier for Muslims than for other peoples to place outsiders in responsible positions. Islam shows astonishing power in capturing 88  as-Sarakhsī, p. 908. This brings to mind a US Civil War anecdote, when a master “told his slave to look after his property, and when some shot fell nearby, the Negro fled, and when the master reprimanded him he said, ‘Massa, you told me to take good care of your property, and dis property (placing his hand on his breast), is worf $1500’” (CW, p. 278). 89  My “Mawlas” article discusses this at length. 90  For an exception, see KM, p. 203. 91  M. Lombard, L’Islam dans sa premiere grandeur (Paris, 1971), p. 194. 92  Ibid., pp. 194–202; G. Hambly, “Islamic Slavery: An Overview” (mimeographed: n.p., n.d.), pp. 21–25.

An Explanation Of Military Slavery

835

a person’s primary identity; converts usually view themselves as Muslims first and as members of an ethnic group or region second. While many differences distinguished Muslims from each other, they felt their common bond more than those differences. So powerful was this affiliation that, for example, the Turks “submerged their identity in Islam” to the point where, in premodern times, they had almost no self-conception other than as Muslims.93 Similarly, for Berbers in North Africa, “notions between tribal and Islamic were hazy and of doubtful social significance.”94 The power of the Islamic bond gave a master confidence that a slave who converted to Islam really did transfer his allegiance to the Muslims. When nonMuslims attempted to bring about comparable changes in identity, they did not meet with the same success; both the Byzantines and the Crusaders tried to effect a similar transformation in marginal area soldiers;95 but Christianity does not bind its adherents as tightly as Islam does. Besides the strength of the Islamic ties, the readiness of Muslims to accept new converts to Islam—even slaves—contributed to their enthusiasm for Islam. The manumitted slave became a full-fledged member of the community. A Muslim could not be a complete stranger to other Muslims; even if a foreigner or a slave, he shared a vital bond with them. Other societies generally showed less openness to outsiders and greater reluctance to allow them to hold sensitive positions. 8. The Sharīʿa ruling that one-fifth of all booty (ghanīma) be-given to the government derives from the Qurʾān: “When you have taken booty, one-fifth belongs to God, the Prophet, his near of kin, to orphans, the poor, and way­ farers.”96 While the jurists disagreed about the details of this arrangement,97 they did concur that this one-fifth (Arabic: khums; Persian: penchik) of the 93  B. Lewis, The Emergence of Modern Turkey (London, 1965), p. 13; S. A. Zenkovsky, PanTurkism and Islam in Russia (Cambridge, Mass., 1967), p. 8. Islam had the very special power (shared, perhaps, only by the United States) to change not only the identity of a new member of its community but also that of his ancestors. 94  Gellner, p. 15. 95  On Byzantium, see Vryonis, “Byzantine and Turkish Societies,” pp. 125–52, comparing Byzantine and Islamicate attempts; also C. Cahen, “D̲ j̲aysh” in EI2. On the Crusaders, see Smail, pp. 111–12, discussing the Turcopoles, the closest the Crusaders came to military slavery. The Crusaders had an acute shortage of military manpower (pp. 88–97); all the Muslim armies they faced relied on military slaves; yet despite the need and the availability of a model, the Western Christians appear never to have tried to imitate the Muslims.  The Ethiopians perhaps tried military slavery too (see n. 12 to chapter 1). 96  Qurʾān, 8:41. 97  Khadduri, pp. 121–22.

836

Pipes

booty belonged to the community as a whole as represented by its government and personified by its ruler. The fifth generally included all forms of wealth acquired by force of war, both property (movable and not) and persons— namely, slaves.98 Until now, no study has been made of the application of the khums regulation throughout Islamicate history; but wherever it held firm, it could supply the ruler with massive numbers of slaves.99 Extensive conquests could provide the Muslim ruler with thousands of slaves; in such instances, it made no sense for them all to serve him personally or work in his household. They could better labor in the fields or in industry, but the ruler might prefer to use them to support his regime by enrolling them into the army or the administration. The khums ruling probably made the number of slaves available to a Muslim ruler greater than those available to other rulers, and their being at his disposal then encouraged their use in public life. 9. The Qurʾān prescribes the inheritance laws for Muslims and leaves little room for individual discretion. By assuring a fairly equal division of properties among members of a family, it prohibits the concentration of wealth for more than a generation or two. Muslims regularly applied these laws, despite widespread maneuvering around them and some outstanding long-lasting families.100 Polygamy exacerbated this diffusion of wealth, for rich men tended to have large families, and so the share of each individual heir was often quite small. No matter how rich the grandfather, two generations later his grandchildren usually received modest inheritances. Unable to concentrate their resources, great families did not often gain a hold on important positions. Islamicate society knew no rigid social boundaries but was a constant flux of persons and families; as a result, there was always room for new blood. Only in religious officialdom, where special skills (not money) formed the basis of power, does one find consistent hereditary patterns. Because no hereditary aristocracy dominated military and political offices, they were open to social climbers—including slaves. Beyond keeping the positions open, this fluidity in social rank cut down on birthrights. The daintiness of born aristocrats in Hindu India or feudal Europe 98  In the Ottoman case, for example, this factor is commonly ascribed considerable importance: Káldy-Nagy, pp. 164–65; S. Shaw, History of the Ottoman Empire and Modern Turkey, vol. 1: Empire of the Gazis (Cambridge, 1976), s.v. pençik in the index. 99  az-Zabīdī, p. 96. 100  In Iran especially, the aristocracy exploited every loophole and even committed incest in order to preserve a fortune. For the description of an enduring aristocracy, see R. W. Bulliet, The Patricians of Nishapur (Cambridge, Mass., 1972).

An Explanation Of Military Slavery

837

derived in large part from their assured superiority. They never allowed their ranks to be filled by persons of slave origins; but in Islamicate society, social fortunes were too transient for a person’s birth to play too much of a role in his career.101 10. The Sharīʿa allows few ties of subordination other than slavery, so Muslim leaders fully exploited the slave-master bond. As a result, slaves attained diverse positions which elsewhere would have gone to persons with different statuses. Also, the disjointed nature of Islamicate society and its weak institutional ties meant that only a bond as strong as slavery insured durable relations. Ties with free men, be they relatives, kinsmen, mercenaries, allies, religious brethren, or notables, often did not hold. 11. A strict separation of the sexes may also have favored the use of military slaves. The exclusion of women from public life and the stringent sexual code between men and women encouraged widespread homosexuality among men and made it generally accepted in Islamicate society. In this context, military slavery benefited the leaders by supplying them with a pool of subservient men available for sexual relations. Better yet, the young recruits offered a choice of “beardless ones.” Two practices reflected this homosexual element: the fact that eunuchs trained the recruits102 and the meteoric careers of male slaves—often military slaves—to whom the ruler or a high official took a liking.103 Presumably sexual bonds, not just good looks, accounted for this favor. Conclusion This chapter has suggested why military slavery existed and why it existed only in Islamdom. It provided a most effective way for governments to acquire and control marginal area soldiers; and Muslim rulers developed it because their own subjects withdrew from public life and would not fight for them. Several distinctive Islamicate elements have emerged here: (1) the extent and seriousness of the gap between ideals and reality in public life; (2) the withdrawal of insiders from politics and warfare; (3) the reliance on outsiders 101  For an entirely different approach leading to this conclusion, see Halpern, p. 70. Also Hodgson, 1:320, 2:117–18. 102  Ayalon, L’Esclavage, pp. 14–15; idem, “Eunuchs,” p. 268. 103  Ayāz, the slave and favorite of Maḥmūd of Ghazna, was probably the most renowned catamite in Islamicate history (P. Hardy, “Ayāz,” in EI2); also Sīmā ad-Dimashqī, al-Muʿtaṣim’s favorite (Aghānī 18:93); and al-Khālidiyaynī, p. 49; Mez, p. 358. Even a eunuch gained fame as a sultan’s lover! (S. Digby, “Kāfūr, Mālik,” in EI2).

838

Pipes

to staff administrations and armies; (4) the development of an institution— military slavery—to bring in and control soldiers from the outside. The truly unusual feature of military slavery has little to do with the enslavement of soldiers; it lies, rather, in the fact that a cultural rationale lies behind this phenomenon. The existence of military slavery has almost nothing to do with material circumstances (geographic, economic, social, political, technical, and so on), but follows from the needs inherent in Islamicate civilization. In contrast to other forms of military recruitment—tribal levies, mercenary enrollment, militia conscription, or universal service—this one existed in only one civilization and occurred there almost universally. To the best of my knowledge, no other method of military organization has comparable connections to a single civilization. Herein lies the purpose of this analysis: to show that Islam, a religion, must be taken into account when assessing public affairs in premodern Islamdom. Some issues, no doubt, can be fully understood without reference to Islam; it would be foolish to claim that nothing significant took place in this period without Islam affecting it. But it would also be unwise to ignore Islam; for although it is a religion, it has affected many nonreligious aspects of life among Muslims, even in areas—military organization a case in point—which ostensibly have nothing to do with it. If the connection drawn here between military slavery and Islam is valid, then much else in the public life of Muslims occurred within an Islamicate matrix. The reaction of the reader to this assertion may well be influenced by his own background. Of the three great monotheistic religions, Christianity differs from Judaism and Islam in this regard. To look at Islam from a Christian point of view can make it more difficult to understand how Islam permeated the lives of its adherents; whereas Christianity is a system of belief, Judaism and Islam are all-inclusive ways of life. Above all, the Jew fulfills his religious obligations by maintaining an allinclusive law, both the biblical and the expanded rabbinic Halakha. Although Jesus himself seems not to have rejected this law, Saint Paul did: “If righteousness comes by law, then Christ died for nothing.”104 Among other consequences, this belief had the effect of withdrawing religion from many aspects of life for Christians; the emphasis on faith made the many regulations which defined the everyday life of the practicing Jew unnecessary. While in time the Church built up a body of regulations, the Canon Law, which in some ways approached those of Judaism, these never touched on spheres of life to the extent that the Halakha did, in respect both to quantity and to basic principles. 104  Galatians 2:21.

An Explanation Of Military Slavery

839

The Protestant Reformation included a passionate denunciation of the regulations which had proliferated in Catholicism; whatever their other differences, the Protestants agreed on reemphasizing the paramountcy of faith. Moreover, there arose among Lutherans a theory often alluded to as that of “two kingdoms,” one of God and the other of the state. An indirect consequence of this theory was the shrinking of religion from being all-inclusive (as was the Jewish and Muslim view) into a limited domain of worship, personal piety, and the like. This reduction in the scope of religion has had a far-reaching impact, not only on Protestants, but also on Catholics, Orthodox, and even on Jews (since the late eighteenth century) and Muslims (since the late nineteenth century). The Protestant ethos has become the modern way; modern man restricts religion to belief and feeling; he is not familiar with the pervasiveness and absolute authority of religious injunctions.105 Thus, Catholics have come to view confession in terms of its psychological value; Jews point out the medical advantages of circumcision; and Muslims argue the social benefits of banning liquor. The modern person has nearly forgotten the original religious mandate that established these requirements; he is therefore psychologically distant from the notion that religion could both directly and indirectly affect nearly all aspects of life; he is unprepared for this because he looks at religion in a Protestant way. Although in both Judaism and Islam faith is not ignored, both center around a code of law. The Muslim fulfills his religious obligations by maintaining the laws of the Sharīʿa. Not only does it resemble the Halakha in general purpose; many of its specific points derive from Jewish prototypes. For example, Islamic requirements for slaughtering animals for food closely parallel Jewish ones. Other regulations reflect their Jewish origins by being the opposite; if Jews almost never speak the name of the Lord, Muslims do so incessantly. The important point is clear: Jews and Muslims in premodern times lived in societies molded by divinely inspired regulations in a way far different than did Christians. Many modern persons, regardless of religion, are skeptical about the notion of Islamicate patterns permeating a society; traditional Jews and Muslims have little trouble understanding and accepting this fact. The dubious reader may inquire of himself if he does not look upon Islam in too modern a manner.

105  Our alienation from comprehensive religion can be compared to our alienation from animals. The modern person no longer lives among them as his ancestors did, but encounters them only as pets, pests, clothing, and food.

840

Pipes

This excursus into comparative religions prompts a final remark: from a broader perspective, Islam extended many Jewish patterns and values. It took them to new geographical areas (sub-Saharan Africa, Southeast Asia) and into new areas of life (warfare, politics) which the Jews, because of their smaller dispersal and numbers, hardly explored.

War, Servitude, and the Imperial Household: A Study of Palace Women in the Chola Empire Daud Ali

Problems in Historiography

The nature and history of the forms of servitude and slavery in early medieval India are still not clearly understood. This lack of understanding is only partly due to the sources, which are fragmentary, episodic, and often obscure. More often, formulaic theories of society and colonial definitions of productive work have weighted down the interpretation of available sources. Perhaps unsurprisingly, the most important discussions have taken place around the question of agrestic labor. Some of the initial assessments in this field stretch back to the nineteenth century, when colonial officials sought to understand the land-holding and tenurial systems which they encountered across the subcontinent. In south India, the situation seemed inconsistent, though varieties of agrestic servitude which bore the characteristics of both serfdom and slavery were palpable and widespread.1 Such debates formed the backdrop for the rise in south India of historical studies of the great lowland empires like the Cholas, though the results have been equally uncertain. Though the corpus of some ten thousand inscriptions2 so far recorded from the Chola period (c. 950–1250 ce) in south India (see map 3) constitute an Source: Ali, Daud, “War, Servitude, and the Imperial Household: A Study of Palace Women in the Chola Empire,” in Indrani Chatterjee and Richard M. Eaton (eds.), Slavery & South Asian History, Bloomington: Indiana University Press, 2006, 44–62. © Indiana University Press, Reprinted with permission of Indiana University Press. * I would like to thank Sascha Ebeling, A. Murugaiyan, Y. Subbarayalu, S. Swaminathan, and especially Leslie Orr and P. Sundaram for discussions in person and through e-mail of various readings of inscriptions. Leslie Orr was particularly generous in sharing her extensive knowledge of women in Tamil inscriptions. I would also like to thank Indrani Chatterjee for a very careful reading of the essay, which has benefited greatly from her suggestions. 1  See Dharma Kumar, Land and Caste in South India: Agricultural Labour in the Madras Presidency during the Nineteenth Century (Cambridge: Cambridge University Press, 1965), 49–55. 2  Some three thousand inscriptions have been published in various journals, including South Indian Inscriptions (henceforth sii), Epigraphia Indica (ei), and Travancore Archaeological

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_027

842

Ali

undoubtedly rich bequest when compared with their north Indian counterparts, they by no means present us with an entirely clear picture of contemporary social dynamics. Like nearly all the lithic records from this period, Chola inscriptions are chiefly concerned with temple affairs. Most often they record various types of donations for the establishment and maintenance of services to gods in temples, and do not seek to portray social or economic relationships not directly relevant to such affairs in any systematic fashion. The names, titles, corporate bodies, and financial and revenue arrangements they record give us only an incidental and fragmentary picture of the wider dynamics of Chola society. The records also typically relate to elite groups of the social order, and information relating to the lower echelons of society is indirect at best. Despite these difficulties, most historians now concur in characterizing the lowest ranks of cultivators, mostly paraiyars and pallars, as substantially “unfree” by the tenth century. Though the precise conditions of this unfreedom are unclear, some of these cultivators appear to have been “owned” by landholders and wealthier cultivators. Yet it has not been definitely resolved whether such ownership constituted “slavery” or “bondedness.”3 At stake in this debate, following Marx, has been the “feudal” or “Asiatic” character of the early medieval state. It appears that this debate has been conducted in a partial vacuum, since there is little evidence for the Chola period which can yield a sense of the scale of unfree labor—the degree to which productive processes as a whole relied on people of such status.4 The evidence of “ownership” of cultivators leads us to infer that such cultivating serfs or slaves were probably found in areas of intensive irrigated agriculture. However, the majority of references to slavery are not connected to the transfer of men and women between landowners. The majority refer to the dedication, sale, or gift, by men of various stations, of slaves to temples— the famous tevaratiyar (or devadasis) who have been the subject of so much interest in modern times. In many cases, these men and women seem to have been the household or personal servants of the donors in question rather than Series (tas), while the remaining unpublished inscriptions have been noted in the Annual Report on Indian Epigraphy (henceforth are), published by the Archaeological Survey of India, and are available in estampage or transcription form only. Citations of epigraphical publications will give the volume number followed by the inscription number or, for citations of the Annual Report on Indian Epigraphy, the inscription number followed by the year. 3  See Y. Subbarayalu, Studies in Cola History (Chennai: Surabhi Pathippakam, 2001), 92; and Kesavan Veluthat, The Political Structure of Early Medieval South India (Bombay: Orient Longman, 1993), 230–31. 4  James Heitzman, Gifts of Power: Lordship in an Early Indian State (New Delhi: Oxford University Press, 1997), 69.

War, Servitude, And The Imperial Household

843

agrestic laborers granted by landlords.5 This points to an important lacuna in the historiography to date, a comparative indifference to the apparently widespread practice of domestic servitude and slavery in Chola times. Not only did many tevaratiyar probably come from domestic contexts, but they performed similar labors in the temple setting—service and attendance functions, and menial tasks. The temple was, after all, a palatial residence, the house of a god. Many temple servants mentioned in Chola period inscriptions, then, might be usefully considered a special and highly ranked subset of the larger category of domestic and personal servants whose prevalence must have been extensive. This wider world of social relations has yet to receive its due attention. Since nineteenth-century conceptions of productive work largely excluded domestic labor from the analytical field, historiography shaped by such notions has also been disinclined to study the complexity of domestic servitude in medieval south Indian society. In the case of tevaratiyar, this indifference has been compounded by the fact that such men and women in many cases seemed to enjoy certain economic and social privileges. Inscriptional references, for example, often record their donations to temples, implying the ability to own property. Their association with religious institutions has predictably created a historiography of “sacrality” and “auspiciousness,” obscuring, as Leslie Orr has forcefully demonstrated, both the mundane and quotidian nature of their work as well as the considerable agency they exercised.6 Orr’s careful study of temple women, as well as recent scholarship on domestic slavery in other historical contexts, suggests that historians of medieval south India revisit their records with new questions and sensitivities. One of the most intractable problems in assessing the status of servile men and women in medieval India is terminological. Though it might be easy to identify “slaves,” or atiyar, a term derived from the word for “foot” (ati)—meaning literally those “at the feet” of another—in the inscriptional records, it is far from certain what such status actually meant.7 This is in part because of the great importance placed on the language of servitude to represent the reciprocal bonds of affiliation between superiors and inferiors in medieval south India. Relations of hierarchy during Chola times were expressed and experienced through vocabularies of submission, devotion, and intimacy drawn from an existing terminology of servitude and slavery. In courtly and religious contexts the language of servitude was imbricated with power dynamics in particularly 5  Ibid., 68. 6  Leslie C. Orr, Donors, Devotees, and Daughters of God: Temple Women in Medieval Tamilnadu (New York: Oxford University Press, 2000). 7  See the useful discussion in Orr, Donors, Devotees, and Daughters, 52–53.

844

Ali

complex ways. To assume a posture of exaggerated deference, submission, or even self-deprecation often served to draw attention to the humility, devotion, and even eminence of a man or woman. Slave-status terminology, therefore, is often particularly difficult to understand, as it functioned both as a marker of subordination and as a language of distinction. Moreover, the widespread use of this and related terminologies of subordination to refer to people of varying constraints and privileges suggests a continuum of servitude and lordship rather than clearly opposed categories.

The Organization of Labor in the Imperial Household

This chapter seeks to explore a group of people, mostly women, and the establishments to which they belonged, which are both poorly understood and rarely treated in the historiography of the period—the lower echelons of imperial servants known as pentattis who were organized into palace establishments, or velams, attached to the Chola household. The working assumption of this chapter, which will be borne out and refined through an examination of the evidence, is that palace women, like their temple counterparts, despite possessing certain privileges, constituted a category of unfree labor. But once again, the treatment of these women in studies of medieval south India has been minimal. The eminent historian Nilakanta Sastri, who, to his credit, took more time than most to consider their status, concluded that “we have to look upon this crowd of personal servants as in the enjoyment of a fair competence in return for generally very light work; the status of the members of the velams was perhaps that of a not unpleasant servitude to which the less sensitive among them might have reconciled themselves in a short time.”8 Such characterizations of domestic labor, perhaps comfortably familiar to historians of the nineteenth century, have remained unchallenged interpretive obstacles to coming to grips with the nature of both the royal household and the gendered labor relations which sustained it. To convey a sense of the scale of this work force, we may turn to the Chinese traveler Chau Ju-Kua, who claims to have visited south India sometime in the late twelfth or early thirteenth century, and reports that the Chola king retained for his escort some ten thousand “dancing girls,” three thousand

8  K. A. Nilakanta Sastri, The Colas, 2nd ed. (Madras: University of Madras, 1955), 450.

War, Servitude, And The Imperial Household

845

of whom attended him at a time, in rotation.9 Such observations are legion among travelers to other parts of India in medieval times, and would seem to be corroborated by a variety of indigenous sources. Architectural and sumptuary manuals rank the majesty of kings by the size of their palaces, the splendor of their regalia, and the number of women attached to their households. Indeed, like the gods who ruled in the heavens, the lordly personages who walked the earth were to be surrounded by a luminous female presence. The imperial king, according to the architectural treatise Manasara, was to have an entourage of millions of women.10 Women like these, so close to imperial authority, yet devoid of the benefit of name or lineage, litter medieval sources. In medieval poetry they appear as voluptuous nayikas thronging the streets of the royal city and seeking the attention of the king as he moves in procession. Such representations certainly had pronounced symbolic dimensions relating to both kingship and religious experience, which have been effectively drawn out by scholars like David Shulman and Velcheru Narayana Rao.11 Yet this scholarship has occasionally dismissed enquiries into the sociological world which formed the backdrop to these exquisite literary depictions.12 As Cynthia Talbot points out, inscriptions too are literary artefacts, and deserve the same regard as other literary sources.13 Noboru Karashima has even suggested that scholars deploy other faculties to decipher the “whisperings” of inscriptions on the walls of medieval temples.14 Only then can we understand the scores of women who appear in inscriptions associated with the Chola household—a category of servants Leslie Orr has usefully called “palace women.”15 9  Reported originally in the account called Ling-wai-tai-ta. See Chau Ju-Kua: His Work on the Chinese and Arab Trade in the Twelfth and Thirteenth Centuries Entitled Chu-fan-chi, trans. F. Hirth and W. W. Rockhill (Taipei: Cheng Wen Publishing Co., 1970), 95, 100. 10  See Manasara, ed. and trans. P. K. Acharya (Delhi: Oriental Reprints, 1980), 41.10–43. 11  See David Shulman, The King and the Clown in South Indian Myth and Poetry (Princeton, N.J.: Princeton University Press, 1985), esp. 303–39; and Velcheru Narayana Rao, David Shulman, and Sanjay Subrahmanyam, Symbols of Substance: Court and State in Nayaka Period Tamilnadu (Delhi: Oxford University Press, 1992), 57 ff. 12  See the remarks of Narayana Rao, Shulman, and Subrahmanyam on the utility of inscriptions as historical sources, in Symbols of Substance, 31. 13  Cynthia Talbot, Precolonial India in Practice: Society, Region, and Identity in Medieval Andhra (New York: Oxford University Press, 2001), 8–9. 14  Noboru Karashima, “Whispering of Inscriptions,” in Kenneth R. Hall, ed., Structure and Society in Early South India: Essays in Honour of Noboru Karashima (Delhi: Oxford University Press, 2002), 56–57. 15  See Orr, Donors, Devotees, and Daughters, 40–41.

846

Ali

From what we can glean from inscriptions, the Chola kings had complex extended households. In order to secure political alliances with powerful local lineages, they took numerous wives, many of whom probably lived in separate quarters within the royal palace complex.16 Beyond members of the royal family, there were functionaries at the royal court, whose number and variety grew significantly in the eleventh century. The person of the king was also surrounded by bodyguards as well as hereditary military retainers who were attached to the royal household. Then there were special “intimates” (masc. anukkan, fem. anukki), “concubines” (poki), and “friends” (saciva) who enjoyed elevated status and sometimes lordly titles. These people often appear as members of the heterogeneously staffed retinues or entourages (parivarams) of high-ranking family members as they toured the kingdom making religious donations and such. Such people may have been drawn from a wider cadre of domestic servants of lower rank, the men of which were known for a brief period in the eleventh century as “servants,” literally “work-sons” (pani makan).17 Many of these lower-ranking men and women who served the royal household, mentioned in inscriptions, are said to be attached to institutions called velams, a problematic and insufficiently understood term, rendered variously by scholars, but which most likely denoted a collection of palace servants.18 The handful of references to velams in Tamil inscriptions (and the term’s single literary occurrence) all date to the Chola period and the term does not seem to have any prevalence in the literature and epigraphy of either previous or successive royal houses in south India. Chola inscriptions furnish the names of some twenty-seven different velams between the reigns of Parantaka I (907–55) and Kulottunka I (1070–1120). The royal household was often served by many velams at any given time, and some seem to have been quartered outside of the palace itself. The nomenclature of velams gives us some clue as to their organization and function. The titles of many velams clearly derived from the names 16  At least two kings in the tenth century, Parantaka I (907–55) and Uttamacola (979–85), are known to have had at least ten wives each. Notable are the marriages secured with the Malaiyamans of Miladu, the Malavars of Maladu, and the Irukkuvels of Kodambalur, as well as the Vallavaraiyar and Paluvettaraiyar lineages. See George Spencer, “Ties that Bind: Royal Marriage Alliance in the Chola Period,” Proceedings of the Fourth International Symposium on Asian Studies (Hong Kong: Asian Research Service, 1982), 717–36. 17  Subbarayalu, Studies in Cola History, 107. 18   See T. N. Subramaniam, “Glossary,” in South Indian Temple Inscriptions (Madras: Government Oriental Manuscript Library, 1957), vol. 3, pt. 2, s.v. velam; Nilakanta Sastri, The Colas, 449–51; B. Venkataraman, Rajarajesvaram: The Pinnacle of Chola Art (Madras: Mudgala Trust, 1985), 251; Heitzman, Gifts of Power, 149; and most usefully, Y. Subbarayalu, Tamil Kalvettuc Collakarati (Chennai: Canti Cathana, 2002), s.v. velam.

War, Servitude, And The Imperial Household

847

of members of the royal family and the various titles which they bore, but the royal names in their titles do not seem to provide any certain indication of the persons to whom the services of their members were directed. The royal titles of velams seem to have had a commemorative rather than functional sense. During the reign of Rajaraja I (985–1014), inscriptions at the Chola capital of Tanjavur and elsewhere record the names of no fewer than nine velams bearing different royal titles (most of which were associated with Rajaraja himself), and at least one velam named after his queen Pancavanmateviyar.19 Beyond such titulary names, some velams were designated by terms suggesting more hierarchical and functional roles. Inscriptions of widely differing dates mention the “old” (palaiya) and “big” (periya) velams, and a single record uses the term alvar, referring to junior member(s) of the household.20 The titles of at least two velams suggest they were named after a particular conquest by the king, and may have indicated the regional origin of their inhabitants—a significant fact in considering the personnel of these palace establishments, below.21 Various other titles suggest particular functions in the daily routine of the king and royal family. Numerous records across the tenth and eleventh centuries, for example, mention velams connected with the royal bath (mancanam, tirumancanam), and occasionally records also refer to velams connected with the handling and supply of ceremonial vessels (tiruparikalam) and the performance of protective evening rites (antikappu).22 Many of these more specialized velams are referred to as composed of “select” (terinta) personnel, suggesting a hierarchy among the personnel who filled the velams. 19  For velams referring to titles of the royal family during the reign of Rajaraja I, see sii 2.94; sii 2.95; are 340 of 1927; are 62 of 1928; sii 7.678; sii 26.669; sii 23.342; sii 23.278; and sii 23.356. 20  For mention of the Palaiya velam, see sii 3.204 and sii 5.697. For the Periya velam, see sii 17.480; sii 19.10; are 106 of 1925; sii 22.291; are 401 of 1921; and are 424 of 1962. For the Alvar velam, see sii 23.45; and on the significance of the term alvar, see Nilakanta Sastri, The Colas, 142. 21  These are the Pandi velam and the Rajaraja terinda Pandi tirumancanattar velam, both mentioned at Tanjavur (sii 2.94; sii 2.95), which I take not only as commemorating Chola victories over the rival Pandya dynasty but as indicating that they were at least partly staffed by women captured during these campaigns. Though it is possible (though unlikely) that “Pandi,” in the title “Pandi velam,” may have referred to a title borne by Rajaraja I, this would seem to be precluded by the title “Rajaraja terinda Pandi tirumancanattar velam,” which would be more logically rendered as the “velam known as Rajaraja of those from the Pandya kingdom selected for the ceremonial bath.” 22  SII 19.193; sii 22.27; sii 678; sii 2.94 and 95; are 323 of 1927; are 142 of 1919; and are 212 of 1911.

848

Ali

The individuals most typically associated with velams were identified as pentattis and, to a lesser extent, kaikkolars. The term kaikkolar, literally “of strong arms,” referred to a member of a class of apparently hereditary military retainers who often resided near the palace and who formed an integral part of the Chola armies. While a small number of velam references mention kaikkolars, the incidence of kaikkolars in Chola inscriptions is far larger, as they often appear as part of the elite military coteries of the Chola kings, being selected for staffing personal entourages (parivarams) and perhaps acting as bodyguards, but at the very least constituting, on occasion (but not always), part of an inner core of permanent troops around the royal household.23 It may be that these identifications were not mutually exclusive, and that all kaikkolars were in fact associated with velams, but the evidence is far from certain in this regard.24 Even in early Chola inscriptions, these men possessed a strong corporate identity, which, like that of other hereditary military groups, was transformed into a caste status in post-Chola times.25 Female members of velams are almost always described as pentattis, a difficult word because of a long historical sedimentation and multiple usage. Though used in contemporary Tamil (in informal contexts) to mean “wife,” in medieval times the term denoted more particularly a woman of generally servile status and most usually one connected with the royal palace in some capacity. It literally meant a woman “ruled” or a “slave/servile woman,” but the generic nature of the vocabulary of servility mentioned above prevents any conclusions about the status of such women on the basis of terminology alone.26 In at least one reference a pentatti seems to also be identified with a term which less ambiguously denoted a “slave” (atiyal), but for the same reasons this may mean very little.27 Not all inscriptional references to pentattis 23  See P. Sundaram, “Chola and Other Armies—Organization,” in S. N. Prasad, ed., Historical Perspectives of Warfare in India: Some Morale and Matériel Determinants (New Delhi: Centre for Studies in Civilizations, 2002), 190–91. 24  L. Thyagarajan has argued this position, and consequently interpreted the velam as a military barracks. See Pierre Pichard et al., Vingt ans après Tanjavur, Gangaikondacholapuram, 2 vols. (Paris: Ecole française d’Extrême-Orient, 1994), 1:184. 25  Later the term kaikkolar came to denote a caste of weavers tracing their origin to military groups of ancient times, but in Chola inscriptions the term referred exclusively to military units. See Nilakanta Sastri, The Colas, 454 and passim; also Heitzman, Gifts of Power, 150; and Subbarayalu, Studies in Cola History, 108. 26  The word is formed by adding the suffix -al (a verbal root meaning “to rule, receive, control, or maintain,” or a noun meaning “man, servant, slave, laborer”) to the noun pen, meaning “woman.” See Orr, Donors, Devotees, and Daughters, 212n5. 27  The spelling is irregular and probably a scribal error; see sii 23.278.

War, Servitude, And The Imperial Household

849

mention velams. They are sometimes identified simply by the term or are designated as pentattis of particular members of the royal family or their entourages (parivaram).28 In two cases pentattis are mentioned as cooks for the imperial household.29 It may be the case that pentatti referred to a more generic category of female servant who took on a variety of roles at the Chola court, but the majority of records do place them within velams, and I will assume that this was their typical affiliation.30

War and the Gendering of Imperial Households

The inscriptions are for the most part silent as to the origins and identities of pentattis, but we are not entirely bereft of information bearing on this question. Exceptional evidence comes from the late twelfth century, when an inscriptional eulogy (meykkirtti) of the Chola king Kulottunka III (1178– 1218), describing his protracted struggles with Vira Pandya of Madurai, boasts that, having beaten the Pandya king on the battlefield, he “caused the best of his women to enter his velam.”31 A later version of the same eulogy adds that the Chola king caused Vira Pandya’s “young queen” to enter his velam (matakkotiyai ve[lam] erri).32 Though we have no further record of who these women might have been or their subsequent fate, this record makes it clear that at least one source of women in the velam was warfare. To this extent, the status of pentatti was born from the violence of war and from an initial but decisive severance from natal kin. Tamil literature patronized by the Chola court also provides some relevant, if indirect, information on velams and the identities of the women within them. 28  See are 88 of 1928 and 69 of 1926; and sii 5.700. 29   s ii 19.98; and tas 1.8.1. 30  It is difficult to know whether the absence of velam affiliations for pentattis was accidental or indicates a wider usage of the term. In her survey of women in Chola inscriptions, Leslie Orr has identified forty-one instances of pentattis in Chola inscriptions, with twenty-six (over 60 percent) mentioning some association with a velam (Orr, Donors, Devotees, and Daughters, 212n5). My own data suggests a slightly higher number. We also have an instance in the tenth century of two adjacent inscriptions at Utaiyarkuti (South Arcot) commemorating gifts probably made on the same occasion by two women, one identified as “singing” in the Periya velam at Tanjavur and the other simply as a pentatti, sii 19.10, 12. The apparently accidental omission of the term pentatti in the first inscription may parallel the omission of a velam affiliation in the latter. 31   s ii 22.42; also are 254 of 1925. 32   s ii 3.88.

850

Ali

The Kalinkattupparani, a poem celebrating the victory of Kulottunka I over the Eastern Ganga king Anantavarman Codaganga, composed by the court poet Cayankontar at the beginning of the twelfth century, devotes its first substantive canto to a long entreaty to the women of the royal city to “open their doors” to the returning Chola army. A string of verses are specifically addressed to women of the velam: You gentle women of the Pandya country, the flag of which bears the fish, who have entered the velam after running through the wilderness in tears, open your doors! Women of Tulunatu, women of Malainatu, give tribute to Kulottunka, from the land of the splashing waters, open the doors to your houses … You Karnata women, approaching uttering a confusing mix of beautiful words in Tamil and Vaduku in your gentle speech, open your doors!33 It is clear enough here that the women of the velam were captives of war, taken in the military campaigns of the Chola kings.34 The practice of capturing or forcibly abducting women as part of annual military campaigns in rival kingdoms is well attested. “Seizing women” is a conventional boast in the royal eulogies that cover the walls of hundreds of Chola-period temples in south India. Medieval south Indian armies traveled with large trains of supporting personnel—children of the royal family, slaves, and various ranks of male and female servants—and if they were routed these people fell into the victor’s hands. Chola meykkirttis are often quite particular about the fate of the women captured from their rivals.35 In some instances they were “defaced”—their noses cut off36—and in others they were simply added to the Chola household as war booty. Typical is a meykkirtti which boasts that, having defeated the Chalukya king Ahavamalla, the Chola king captured a 33   Kalinkattupparani, ed. Pe. Palanivela Pillai (Chennai: South India Saiva Siddhanta Publishing Works, 1961), vv. 40–43. 34  See C. Ilavaracu, Parani Ilakkiyankal (Chidambaram: Manivacakar Nulakam, 1978), 53–54. Ilavaracu contends that among the women of the royal capital depicted in the second canto were contingents of women received as tribute from subordinate rulers or captured during wars in other kingdoms. 35  See especially the meykkirttis of Rajendra I’s successors, particularly Rajendra II (1052– 64), sii 22.80; and Virarajendra (1063–70), ei 21.38. 36  The meykkirttis of Virarajendra often boast that the Cholas decapitated the dead body of the Chalukya mahadandanayaka Chamundaraja and severed the nose of his only daughter, the beautiful Nagalai, the queen of Irugaiyan, ei 21.38; sii 3.20. An inscription of Rajadhiraja (1018–54) says the same with regard to a Pandyan king’s mother, sii 3.28.

War, Servitude, And The Imperial Household

851

number of famous elephants and well-bred horses and camels, the banner of the Chalukya army and other insignia of royalty, and in addition the illustrious queens Sattiyavai and Sangappai, as well as lesser wives, along with crowds of women left on the battlefield by the retreating king.37 The forcible abduction of women of lesser rank from the countryside is also attested. A famous Chalukya inscription dated 1007, at the village of Hottur in contemporary Dharwar district, describes the campaign of a large Chola army from the perspective of the other side, as it “ravaged the whole country, murdering women, children and Brahmans, seizing women (pendiram pididu) and overthrowing the order of castes.”38 Though the claims of some meykkirttis seem exaggerated, the repeated emphasis on capture, and the often very specific details given in the accounts, cannot be ignored. Medieval ula (poetic descriptions of royal processions), which focus on the reactions of crowds of women of different ages to the king as he moves through the royal city, also have some significance for locating the origins of these women. One of these describes the crowds of women who lined the royal street during processions as women of diverse birth descended from women brought to the Chola country as war booty from various regions and settled by the king in areas assigned to them by royal order.39 Though the word velam is not used in the poem, the passage clearly invokes these establishments. This arrangement is broadly corroborated by a contemporary Sanskrit text on architecture composed in south India, Mayamata, which recommends that the royal street (rajavithi) should be lined with mansions (malika), where the king’s retinue was to reside—these may have been the residences of the velams we hear about in the Tanjavur inscriptions, which extended into the suburbs around the palace, and past which royal processions would have been made, forming the immediate setting of the ula as a poetic genre.40 Closely related to capture through war was the receipt of women as tribute from subordinate kings, a practice which was not unknown elsewhere in early medieval India. The appearance of women in descriptions of battlefield loot suggests that lower-ranking servants were incorporated into the 37   s ii 3.29. 38   e i 16.1 1a. 39  “Iracaracacolanula,” in Muvarula, ed. U. V. Caminathaiyar (Chennai: U. V. C minathaiyar Nul Nilaiyam, 1992), vv. 70–79. Also see the remarks of G. Thirumavalavan, Political, Social, and Cultural History of the Cholas as Gleaned from the Ula Literature (Thiruvathipuram: Ezhilagam Publishers, 1991), 134–35. 40   Mayamata: Traité sanskrit d’architecture, ed. and trans. Bruno Dagens (Pondichéry: Institut français indologie, 1970), 10.74–75.

852

Ali

transactional economy of gifts and tributes between men of lordly rank. The Kalinkattupparani, which portrays the magnificence of the assembled Chola court, describes among the annual tribute gifts required of subordinate kings “the forehead bands [pattam] of women who are rightfully yours.”41 At least one pentatti, a woman who served one of Rajaraja’s queens, Pancavanmateviyar, is known to have been in the Kotantarama velam at Tanjavur, and her name, Vanakovaraiyan Porkali, identifies her with a lineage known to be subordinate to the Cholas. It is possible that this woman, who was clearly not a wife but a servant, was presented as a gift to the Chola family. It is quite likely that the women of the velams entered the palace as a result of the military campaigns of Chola kings or as tribute received from subordinate lords. Indeed, the period when the greatest number of velams are mentioned in the epigraphic record coincides with the successful military campaigns and territorial expansion of the Chola empire in the eleventh century under Rajaraja I and his son Rajendra I. The extent to which velams were filled with captive and tribute women remains an open question, however, as does whether there were other methods of recruitment into velams. Though pentattis shared a number of characteristics with temple women, there is no existing epigraphic evidence of the presentation, sale, or purchase of pentattis to or by the royal court. It is likely that the mechanisms of war and tribute obviated the need to acquire personnel through purchase. It is also true that such transactions, being irrelevant to temple affairs, would not have appeared in the inscriptional record, and must be inferred indirectly. Important in this regard is an inscription dated to the reign of Kulottunka I, which records the transfer of a number of temple slaves found in the king’s retinue back to the temple authorities. This “return” involved removing the king’s mark from their bodies and rebranding them with a trident, sign of their proper master.42 The practice of branding, encountered in a significant handful of temple inscriptions relating to the temple’s own slaves, seems to have been a royal practice as well. Thus the confusion of the women in this case offers an extremely suggestive glimpse into such lives. While the lower-ranking women among palace servants clearly overlapped with their temple counterparts, such similarities might in turn have provided avenues for “escape” from one holder to another. However, the temple authorities’ reclamation of “their” slaves from the royal entourage also suggests a degree of surveillance of servants, and the fact that marks on servile bodies pronounced the “names” and identities of holders first. 41   Kalinkattupparani, v. 336. 42   a re 141 of 1922, discussed in Nilakanta Sastri, The Colas, 556.

War, Servitude, And The Imperial Household

853

Obviously, such inscriptions disable overly simplistic attributions of gender and identity that assume that the personal names and affiliations of such servants are transparent. For instance, some pentatti names appear to be “male.”43 If body-marks spoke of the identities of holders, then there is no necessary identification between the bearer of a name and the persona of the bearer. The “male” named might be the holder of the pentatti. However, we could also infer that the bearer of the name, while embodied as a woman, performed tasks that were identified as masculine and might therefore have merited a “socially” male name. Or we could see such names as having been given to androgynous beings in those societies. At the very least, such names ask for a refinement of modern dimorphic perspectives when we engage in gendering the pasts of older societies. Pentattis often had compound names, like Katan Accatevi, Kallici Uttamata, or Kari Catti, which raises the question of the significance of each name segment. It is possible that in some cases, paternal names were prefixed to proper names, as has been common practice in south India during more recent times, and, some have argued, was equally so in Chola times.44 The evidence, however, is far from conclusive. In some compound names, the first element is either clearly feminine or incorporates the name of a deity or place. There is, therefore, no consistent naming pattern among pentattis. Nevertheless, the names of pentattis do reveal clear evidence of stratification among the women of the velam. Some women may have entered the velam with elevated rank—women of high standing given as tribute or captured from the chiefly and royal families of subordinate lineages. For instance, Vanakovaraiyan Porkali, a servant of the Chola queen Pancavanmateviyar in the Kotantarama velam, seems to have retained the title of an earlier affiliation with a feudatory family.45 Others appeared to have acquired titles of distinction once within the velam, like the pentatti Tevayan Pulalakkan of the Kilai (Kilanatikal?) velam, who was also known as “Crest-Jewel of the Earth” (Avanisikhamani), and Cattan Ramadevi of the Rajendracola Periya velam, who had the title “Ruby of the Sacred Jambu Fruit” (Tirunnavalmanikkam).46 The term manikkam, or “ruby,” seems to have been a title incorporated into a 43  As in the case of a pentatti of the Melai velam with the single name Raman, are 340 of 1927. It is also possible that men were given the title pentatti (personal communication of P. Sundaram), but this is difficult to confirm. 44  On the practices of naming in medieval south Indian inscriptions, see Noboru Karashima, Y. Subbarayalu, and Toru Matsui’s introduction to their A Concordance to the Names in the Cola Inscriptions (Madurai: Sarvodaya Ilakkiya Pannai, 1979), 1:ix–lxvii. 45   s ii 23.278. 46   s ii 3.201; and are 424 of 1962.

854

Ali

number of personal names of pentattis in the eleventh century, and was even more widespread among temple women.47 Its significance, however, is uncertain. Clearer in meaning are the titles anukki and poki, which referred to an intimate or concubine of a member of the royal family. In some cases such women appear as donors in inscriptions without any association with velams. So we have two records at different temples noting the gift of lamps by one Nankai Cattaparemanar, a “concubine” (pokiyar) of the king Aditya I (871–901), and very extensive gifts made by one Nakkan Paravai, identified as an “intimate” (anukki) of Rajendra I—neither of which mentions pentatti status or affiliation with a velam.48 On the other hand, Cattan Ramadevi, the pentatti mentioned above, has the title of “intimate” (anukki), suggesting that women of the velam were eligible for such distinction.49 We also have cases of the children of favorites appearing among the women of the velam.50 What these host of titles suggests is that there was considerable scope for favor and advancement among the women of the velam. Such advancement can be assessed along many grids at once. The most obvious is that of their provisioning. Like their temple counterparts, velam pentattis received maintenance and appear to have accumulated modest amounts of wealth. It was not unusual, either in northern parts of the subcontinent in this period or in Chola contexts, for slaves and others of servile status attached to powerful households to enjoy specific, albeit circumscribed, privileges and material support. In one record, for example, we have a slave (atiyal) donated to the temple with a maintenance grant ( jivanam) for picking flowers in a temple garden.51 Once again, while temple inscriptions have much to say about tevaratiyar, they provide little information about maintenance arrangements within the palace. The inscriptions relating to pentattis indicate only that they were able to donate wealth to temples. As is the case with tevaratiyar, most inscriptions relating to pentattis record gifts of money for the establishment of perpetual lamps in temples. How pentattis acquired this wealth is unclear—it may have been given to them as maintenance (clothes, jewelry, daily or yearly allowances for food or land) by the royal household, or as support from children or others, or it may have accompanied them into servitude. 47   s ii 22.2911 and are 323 of 1927. On the occurrence of the title among tevaratiyar, see Orr, Donors, Devotees, and Daughters, 148. 48   s ii 23.219; sii 23.247; and sii 4.223. 49   a re 401 of 1921. 50  One At[n]ukkan Mahamalli, of the Utaiyar Iracakesari velam, appears as the mother of a kaikkolar donor, sii 26.669. 51   s ii 22.141, discussed along with other cases by Orr, Donors, Devotees, and Daughters, 127.

War, Servitude, And The Imperial Household

855

Given this larger silence, we might gainfully interpret an unusual record (dating from the time of Rajaraja) of the intervention of the palace kaikkolars on behalf of the royal household. The record outlines their attempt to collect payment on a loan that a powerful Brahman member of the local assembly (sabha) had taken from a pentatti belonging to the Mancanattar velam.52 The debtor had died and his son had inherited his property, but neither he nor other relatives were willing to settle the debt. So the sabha took over the debtor’s property and sold the land to the temple to repay the debt. Throughout the entirety of this process the pentatti seems to have been a passive conduit of wealth actually held by the royal palace. Clearly great caution needs to be exercised in attributing control over the use and disposal of such wealth to the pentattis. Some inscriptions suggest that the gifts by pentattis were made on behalf of brothers, sisters, mothers, or daughters (but never fathers, who remain entirely absent from the limited kin world of the pentatti). An eleventh-century inscription from Tiruvisalur, for example, records the gift of temple lamps by a pentatti of the Alvar velam for the benefit of her brother Vicchadiran Kunjiramalla and her mother Karoki.53 Most records, however, do not mention merit-recipients. The great majority of economic and fiscal activities associated with pentattis have been recorded in the central region of the Chola empire (Tanjavur, Trichurappalli, and South Arcot districts). They appear to follow the patterns of, or are made in conjunction with, queens and other members of the royal household. They rarely appear in isolation. So during the reign of Aditya II we find separate records of some ten donations by various members of the royal household, including the queen Sembiyanmateviyar, two pentattis, and members of five different kaikkolar units, apparently donated on the same occasion, to the Anantesvarasvamin temple in Utaiyarkuti, South Arcot.54 At this stage, we can only speculate that the merit accruing from the “gifts” and donations they made to the temples may have been meant to be shared with the other, more important, donors who sometimes accompanied them. Given the often collective and ritualized nature of religious gifts and the potentially more complex property relations which may have underlain them, it would be hasty to assume that gifts donated by pentattis merely represented personally accumulated wealth which could be disposed of at will.

52   s ii 22.27. 53   s ii 23.45. See also are 212 of 1911 and sii 17.480. 54   s ii 19.10 ff.

856

Ali

The Reproduction of Devotion

Another way to understand the stratification and advancement of pentattis is by exploring the multiple relationships around which pentatti lives were structured. Literary texts of the period portray palace women of all ages. In fact, the court literature pays acute attention to the ages of women who formed part of the retinues of Chola kings. The ula genre mentioned above is structured around the calibration of a woman’s life into seven distinct stages (elu paruva makalir) from age five to forty-four.55 These “stages,” which are not attested in pre-Chola literature, notably exclude both infancy and old age, but cross the threshold of sexual maturity. Almost half of these, from the age of five to fourteen, concern prepubescent girls, and the others relate to sexually mature women. This is explicitly recognized in the poetic descriptions of the first three types of young girls (petai, petumpai, and mankai), which focus extensively on the minutiae of their physical maturation and their impending sexualization. The significant point about these stages is that they seem to reflect a categorization of women not simply as sexual objects, but as potential sexual objects—suggesting an unstated transactional economy in which women were scrutinized and valued in relation to the specific degree of their maturity. Judging from this sexualization of bodies and languages used to describe “palace women” in court literature, the dynamics of favor for women of the velam appear to have been closely tied to sexual relationships with members of the royal household. However, the contribution of the literary genres in situating the advancement of female servants through the palace or temple bureaucracy in the language of “favor” and personal “intimacy” should not be overlooked at the same time. Take, for instance, the language of intimacy deployed by the ula genre in its treatment of the various generations of women in the king’s retinue. The petai (ages five to eight), close to the side of their mothers, their bodies “like fresh and tender leaves” and eyes like drops of honey, are not attracted to the king’s physical beauty, but become fixated on the garland he wears.56 The garland (a sign of favor to be bestowed by the king) is 55  They are petal (5–8), petumpai (8–11), mankai (12–14), matantai (15–19), arivai (20–25), terivai (25–32), and perilam pen (32–44). This division is unknown in earlier literature, and is elaborated in the pattiyal poetic texts, the earliest of which is the anonymous Panniruppattiyal, published with the commentary of K. R. Kovintaraca Mutaliyar (Tinnevelly: South India Saiva Siddhanta Publishing Works, 1963), vv. 131–38. While the specific age spans of each category differ in different pattiyal texts, they all embody the logic of progression from sexual immaturity to maturity. 56  “Vikkiramacolanula,” in Muvarula, vv. 112 ff.

War, Servitude, And The Imperial Household

857

extremely difficult to obtain, explain their mothers, and the girls break into tears as the king moves off down the street. With each successive stage, as the sexual maturity of the girl advances, so does the transformation of her desire, until the entire range of emotions of love in separation are experienced by the retinue of waiting women of different ages. Though there is certainly an artificial and formulaic element to the genre, it nevertheless envisions the feminine transition into maturity as at once a romantic fantasy and a quest for intimacy and favor. While these horizons of possibility are suggested by the literati for the young pentatti, inscriptional evidence does not corroborate the favor of such intimacy in any precise form. None of the children of pentattis we find in inscriptions mention their paternity, and we thus have no way of understanding who their fathers were. This absence is significant in itself, for the mothers seem never to have had the benefit of legalized marriage. The corresponding ambiguity and invisibility of the sexual relationships of the pentatti makes tracing conjugal and natal kin of these women particularly fraught. Leslie Orr has even suggested that the names of many tevaratiyar in the inscriptions may not indicate kinship affiliations of any sort.57 Name segments aside, in no instances are pentattis explicitly identified as either the daughters or wives of men, even when their own children are mentioned. This absence of male kin remains in stark contrast to the contemporary identification of women from the high-status castes and points markedly to the disconnection from natal and conjugal kin experienced by pentattis.58 As though in compensation, and underlining the importance of being “related” at all times, many temple slaves were designated “daughters of god” and male palace servants as “work sons” (pani makan). Whatever the connotations of this quasi-kin terminology, the men and women of the velam did not define themselves through normative natal and conjugal kin relations. The only kin definitively mentioned in connection with pentattis are their children. In a number of records donors identify themselves as the sons or daughters of pentattis, in some cases making gifts on behalf of or with their mothers.59 When children do appear in inscriptions, references to their paternity are usually absent. The fate of these children in relation to the complex 57  Orr’s findings suggest that most women did not incorporate the names of their fathers into their own. Donors, Devotees, and Daughters, 147, 248n16. 58  Ibid., 154–55. 59  See sii 17.530 for a gift by the children (makkal) of a pentatti of the Kilanatikal velam; sii 17.480 for a gift by a man for his elder sister, who is identified as the daughter of a pentatti of the Periya velam; and are 63 and 64 of 1928 for a joint gift by a pentatti and her daughter, both residents of the Sivapadasekhara terinta tirumancanattar velam.

858

Ali

of institutions surrounding the court is not certain. Young males may have entered the ranks of kaikkolars, as we have two instances of kaikkolars either making grants on behalf of velam women or identifying themselves as children of a velam pentatti.60 This might fit well with the evidence we have from separate records, mentioned above, which identify kaikkolars themselves as members of velams.61 But most inscriptional references to kaikkolars do not mention velam affiliations, but instead note their connection to units within the Chola army—which of course does not preclude their originating from pentattis. But kaikkolars couid rise to higher ranks within the court hierarchy and we have records suggesting they were sometimes attached to the personal retinues of various members of the royal household and in some cases enjoyed service tenures known as virabhoga by order of the king.62 Whatever the case, there seems to have been some special relationship between kaikkolars and velam women. They often appear together as donors in clusters of inscriptions at key temples. This seemingly special connection between velam pentattis and kaikkolar units also seems to fit with the Kalinkattupparani’s request to the women of the velam that they “open the doors” to the returning soldiery of the Chola army. It is thus likely that one important function of the velam, as Nilakanta Sastri hinted at long ago, was to supply the Chola court with a regular source of loyal military retainers whose links (and loyalties) were confined entirely to the extended household and its master.63 As for the female children of pentattis, they are even more invisible in the epigraphic record, as we come across only occasional references to daughters or sisters in connection with pentattis. But this hardly rules out the very likely possibility that the daughters of pentattis were born into the same condition as their mothers.64 The comparative frequency of pentattis themselves in the 60  See sii 26.669 for a kaikkolar whose mother was a pentatti in the Iracakesari velam; and sii 23.356 for a kaikkolar making gifts for various women in the Kotanta velam. The editors have assumed in the latter case that the women were relatives of the donor, though the inscription does not indicate what relation the donor may have had to the women of the velam. 61  We have one tenth-century record of a kaikkolar in the velam of Perumanatikal Mateviyar, sii 4.536; and four twelfth-century records, sii 5.697; sii 5.698; sii 23.279; and sii 23.281. 62  See are 69 and 72 of 1926, where the village of Kulottunkacolanallur is designated as virabhoga for kaikkolars from Merka-natu who were of lesser (cirudanam) rank and served in the palace at Gangaikondacholapuram. 63  P. Sundaram has also suggested that velams were training establishments for Chola military personnel. “Chola and Other Armies—Organization,” 191. 64  This is in fact the implication of the Iracaracacolanula, which speaks of generations of women from different lands living by the order of the king.

War, Servitude, And The Imperial Household

859

inscriptions across many generations may account for the absence of explicitly identified daughters. The preponderance of females in the velam may not be so surprising, for if pentatti mothers did not identify themselves with their natal or conjugal kin, being instead identified entirely with the extended household of the royal family, their daughters may have been expected to follow a similar path, while sons would have had the option of pursuing military service and its wider network of affiliations. Conclusion In bringing this review of the evidence to a close I would like to summarize a few features of the world of velams and pentattis. Velams, collections of palace servants organized around the daily needs of members of the Chola royal household, seem to have been large institutions which flourished during periods of Chola imperial expansion. They were staffed largely, but not entirely, by women, who were at least in part acquired through tribute and capture during war, but perhaps also through purchase or birth into servile status. The most notable aspect of the identities of these women, known as pentattis, as they appear in the epigraphic record, is a lack of identification with natal and conjugal kin groups. Velams thus seem to have been institutions of those who had been natally alienated through violence or some other process, a very significant fact considering the importance of kin and lineage among the ruling elite in medieval south India. But this may have had its advantages for the royal household, as it made velams ideal reproductive pools for developing a cadre of men loyal to the royal family alone, a fact which may explain their ongoing links with kaikkolars. But natal alienation, combined with their apparently multigenerational presence at the Chola court, their sexualization in contemporary literary sources, their association with tribute and war capture, and even their enjoyment of privileges by virtue of attachment to the royal household, all suggest that the women of the velams were far from voluntary members of the Chola imperial service, as they are sometimes portrayed. While the sources do not allow us to draw more subtle distinctions between pentattis and temple slaves, with whom they seem to have had so much in common, the evidence suggests that both sorts of women participated in a common world of servitude whose lineaments are far wider and more significant than scholars have been willing to admit.

The Modern World: 1450–1900

∵

The Geography of Slaving in the Early Modern Mediterranean, 1500–1800 Robert Davis The Mediterranean basin has in the past half-century become one of our most recognizable historiographic touchstones, a key organizing principle behind an entire school of thought that holds dear the primacy of economic and cultural connectedness in human affairs. The binaries of Henri Pirenne’s Mediterranean and the broad-brush geodeterminism of Fernand Braudel’s have recently been rendered both more subtle and more particular in the work of Peregrine Horden and Nicholas Purcell, which has, in its turn, brought forth a new generation of Mediterraneanists who still find, whatever their disagreements, that the sea and its shores provide a productive methodological framework for their work. The Mediterranean Principle has indeed become exportable, as its proponents have been busy for some time finding mediterraneans in the Sea of Japan, the Caribbean, the Baltic, in full-fledged oceans like the Atlantic, and even in dry seas, such as the Sahara or Gobi deserts.1 A major, even defining, element of the Mediterranean model has been that of exchange—the networks of long or (increasingly) short-distance commerce that scholars have used to delineate the cultural basin and link its disparate margins. In this essay, I would like to map out a particular form of commercial traffic that was both vast and especially characteristic of the Mediterranean— the trade in human beings. Slave trade across the Mediterranean sometimes seems as ancient as human settlement around the sea itself and has over the millennia passed through so many permutations and entangled so many societies that it undoubtedly merits a study of Braudellian scale in its own right. The practice of slaving around the Mediterranean rose and fell over the long S ource: Davis, Robert, “The Geography of Slaving in the Early Modern Mediterranean, 1500– 1800,” Journal of Medieval and Early Modern Studies, 37(1) (2007): 57–74. Copyright, 2007, Duke University Press. All rights reserved. Republished by permission of the copyright holder, Duke University Press. www.dukeupress.edu. 1  Fernand Braudel, The Mediterranean and the Mediterranean World in the Age of Philip II (New York: Harper and Row, 1972); Peregrine Horden and Nicholas Purcell, The Corrupting Sea: A Study of Mediterranean History (Oxford: Blackwell, 2001); David Abulafia, “Mediterraneans,” in William V. Harris, ed., Rethinking the Mediterranean (Oxford: Oxford University Press, 2005), 64–93; but cf. Michael Herzfeld, “Practical Mediterraneanism: Excuses for Everything, from Epistemology to Eating,” in Rethinking the Mediterranean, 45–63. © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_028

864

Davis

term with the regimes that grew, clashed, and ultimately collapsed along its shores. Fledgling empires fueled their growth with captives taken in the process of conquest: whether Greek, Roman, Arab, or Turk, their cyclical quests for dominance both depended on and were to a great extent expressed through the conquerors’ ability to thrust the vanquished into the social annihilation and institutionalized subservience that slavery represented. On this basis flourished and declined the succession of slaving societies whose presence marked the Mediterranean basin to a greater or lesser extent from the classical to the early modern eras—polities that knew how to put these deracinated humans to use, not just as a source of physical labor, but also as markers of status and more often than not as a fundamental currency of exchange. There were ebb cycles of slaving activity around the Mediterranean—after the fall of Rome, for example, and following the breakdown of the Caliphate— but the practice returned in full force during the centuries of the Crusades, now redefined along a confessional bipolarity. Indeed, it was through the very act of slaving that the circumsized and the baptized, as the case may be, affirmed their religion and identity, since in both Islam and Christianity the community of the faithful was forbidden to enslave any but nonbelievers. Despite the announced intention of several Christian religious orders founded during the Crusades to ransom captives of either faith, the infidel other was epistemologically wedded to the enslaveable other, a practice underscored through treating slaves of the faith as basically less than human.2 Not without reason were such chattel routinely called “degenerate dogs” or “dogs without religion” by both Christians and Muslims.3 The comparative decline in slave-taking following the Crusades may well have signaled less the emergence of a genuinely alternative social economy across the Mediterranean basin than a period of exhaustion and fallow, where a wary balance of forces left societies too weak to prey on each other in the customary fashion. Although the standard historiography has preferred to focus 2  The Trinitarians (Order of the Most Holy Trinity), were established by Jean of Matha and Felix of Valois in 1193; and the Mercedarians (Our Lady of Mercy) were begun by Pedro Nolasco in 1203. Though over a century old, two of the most thorough available histories of these orders and their ransoming efforts remain P. Deslandres, L’Ordre des Trinitaires pour le rachat des captifs, 2 vols. (Toulouse: E. Privat, 1903); and Emile Ledermann, Les Frères de N.-D. de la Merci et la rédemption des captifs (Paris, 1898). 3  E.g., João de Carvalho Mascarenhas, Récit de captivité de João Mascarenhas (1621– 1626), trans. and ed. Paul Teyssier (Paris: Editions Chandeigne, 1993), 144; Archivio di Stato di Napoli, Santa Casa per la Redentione dei Cattivi, busta 15, case 2881, 29 January 1735. Steven A. Epstein, Speaking of Slavery: Color, Ethnicity, and Human Bondage in Italy (Ithaca, N.Y.: Cornell University Press, 2001), 42–7.

The Geography of Slaving in the Early Modern Mediterranean

865

on the emergence of capitalism and the free labor market during these years, even such mercantile and manufacturing centers as Catalonia and northcenter Italy remained remarkably active in the slave trade and in the use of slave labor. Elsewhere, over large swaths of the Mediterranean littoral from Andalusia to southern Italy to the Aegean, the presence of slavery remained, universal even if quiescent. To an extent, this only reflected the stability of the overall political geography of those years. States were small, their powers balancing or cancelling one another out. Even the religious divide, which had so animated men in the twelfth and thirteenth centuries, and which still split the late medieval Mediterranean in half diagonally, was more apparent than real. The north and west, from Valencia to Constantinople, may have been in Christian hands but contentiously so, with fierce rivalries between the many small states sharing the coasts. Where they came into regular contact, in the Balkans and around the Aegean, Catholic and Orthodox could barely tolerate one another, each considering schismatic Christians as no better and perhaps worse than the traditional Muslim foe. Facing Christendom in the south and east was an equally fractious Islam, whose apparent uniformity from Morocco to the Levantine shores was belied by the continual jostling between Turks, Arabs, and Berbers, each seeking dominance over ancient commercial centers where Muslim authority often lay uneasily over large indigenous populations of Christians and Jews.4 The Christian and Muslim shores of the Mediterranean in fact had good reason to cooperate, at least economically, if not culturally. Whereas control of the Mediterranean sea lanes was largely assumed by Christian powers like Barcelona, Genoa, and Venice, with their superior maritime technology and aggressive naval policies, the routes that brought the most desirable and most profitable goods from their sources of supply in the distant East to key points of exchange around the Levant or the Black Sea were solidly in Muslim hands. For the space of several centuries, Christians and Muslims managed to turn the Mediterranean from a conflict zone into a vital link in a web of trade extending from the Spice Islands to the North Sea. As long as such cooperation was possible, there was little incentive to prey on each other for slaves, and it made more sense during the later Middle Ages for those in the Mediterranean slave trade to seek their human chattel as they did their silks and spices, arranging to have their slaves harvested in remote lands—from the steppes of Russia, for example, or from sub-Saharan Africa, “da quelle parti ove più scalda il sole: 4  David Nirenberg, “Muslims in Christian Iberia, 1000–1526: Variations of Mudejar Experience,” in Peter Linehan and Janet L. Nelson, eds., The Medieval World (London: Routledge, 2001), 60–76.

866

Davis

from those places where the sun burns more.”5 Slavers were able to take advantage of many of the same trade routes already established for other exotic goods: though others may have done the initial harvesting, much of the actual transport of new captives was controlled by a Christian-Muslim cartel, where the initial transport overland was firmly in the hands of the Muslims who operated the necessary caravans and chain gangs. Once arrived at such lading ports as Kaffa or Alexandria, these Slavs or Africans were sold to the Christian intermediaries who monopolized the Black or Mediterranean sea lanes and who shipped the chattel to customers waiting in Christendom or elsewhere in the Land of Islam to begin their servitude.6 Like other imported exotica, these slaves were generally bought and sold as luxury goods: the blacks we see in Carpaccio’s cityscapes—very likely he pictured liveried white slaves as well, though the modern eye fails to register them—were baubles of the very rich, whom they served as lackeys, ornaments, and curiosities, markers of status as much as sources of labor.7 This fragile equilibrium would shift drastically after the middle of the fifteenth century, as the new empires of Spain and Ottoman Turkey completed the consolidation of their home territories and began expanding into the Mediterranean arena, in parallel imperialist projects that would quickly bring the two superstates into collision along a broad front running from Morocco to the Balkans. For nearly two hundred years, the regimes that were based in Madrid and Constantinople fought one another for control of the Mediterranean and its shores, until both sides finally collapsed in exhaustion and penury—only to have their various client and successor states pick up the struggle and continue it far into the eighteenth century. Since those on both sides, jihadis or reconquistadores, saw in the contest a replay of the passions of the Crusades as much as a simple grab for wealth and territory, their struggle, as Jacques Heers has put it, “assumed a different and unique character, one tainted by an undercurrent of malice.”8 A significant share of this malice concentrated itself in enslaving religious opponents—in taking, breaking, and 5  “Great lord, we come from those places where the sun burns more; / And to show how by everyone you are honored and adored, / We have brought you these slave girls, graceful and tall, / Whose marvelous beauty can be seen by all.” Quoted in Alberto Tenenti, “Gli schiavi veneziani alla fine del Cinquecento,” in Rivista storica italiana 67 (1955): 52–69, at 53–54. 6  Largely Italian and Greek shippers carried (heathen) slaves to Christian and Muslim customers alike (Epstein, Speaking of Slavery, 122–24). 7  Denis Romano, Housecraft and Statecraft: Domestic Service in Renaissance Venice (Baltimore: Johns Hopkins University Press, 1996), 108–9; Tenenti, “Gli schiavi veneziani,” 54. 8  Jacques Heers, The Barbary Corsairs: Warfare in the Mediterranean, 1480–1580, trans. Jonathan North (London: Greenhill Books, 2003), 179.

The Geography of Slaving in the Early Modern Mediterranean

867

dealing in the faith enemy, with a passion not even seen in the heyday of the Crusades. Certainly slaving activity rose significantly around the Mediterranean after 1500.9 What had been primarily a luxury trade that might amount to only a few hundred humans a year rapidly grew to a much broader enterprise—one might almost say a social movement—that snared some thousands of victims annually. Appropriately enough for the conflict that was underway, slaving’s economic geography also shifted, returning to its format of the crusading centuries, with the primary slaving fields for both Christian and Muslim moving from the Mediterranean’s exotic hinterlands once again to the Inner Sea and its shores.10 The two faiths were no longer tacit partners trading in a few score hapless strangers but were instead intent pursuing slaving as a strategy to weaken their opponent and enrich themselves. In their aggressive desire to punish the other, vengeance played its key role: revenge against Hapsburg Spain by those Moors expelled during and after the Reconquista; revenge by the Spanish and their allies for what were perceived as past Turkish atrocities; revenge by those thousands of renegade Christians who had converted to Islam while maintaining bitter memories of their home villages, their local lords, or the faith into which they were born. Not for nothing have the slave pens where both sides housed their chattel been compared to twentiethcentury concentration camps—locations, in other words, where the inmates were held not for crimes they may have committed nor even for the military opposition they might have presented (though that often played a role in their actual capture), but simply to be punished, for who they were and what they believed. Having violated no laws, they received no protection from any legal system but labored instead under what was, theoretically, a life sentence of retribution and nonpersonhood, effectively detached from all civic authority or personal protection beyond that provided by their own masters.11 The upsurge in Mediterranean slaving was not just about ideology or revenge, of course, for there were other, more pragmatic forces involved as well. 9  Salvatore Bono, Schiavi musulmani nell’Italia moderna: galeotti, vu’cumpra’, domestici (Napoli: Edizioni scientifiche italiane, 1999), 31. 10  The expression, along with “Christian stealing,” is a contemporary one; see C. R. Pennell, ed., Piracy and Diplomacy in Seventeenth-Century North Africa: The Journal of Thomas Baker, English Consul in Tripoli, 1677–1685 (London: Associated University Presses, 1989), 120, 124, 129, and 138. 11  Stephen Clissold, The Barbary Slaves (New York: Barnes and Noble, 1992), 56; also Robin Blackburn, “Slave Exploitation and the Elementary Structures of Enslavement,” in M. L. Bush, ed., Serfdom and Slavery: Studies in Legal Bondage (London: Longman, 1996), 158–80, at 158–60.

868

Davis

Slaves were, for Christians as much as for Muslims, a commodity, and a particularly valuable one at that. The custom of assigning monetary value to those enslaved during battle was already well established during the Crusades, when both faiths negotiated for prisoner exchanges or the ransom of particularly valuable or well-to-do captives. The commodification of captivity, especially with regards to Christian slaves, was greatly expanded during the sixteenth century, as the Spanish and their client states in particular began to accept that the ransoming of enslaved subjects was both a religious and a national obligation. State magistracies in Madrid, Milan, Genoa, Florence, Rome, Naples, Venice, and elsewhere, for the most part ably aided by the Christian redemptive orders, set about enthusiastically soliciting donations from individuals and organizations, with the aim of extending the ethos of ransoming, for centuries the preserve of the wealthy and well connected, to include even the meanest sailors, peasants, and fishermen. Their efforts not only brought tens of thousands of poor captives back home, but also made it obvious to many Muslim slave owners and dealers that such chattel was worth acquiring if only as a good investment opportunity. Most of a slave’s value to his master, however, lay in his potential as a worker—male and female concubines being a special case in this regard. In the Barbary regencies of Algiers, Tunis, and Tripoli, as well as in the Christian states of Malta and Spain itself, slave labor was always necessary to take care of the “tasks that free men were no longer willing to do”: cutting timber, for example; mining salt or mercury; cutting and hauling stone; building fortifications and ships; doing agricultural work; or simply hawking water, spirits, or tobacco around the harbor.12 In particular, slaves were needed to row in the galleys that were the backbone of all early modern Mediterranean fleets, whether Christian or Muslim. To keep these ships going required a staggering number of oarsmen: 20,000 or more rowers were needed even during the rare moments when the Inner Sea was at peace, and the demand escalated vertiginously when war broke out—the single Battle of Lepanto in 1571 required (and consumed a good part of) an estimated 80,000 oarsmen. Any state with even modest pretensions for a strong navy had an insatiable need for rowers— around 180 to 200 galeotti were required on every ship—which helps explain why the Christian powers kept hunting Muslim slaves so avidly, even after it had become obvious that the rulers in Morocco, Algiers, and Constantinople had little interest in redeeming their subjects from captivity. With the naval arms race that marked the Turkish-Hapsburg duel, both sides were soon forced 12  Epstein, Speaking of Slavery, 45; Pierre Dan, Histoire de Barbarie, et de ses corsaires, 2nd ed. (Paris, 1649), 404–7; Ruth Pike, Penal Servitude in Early Modem Spain (Madison: University of Wisconsin Press, 1983), 27–40.

The Geography of Slaving in the Early Modern Mediterranean

869

to give up on their traditional means of galeotti recruitment—the Christians had used paid volunteers and convict labor, while the Turks relied on a draft levy imposed on rural townships—and turn to slave labor to make up the ever-widening shortfall. It was a problem that plagued not only procurement boards in Madrid and Constantinople, but also their client states and their independent counterparts throughout the region: by the mid-sixteenth century, the French, the papacy, the Florentines, Moroccans, and Maltese were all turning to slave labor to power their ships.13 No one who wanted to avoid ending up as the Turks so often did really had much choice: though the Ottomans had almost infinite supplies of material and the largest shipyards in Europe, they often could not put their fleet to sea, as the Venetian ambassadors repeatedly noted, since they “could not equip their ships, for the lack of [Christian] slaves.”14 Many of these oarsmen, whether Christian or Muslim, fell into the category of “public slaves,” that is, they belonged to the king or sultan, or the local pasha, bey, duke, or viceroy. These galeotti were typically put to work on state projects like building fortifications or harbor works when their fleet was wintering over in port. Such public slaves only represented a quarter or less of the total human chattel held in most slaving centers. The rest belonged to private individuals and spent their days at the tasks their masters set them, usually with the requirement that they remit a certain sum from their earnings back to their owner at the end of every week.15 This proportion of public to private slaves was less skewed when it came to Christian women, who were prized by slave raiders as “being of more worth” than men, not only because they had both sexual and economic value, but also because their home countries were often more desperately willing to ransom them back. Many Muslim women were slaves in Italian and Spanish cities, typically as serving maids, though often as concubines.16 A female presence was much more pronounced among slave populations in the Muslim world, thanks in large part to harems that could 13  For a breakdown of the changing proportions of free, convict, and slave labor in Italian galleys of the sixteenth and seventeenth centuries, see Bono, Schiavi musulmani, 166–90. 14  Archivio di Stato di Venezia, Dispacci degl’ Ambasciatori Veneziani a Costantinopoli, filza 120, 9 April, 30 April, and 20 June 1639; Nicolò Barozzi and Guglielmo Berchet, eds., Relazioni degli Stati Europei lette al Senato dagli ambasciatori Veneti nel secolo decimosettimo, vol. 5, Turchia, 2 vols. (Venezia, 1871), 1:226, 2:174. 15  Heers, Barbary Corsairs, 192–93; Robert Davis, Christian Slaves, Muslim Masters: White Slavery in the Mediterranean, the Barbary Coast, and Italy, 1500–1800 (Basingstoke, Hampshire: Palgrave Macmillan, 2003), 96–97. 16  Their names turn up only sporadically, for the few among them who converted to Christianity received their freedom or in some way fell afoul of the Holy Office. Epstein, Speaking of Slavery, 136–37; Bono, Schiavi musulmani, 331–41.

870

Davis

reach astonishing size. According to the Venetian ambassadors reporting from Constantinople in the seventeenth century, sultans kept between a thousand and fifteen hundred slave women divided into three separate seragli situated in and around Constantinople; many of their imperial lieutenants and pashas maintained their own, miniature harems, of anywhere from a few dozen to a hundred captive women.17 Although scattered slaves could be found almost anywhere around the Mediterranean, from inland cities like Milan or Madrid to such flyblown outposts as Suez in Egypt or Linguaglossa and Giarratana in eastern Sicily, they tended to be concentrated in the same port town that was the region’s principal base of slaving operations.18 On the outermost edges of the Mediterranean cultural basin, such Atlantic ports as Lisbon, Seville, and Salé in Morocco, all had thriving slave markets, though the captives they offered for sale were at least as likely to have been harvested out in the wider Atlantic world as from within the confines of the Inner Sea. Thus, of the 6,000 or so slaves one could typically find in Seville in the sixteenth century, Turks and other Muslims were little more than leavening in the mass of black African, Amerindian, and East Asian chattel held there.19 Most of the slaves in Salé and elsewhere in Morocco were also from beyond Mediterranean confines: in the early 1600s, more than half of their 1,500 or so captives were English or Irish, taken from captured ships or right off British shores: for decades thereafter the English commonly called any corsair from anywhere in Barbary a “Sallyman.”20 This sort of exogamous slaving was also the rule in the colonial ports of Lisbon and Cadiz, as well as in Provence, though the broad mixture of slaves to be found in Marseilles had less to do with reasons of geography than with diplomacy. The French, much like the Venetians, had treaties with Constantinople that forbade, among their other clauses, mutual enslavement. Still, the need for galeotti often proved stronger than any desire to stick to the diplomatic niceties, and the French, 17  Barozzi and Berchet, Relazioni, 1:103, 2:136 and 128; Heers, Barbary Corsairs, 200. 18  Bono, Schiavi musulmani, 23 and 34; Archivio Segreto Vaticano, Arciconfraternita del Gonfalone, busta 1139, fol. 157r. See also Salvatore Bono, Corsari nel Mediterraneo: cristiani e musulmani fra guerra, schiavitù e commercio (Milano: Mondadori, 1993), 194–96, who finds as many as 10,000 slaves in the city of Malaga alone, and perhaps 100,000 in the whole peninsula during the sixteenth century. 19  Mary Elizabeth Perry, Crime and Society in Early Modern Seville (Dartmouth, N.H.: University Press of New England, 1980), 20; Bono, Corsari, 195, found 14,670 slaves in the archdiocese of Seville in 1565, adding that “certainly many [of them were] black [Africans].” 20  Dan, Histoire de Barbarie, 319; Great Britain, Public Record Office, Calendar of State Papers, Domestic series, of the reign of Charles I, 1625–1649 (London, 1858–97), 1635–36, 15; 1637, 75.

The Geography of Slaving in the Early Modern Mediterranean

871

again like the Venetians, regularly found embarrassingly large numbers of supposedly protected Muslims chained to the oars of their galleys: estimates for the later 1600s have put the number at anywhere between one and three thousand, or as much as a quarter of all of the rowers in the French fleet.21 At the opposite end of the Mediterranean, Constantinople also drew its many slaves from the broad reach of Ottoman imperial power. At the peak of its slave-taking, which coincided with the culmination of its sixteenth-century clashes with Spain, Constantinople maintained somewhere between twelve and twenty thousand Christian galeotti, of whom about a third were the sultan’s own “public slaves,” the rest belonging to individual pashas and beys.22 The thoroughly heterogeneous nature of slavery in the Ottoman capital is revealed by the tallies of Christians ransomed in that city during the later seventeenth and eighteenth centuries, when such lists often included captives’ places of origin. These included the usual crop of Spaniards and Italians, as well as captives from all the contested Turkish borderlands in eastern Europe, ranging from the Hapsburg heartlands in Germany and Austria, to Poland and Lithuania, as well as Orthodox Russia, though these latter rarely appear on the ransoming lists produced by the Catholic redemptive orders.23 Slaves were also brought in great numbers from Africa, many taken in skirmishes in Upper Egypt and others purchased through the same network of trans-Saharan caravans that had supplied medieval Europe. Crowded in Constantinople’s numerous “baths” and seragli together with a scattering of unfortunate Protestants from England, Holland, and northern Germany, they made this the Mediterranean’s most thoroughly cosmopolitan slaving center since classical Rome. Not all the slaves who turned up in Constantinople had been captured by the sultan’s armies or supplied by dealers in exotic human flesh, however. Many Italians, and certainly more than a few Dalmatians and Greeks, ended in imperial bondage having been seized from fishing boats or merchant ships or during land raids carried out by corsairs sailing under the flag of one of the 21  Bono, Schiavi musulmani, 122; Gillian Weiss, “Commerce, Conversion, and French Religious Identity in the Early Modern Mediterranean,” in Keith Cameron, Mark Greengrass, and Penny Roberts, eds., The Adventure of Religious Pluralism in Early Modern France (Oxford: Peter Lang, 2000), 275–88, at 279–80; Moulay Belhamissi, Les captifs algériens et l’Europe chrétienne, 1518–1830 (Alger: L’Enterprise Nationale du Livre, 1988), 38. 22  The number of galley slaves (though not necessarily of domestic or harem slaves) dropped considerably after the 1570s: see Barozzi and Berchet, Relazioni, 2:174, 273, 355. 23  E.g., Redemptiones Captivorum Constantinopoli & Tripoli liberatorum … Discaleatorum Ssme Trinitatis, Provinciarum Germaniae & Status Veneti (Roma, 1729–30); Catalogo de’ Cristiani schiavi riscattati dalla Provincia di S. Giuseppe (Roma, 1763); Catalogus capitvorum Christianorum quos Provincia S. Josephi (Roma, 1771–73).

872

Davis

Ottoman client states, either in Barbary or in Albania—in particular, Dulcigno (modern day Ulcinj). Certainly the Ottomans preferred slave galeotti to those raised in inland Anatolia by their own levies: especially during the sixteenth century, the imperial fleet was regularly augmented by large contingents of galleys from Algiers or Tunis, manned almost exclusively by Christian slaves, who were generally agreed to be far superior to drafted villagers. Even after the Algerians stopped sending thirty or forty galleys to serve the sultan, the Turks kept trying to supplement their native rowing force whenever possible, and to do this they depended on the corsairs of Barbary or the lower Adriatic to offer Christian galeotti as gifts or tribute, or, increasingly, for purchase.24 This role of slave catcher and supplier to greater powers that the Algerians and the Dulcignotti filled so ably with the Ottoman Turks was also carried out on the Christian side by the aggressive corsairs of Malta and Livorno. In 1530 Malta was given in fief by Charles V to the recently displaced Order of the Knights of Saint John, who promptly set themselves up as the scourge of Islam and, not coincidentally, Christendom’s slaving headquarters. Ideally anchored midway not just between Italy and Barbary but also between Iberia and the Levant, the Maltese could prey on Islamic shipping and coastal settlements anywhere in the Inner Sea, and this they rapaciously did, both directly, through the six or so galleys operated by the order itself, and indirectly, through the many (usually French) privateers licensed under the Maltese flag. A freebooting capital run very much along the lines of Algiers, Malta also resembled Barbary’s principal port in its multitude of slaves: about 1,200 had to be on hand just to row in the galleys of the order, while at least six to eight times that many were needed for the private corsairs or to work locally— constructing ships, constantly refurbishing Valletta’s impregnable fortress walls, or simply acting as servants and petty merchants or shopkeepers on the island.25 Considering that its free population approached 100,000 during much of the seventeenth and eighteenth centuries, something close to ten percent of those living in Malta were probably enslaved Muslims.26 If Malta had any competitor as Christendom’s primary slaving center during these years, it was Livorno. Home of the crusading (and corsairing) Knights of 24  Increasingly Turkish villagers tried to buy themselves out of this onerous burden. Eugenio Alberi, Relazioni degli ambasciatori Veneti al Senato, ser. 3, vol. 1 (Firenze, 1840), 138; also 382–85. 25  On the organizational similarities between Algerian and Maltese corsairing, see Peter Earle, Corsairs of Malta and Barbary (London: Sidgwick and Jackson, 1970), 11–15; on slave populations and activities in Malta, see 168–74. 26  Ibid., 99 and 170.

The Geography of Slaving in the Early Modern Mediterranean

873

St. Stephan, established there in 1561 by the Florentine Grand Duke Cosimo I, Livorno rapidly became known for its aggressive approach to the Muslim world, much more inclined to use its forces for attacking its faith enemy all over the Mediterranean than merely to deploy them defensively around Tuscany. For the remainder of the sixteenth and first half of the seventeenth century, the knights and their private allies were bringing in at least as many Muslim captives as the Maltese: so many, in fact, that Cosimo’s grandson Ferdinand was moved to build a bagno outside the city to house them all—a monument to an aggressive slave-taking culture that, like the larger-than-life bronzes of four Moors situated at the harbor, soon became something of an attraction for passing Grand Tourists.27 With only one or two thousand Muslim slaves in the city, Livorno could hardly compare with Malta, or with Naples, which typically was home to as many as 20,000 Muslim captives.28 In terms of percentages, however, it was a very different story, since Livorno only had about ten to twelve thousand free residents (as opposed to over a quarter million in seventeenth-century Naples), and thus might have upwards of a quarter of its population as slaves. Moreover, as in Muslim Dulcigno, another smallish port with a high proportion of slaves, the size of Livorno’s captive population also fluctuated widely because of the town’s role as a center of the slave trade. As Peter Earle has noted, only Livorno “had a prize and slave market that could rival Malta,” and both cities grew rich off supplying fresh galeotti and domestic slaves to those states that, either due

27  The four chained, gigantic “Moors,” designed by Pietro Tacca and cast in bronze, were added to an existing statue of Grand Duke Ferdinand II at the Livorno harbor between 1623 and 1626. While the Moors are still in their place, the bagno dei schiavi, built near the galley harbor around 1600, has long since been torn down, and the space it once occupied is now used as a dog racing track. Bono, Schiavi musulmani, 164–66, 170–71, and 342–43; Michel Fontenay, “Course et piraterie méditerranéen de la fin du moyen-âge au debut de XIX ème siècle,” in Course et piraterie: études présentées à la Commission Internationale d’Histoire Maritime à l’occasion de son XV è colloque international pendant le XlV è Congrès International des Sciences historiques (San Francisco, août 1975), 2 vols. (Paris: Institut de Recherche et d’Histoire des Textes, Centre National de la Recherche Scientifique, 1975), 1:78–136, at 90–91. 28  Bono, Schiavi musulmani, 21–35, finds significant numbers of (generally Muslim) slaves, at one time or another in this period, in a host of Italian cities: somewhere between 12,000 and 50,000 in all of Sicily during the 1500s (with about a thousand “public” galeotti); “some hundreds” at Caserta; 300 or so in Civitavecchia; perhaps 1,500 to 2,000 in Genoa; and so on, to reach a total estimated population range between 20,000 and 45,000 throughout the peninsula.

874

Davis

to weakness or to treaty obligations, were unable to harvest their own.29 As already noted, the French and Venetians, but also the Genoese, often had to pick up their galeotti and other chattel on the sly, in semiclandestine purchases that the Maltese and Livornese slave merchants were more than willing to fill.30 Nor were those offered for sale always captured Turks or Moors. When out on the corso, Maltese corsairs were willing to snap up anyone whose religious or ethnic background seemed the least ambiguous or suspect: schismatic Greeks, Protestants, Jews, either practicing or conversos, converted Muslims, along with many who simply did not speak any recognizable language. Eventually the Maltese were forced by diplomatic pressure to leave the Greeks alone, but since the founding charter of the Knights of St. Stephan only forbade them from attacking fellow Catholics, the Livornese were happy to take their place.31 Once captives were taken back to Valletta or Livorno and sold, especially if they were brought in by privateers and bought by outsiders, the legal niceties about their enslaveability no longer much mattered, and those putatively in control of the corsairs “jeered at the request” of anyone else who spoke up on their behalf.32 Mediterranean slaving was, like its African analogue, essentially extractive in nature; as such, the corsairs, Muslim or Christian, who pursued this trade tended to carry out their human harvests in the areas they found to be the richest and most advantageous, until, having exhausted the local supply below the point of profitability, they moved on. Driven by this imperative, as well as by diplomatic and military pressures both internal and external, slave-taking in the Mediterranean shifted its focus over the course of the centuries after 1500, as some regions that had once been relatively lucrative became much less so. From the time of the renegade brothers Oruç and Kheir-ed-din Barbarossa— whose arrival in Algiers with troops and Turkish support in 1518 helped coalesce the vengeful energies of thousands of Moors recently exiled from Andalusia— until after the emergence of Tunis, Tripoli, Malta, and Livorno as major corsairing centers roughly half a century later, the primary Mediterranean slave fields were to the west. The long-range slaving run carried out by the renegade commander Hassan Pasha, known as “the Venetian,” ably illustrates a Barbary corsair’s typical arc of attack: setting out from Algiers with twenty-two 29  Earle, Corsairs, 178. 30  Fontenay, “Course et piraterie,” 87–90; Earle, Corsairs, 169–70. 31  Earle, Corsairs, 118. 32  Barthelémy Carré, The Travels of the Abbé Carré in India and the Near East, 1672–1674, trans. Lady Fawcett, ed. Sir Charles Fawcett, 3 vols., Hakluyt Society, 2nd ser., vols. 95–97 (London: Hakluyt Society, 1947), 3:937.

The Geography of Slaving in the Early Modern Mediterranean

875

galleys in early summer 1582, Hassan first plundered Sardinia, where he raided nearly forty miles inland and carried off around 700 men and women. He then proceeded to Corsica and the Ligurian coast, skipped over Provence for diplomatic reasons, and wound up sacking some Catalan villages before heading home, laden with around 1,300 fresh slaves.33 Others aimed their attacks in a more concentrated fashion, sailing to the coasts of Valencia and Granada to seek prizes among the English and Dutch merchants who came with seasonal regularity to load up on new Spanish wines. Like any hunters, the Algerians adjusted their movements to fit the cyclical migrations of their prey: “[They are] intending Westward to Fish for Dutchmen, whom … they hope to meet in Great Scooles,” as one seventeenth-century observer put it.34 By the later sixteenth century, however, this pattern had begun to shift. After its epochal defeat at Lepanto in 1571, the imperial Turkish fleet largely gave up naval expeditions and slaving activities anywhere west of Sicily. The Ottoman withdrawal was hastened by the aggressive galley commanders of Livorno and Malta, fresh on the scene and filled with a jihadi spirit of their own. These latter-day crusaders were eager to take on the Turkish squadrons and liked nothing better than running down and enslaving the Barbary corsairs, who soon became more circumspect in their slavetaking. Also in the last decades of the sixteenth centuries, those Christian states that had been the slavers’ prime targets began defending themselves more coherently. Especially in Spanish controlled territories, coastlines were bastioned with strings of fortresses and watch towers, forming a cohesive defensive network that (at least theoretically) allowed resident commanders to notify and bring into play the local cavalry—the only counter force the raiders really feared.35 In any case, the rich slaving grounds of Tyrrhenian Italy and southeastern Spain had by this point been largely played out. Large swaths of Spanish and Italian coastlands had been all but abandoned by their inhabitants, as those lucky enough to have escaped capture had either fled to the interior or relocated in larger and better fortified port towns, leaving behind for the corsairs little more than “wretched beaches, abandoned islands, huts [reduced to] ashes, fishermen put to flight.”36 Such desolation was everywhere by the seventeenth century, as a French traveler noted while sailing past the Tuscan island of Pianosa, which he 33  J. Morgan, A Complete History of Algiers, 2 vols. in 1 (London, 1728–29; repr. New York: Negro Universities Press, 1970), 576–78. 34  Pennell, Piracy and Diplomacy, 124; Earle, Corsairs, 56–58. 35  Braudel, The Mediterranean, 2:849–54; Bono, Corsari, 164–72. 36  Salvatore Bono, I corsari barbareschi (Torino: ERI-Edizion, 1964), 140; also Davis, Christian Slaves, 204 n. 110.

876

Davis

found “desert[ed] by its former inhabitants … [who had] abandoned the place on account of Barbary corsairs, who used to ravage the island and take as slaves anyone they found there.”37 In search of new prey, the Mediterranean slavers moved in two opposite directions, away from their old hunting grounds. The Algerians, who had the largest corsairing contingent by far and who were the most committed to piracy and slaving as a community venture, were also the most aggressive at seeking out fresh territories to harvest. By 1600 they had equipped themselves with larger vessels and had begun to venture past Gibraltar, to join the Salé Rovers at preying on Atlantic shipping. In the decades that followed, their onslaught against western Europe was unremitting, as they pillaged the coasts of France, Flanders, England, Ireland, and even Iceland, while also taking prizes by the hundreds—no fewer than 466 English ships just betwen 1609 and 1616, and 80 French vessels between 1628 and 1634—each ship with a dozen or more potential slaves on board.38 The Tunisians and Tripolitans, by contrast, sought new prey in the opposite direction, and while they continued their depredations against Sicily and southern Italy, they also expanded their efforts beyond the Ionian Sea, into the Aegean and the lower Adriatic. If their ability to mount land raids on Italy and Spain diminished sharply during the early seventeenth century, these corsairs of eastern Barbary still found a rich hunting ground in the sea routes running from Sicily to the western Peloponnese. In the process they also replaced the Tyrrhenian island redoubts they had once used to water and careen their galleys with other hideouts further east, among the hundreds of islets scattered across the Ionian and Aegean; their primary way station for heading into the Levant became the island of Lampedusa—supposedly Sicilian territory but midway between Malta and Sousse, in Tunisia—which they turned into something of a Muslim corsair colony.39 Moreover, these Muslim corsairs were not alone in relocating their raiding and slaving to the east: the Knights of Malta and of St. Stephan, along with 37  Abbé Carré, Travels, 1:22. 38  British figures are cited from W. Laird Clowes, The Royal Navy: A History from the Earliest Times to the Present, 5 vols. (London, 1897–98), 2:22; and Philemon de La Motte et al., Voyage pour la redemption des captifs aux royaumes d’Alger et de Tunis (Paris, 1721); trans. J. Morgan as Several voyages to Barbary, 2nd ed. (London, 1736), 99; on the French, see Dan, Histoire de Barbarie, 320; also, Earle, Corsairs, 58. On the famous Icelandic raid of Murat Re’is in 1627, see Borsteinn Helgason, “Historical Narrative as Collective Therapy: the Case of the Turkish Raid in Iceland,” Scandinavian Journal of History 22 (1997): 275–89. 39  John H. Pryor, Geography, Technology, and War: Studies in the Maritime History of the Mediterranean, 649–1571 (Cambridge: Cambridge University Press, 1988), 194; Earle, Corsairs, 55–56.

The Geography of Slaving in the Early Modern Mediterranean

877

some of the smaller Christian slavers such as the Majorcans, were also moving aggressively into the new hunting grounds. John Pryor has suggested that the Aegean and the Levant, which by the seventeenth century, and with the sole exception of Venetian Crete, were solidly in the Turkish sphere, represented a hostile territory beyond the cruising limits for Christian galleys, whose available space for fresh water was inversely proportional to the huge numbers of corsairs and oarsmen who needed to drink it. There is good evidence, however, that the knights laid the groundwork well for their forays into the Aegean, where by the 1600s “many islands were virtually Maltese colonies,” despite their nominally Turkish status.40 Sailing out from their bases in the Cyclades, the knights were in part looking for the Barbary corsairs, whom they considered their primary, or at least official, target. Yet they spent at least as much of their energies scouring the sea lanes that connected Constantinople with Egypt and Barbary in search of Muslim merchants, who had to take these routes to carry on business within the disparate wings of the Ottoman world. Such traders were fair prize for the knights, as was anyone operating or carrying cargo on their behalf: the ships could be claimed, the goods confiscated, and not only the merchants but also the crewmen and other passengers taken back to Valletta or Livorno and sold into slavery. By the second half of the seventeenth century, Europe’s great naval powers had, by organizing merchant convoys escorted by massively gunned sailing ships, neutralized much of the corsair activity in the western Mediterranean, together with the slaving traditionally associated with it.41 The eastern reaches of the Inner Sea, by contrast, continued to sink into a normless general piracy that made slave-taking both easy and attractive. As Ottoman control over the Greek isles effectively shrank down to little more than the sporadic, fleeting, and often punative annual visits made by a pasha based in Cyprus, each island community was left to strike what bargains it could with the corsairs of both religions, some of whom apparently remained there year round. Ships from all states that entered the region routinely carried the flags of all the others, ready to masquerade as needed to dupe a potential victim or slip away from a possible foe. In this Hobbesian world, where even one’s own coreligionists would make up any excuse to seize his ship, cargo, or person, only those sailing in a large, heavily armed convoy could really count themselves as safe—something 40  Pryor, Geography, 195–96; Earle, Corsairs, 144. 41  As the Abbé Carré put it, Travels, 3:884: “I admired the English in this business, for notwithstanding that they were at peace throughout Europe and with all the Barbary corsairs, yet they did not hesitate to keep a man-of-war in each seaport of the Levant to act as escort for their merchantmen, and to guard their trade and commerce.”

878

Davis

that Pedro Teixeira, a Portuguese converso, realized while sailing on a Venetian ship home from the Levant, recording that his captain, though recognizing four approaching “galleys of Malta … did not want them too close, knowing of the unjustifiable doings of Christian ships and galleys upon the Venetian ships from Syria, like ours.”42 Teixeira’s captain, it turns out, was under “orders to trust none, for fear of the wiles and stratagems, which the Moorish and Turkish corsairs are wont to use (and that’s true),” but he was right to doubt the intentions of the Maltese, and so was Teixeira, who, as a former Jew, could justly fear for his own freedom should he fall into their hands.43 The accounts of other travelers of that century offer a litany of other corsairs, Christian and Muslim, Livornese, Maltese, Algerian, Tunisian, or Greek, “who indiscriminately pillage all sorts of traders without making distinction between their allies and their foes.”44 The passage of the eighteenth century saw the decline—for the most part fairly rapid—of the Mediterranean slave trade. Although the Aegean remained a tenuous redoubt of buccaneering until well into modern times, the business of piracy eventually devolved into the hands of the Greeks themselves, who “became the scourge of the eastern Mediterranean, attacking not only Moslems but also English merchantmen and the Maltese corsairs themselves.”45 Sea bandits rather than slavers, the Greeks neither developed nor fed into the kind of slaving networks that had produced such profits for the Maltese and Livornese corsairs. The Knights of St. Stephan in any case went out of the corsairing business with the last of the Medici in the 1730s, by which time their Maltese colleagues were so hemmed in by the restrictions imposed by their French patrons that they essentially gave up as well. Much the same arc of decline was also traced out by the Barbary slavers, though these Muslim slave-takers took advantage of the Napoleonic disorders to stage something of a recrudescence at the end of the century, bagging thousands of new slaves, including a good many Americans, in the 1790s. Slave-taking in the Mediterranean ended in good part because it was bad for trade, in particular for the large-scale commerce that the British, Dutch, 42  Pedro Teixeira, The Travels of Pedro Teixeira, ed. and trans. William F. Sinclair, Hakluyt Society, 2nd ser., no. 9 (London: Hakluyt Society, 1902), 139. 43  For a Maltese galley taking two Venetian ships on their way to Constantinople in 1606, “enslaving the Turks and Jews on board, towing the ships to Cerigo, transferring the goods of the ships to the galley [and then] leaving them there empty,” see Bono, Schiavi musulmani, 41–42. 44  Said by the Abbé Carré of a Majorcan, Travels, 3:937; also 3:889 n. 1. 45  Earle, Corsari, 145.

The Geography of Slaving in the Early Modern Mediterranean

879

and French had managed to establish with the Ottoman Empire during the seventeenth century. The heavy gunned ships-of-the-line that these Atlantic powers used to colonize the Inner Sea economically were as effective as any diplomatic pressure in persuading the Maltese and the Livornese, as much as their Barbary brethren, to give up slaving and the corso, and not only because of their immense fire power. Sails rendered the oared galley obsolete and with it the slave galeotti that states from France to Naples to Turkey had been so desperate to have and that the corsairs proved so adept at providing. Not that the need for slaves had vanished—an entire New World awaited western European exploitation, demanding forced labor on an entirely different scale than that required to propel a few hundred galleys or build palaces and fortifications. To meet these needs, the Atlantic states, even though they had by this time established themselves as Mediterranean powers, soon abandoned the region’s centuries-old traditions of enslaving the faith enemy and turned, as their late medieval predecessors had done, to seek out exotic new chattel from slaving fields far removed from the Inner Sea.

Piracy, Ransom Slavery and Trade: French Participation in the Liberation of Ottoman Slaves from Malta during the 1620s Pál Fodor* Mediterranean historians consider the one hundred years between 1580 and 1680 to have been one of the golden ages of piracy. During this period, in addition to the many local “entrepreneurs”, four large and distinct groups distinguished themselves in this peculiar form of commerce. In the Adriatic Sea (in Dalmatia) the so-called uskoks (South Slav refugees) established an organisation specialising in piracy. Despite their limited numbers, by the end of the sixteenth century they were able—with Habsburg and Papal support—to cause enormous damage to Venetian and Turkish shipping.1 A second large centre of piracy developed in North Africa—or in Barbary as it was known in the West at the time. As their independence from Istanbul increased, the city-ports of Algiers, Tunis and Tripoli began to use their military and economic power for predatory raids. The Barbary corsairs were active primarily in the western basin of the Mediterranean, but they also ventured as far as the Adriatic and the Levant and in the seventeenth century they extended their search for booty to the Atlantic Ocean as well.2 The third group comprised two military orders, S ource: Fodor, Pál, “Piracy, Ransom, Slavery and Trade: French Participation in the Liberation of Ottoman Slaves from Malta during the 1620s,” Turcica, 33 (2001): 119–134. * Pál Fodor is senior research fellow, Institute of History, Hungarian Academy Sciences, Budapest. 1  Ekkehard Eickhoff, “Die Uskoken in der Adria. Ein Kapitel südosteuropäischer Seegeschichte”, Annales Universitatis Saraviensis V/3–4, 1956, p. 196–226; Alberto Tenenti, Piracy and the Decline of Venice 1580–1615 (Translated by Janet and Brian Pullan), London, 1967, p. 3–15; Maurice Aymard, “XVI. yüzyılın sonunda Akdeniz’de korsanlık ve Venedik”, İstanbul Üniversitesi İktisat Fakültesi Mecmuası, XXIII/1–2, 1962, p. 220–223; Kálmán Benda, “Les uscoques entre Venise, la porte ottomane et la Hongrie”, Venezia e Ungheria nel contesto del barocco europeo, Florence, 1979, p. 399–408; Catherine Wendy Bracewell, The Uskoks of Senj. Piracy, Banditry, and Holy War in the Sixteenth Century-Adriatic, Ithaca-London, 1992. 2  Otto Eck, Seeräuberei im Mittelmeer. Dunkle Blätter europäischer Geschichte, Munich-Berlin, 1940, esp. p. 135–188; Sir Godfrey Fisher, Barbary Legend. War, Trade and Piracy in North Africa 1415–1830, Oxford, 1957; Salvatore Bono, I corsari barbareschi, Torino, 1964; Tenenti, Piracy, p. 16–31; Aymard, art. cit., p. 223–225; Fernand Braudel, The Mediterranean and the Mediterranean World in the Age of Philip II, II, New York-Hagestown-San Francisco-London, 1972, p. 880–886; Peter Earle, Corsairs of Malta and Barbary, London, 1970, p. 23–29; Jamil © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_029

Piracy, Ransom Slavery And Trade

881

sometimes operating in alliance with each other. The first of these, the Knights of the Order of Saint Stephen, had been established in 1562 by Cosimo Medici in Pisa and in Livorno. The second order, the prestigious Knights of the Order of Saint John of Jerusalem, had settled on the island of Malta in 1530. After 1580 both orders switched their main area of activity from the West to the Eastern Mediterranean.3 Finally, a fourth group comprised the so-called northerners, that is, Englishmen and Dutchmen who, again after 1580, joined in the piracy on the “Inland sea”. The English pursued banditry as a kind of supplement to their commercial and carrying activities. However, the merciless violence applied by them was a considerable factor in their having pushed Venice and France into the background and their having controlled most of the trade in the Mediterranean by 1610.4 There is a considerable controversy about the nature and extent of piracy in the Mediterranean and the impact on commerce of such warfare. Some M. Abu’-Nasr, A History of the Maghrib in the Islamic Period, Cambridge-New York, 1987, p. 159, 165–166; Michel Fontenay, “La place de la course dans l’économie portuaire: l’example de Malte et des ports barbaresques”, Annales. Économies, Sociétés, Civilisations, XLIII/6, 1988, p. 1321–1347; Andreas Rieger, Die Seeaktivitäten der muslimischen Beutefahrer als Bestandteil der staatlichen Flotte während der osmanischen Expansion im Mittelmeer im 15. und 16. Jahrhundert, Berlin, Islamkundliche Untersuchungen, 174, 1994, p. 36–443; Daniel Panzac, Les corsaires barbaresques. La fin d’une épopée 1800–1820, Paris, 1999. 3  Bono, op. cit., p. 116–13; Tenenti, Piracy, p. 32–55; Aymard, art. cit., p. 225–226; Braudel, op. cit., p. 872–880; Earle, op. cit., p. 97–271; Michel Fontenay, “L’Empire ottoman et le risque corsaire au XVIIe siècle”, Revue d’histoire moderne et contemporaine, XXXII, 1985, p. 185–208; id., “La place de la course”, p. 1321–1347; id., “Corsaires de la fois ou rentiers du sol? Les Chevaliers de Malte dans le ‘corso’ méditerranéen au XVIIe siècle”, Revue d’histoire moderne et contemporaine, XXXV, 1988, p. 361–384. 4  Fisher, op. cit., p. 137 ff; Tenenti, Piracy, p. 56–86; Aymard, art. cit., p. 227–229; Braudel, op. cit., I, p. 621–642; Carlo M. Cipolla, “The Ecomomic Decline of Italy”, in Brian Pullan (ed.), Crisis and Change in the Venetian Economy in the Sixteenth and Seventeenth Centuries, London, 1968, p. 127–145; Richard Tilden Rapp, “The Unmaking of the Mediterranean Trade Hegemony: International Trade Rivalry and the Commercial Revolution”, The Journal of Economic History, XXXV, 1975, p. 499–525; Paul Masson, Histoire du commerce français dans le Levant. Au XVII e siècle, Paris, 1896, p. 27–47, 118–135; Susan A. Skilliter, William Harborne and the Trade with Turkey, 1578–1582: A Documentary Study of the First Anglo-Ottoman Relations, London, 1977; Alfred C. Wood, A History of the Levant Company, London, 1964 (2nd ed.); Alexander H. de Groot, The Ottoman Empire and the Dutch Republic: A History of the Earliest Diplomatic Relations, 1610–1630, Leiden-Istanbul, 1978; Niels Steensgaard, “The Dutch East India Company as an Institutional Innovation”, in M. Aymard (ed.), Dutch Capitalism and World Capitalism, Cambridge-Paris, 1982, p. 235–257; Halil İnalcık, Donald Quataert (eds.), An Economic and Social History of the Ottoman Empire, 1300–1914, Cambridge, 1994, p. 364–379, 487 ff, esp. p. 505–506, 520–524.

882

Fodor

scholars have maintained, for example, that maritime banditry considerably contributed to the weakening and eventual decline of Venetian shipping.5 As regards the Ottoman Empire, others have been of the opinion that these predatory actions inflicted losses that were far from being irrecoverable. As the Venetian ambassador wrote in 1641, the Maltese and Tuscan corsairs caused more irritation than pain to the Ottomans.6 At the same time, a general consensus seems to exist concerning the underlying causes of the phenomenon: the Mediterranean piracy (corso) was the attempt of impoverished societies excluded from the main stream of development to compensate themselves— at least in part—for the losses caused by the commercial ascendancy of the northerners.7 The partial success of this attempt is demonstrated by the fact that while the old commercial ports (such as Barcelona, Genoa and Venice) stagnated or declined during the seventeenth century, the ports that did switch to piracy underwent a certain amount of development.8 Of course, this was largely because ports in this second group enjoyed the support of the northerners and of Marseilles, a centre that continued to remain commercially competitive. This was also true of Malta, the capital of Christian piracy—as Peter Earle called it—an island that acquired this distinguished title while relying upon the support of a whole range of trading centres including Sicily, Naples, Leghorn, Venice, and Marseilles. It is indicative that whereas most of the Mediterranean societies experienced a severe demographic crisis in the

5  Tenenti, Piracy, p. 3–86; id., Naufrages, corsaires et assurances maritimes à Venise 1592–1609, Paris, 1959, p. 11–65. 6   Quoted by Michel Fontenay, “Chiourmes turques au XVIIe siècle”, Le genti del mare Mediterraneo (XVII e Colloque international d’histoire maritime. Naples, 1980), Naples, 1981, p. 880. Cf. Fontenay, “Corsaires de la fois”, p. 366. 7  Fontenay, “La place de la course”, p. 1324–1325; M. Fontenay, “The Mediterranean, 1500– 1800: Social and Economic Perspectives”, in Victor Mallia-Milanes (ed.), Hospitaller Malta 1530–1798. Studies on Early Modern Malta and the Order of St John of Jerusalem, Malta, 1993, p. 75–76. For the parallel phenomenon of continental banditry all around the Mediterranean, see Braudel, op. cit., vol. II, p. 734–756; Henry Kamen, The Iron Century. Social Change in Europe 1550–1660, London, 1971, esp. p. 331–426; Fikret Adanir, “Heiduckentum und osmanische Herrschaft. Sozialgeschichtliche Aspekte der Diskussion um das frühneuzeitliche Räuberwesen in Südosteuropa”, Südost-Forschungen, XLI, 1982, p. 43–116; William J. Griswold, The Great Anatolian Rebellion 1000–1020/1591–1611, Berlin, 1983. 8  Fontenay, “La place de la course”, p. 1324; id., “The Mediterranean”, p. 77. With 100,000 inhabitants Algiers was larger than Genoa, Marseilles, Barcelona and Leghorn in the seventeenth century.

Piracy, Ransom Slavery And Trade

883

seventeenth century, Malta’s population rose from about 32,000 in 1590 to almost 60,000 in 1670.9 In the following article I shall examine one aspect of the Maltese pirateeconomy, namely the capturing and ransoming of Ottoman subjects, and the commercial and political interests that lay behind such transactions. This part of my study was based on a collection of documents preserved in the Bibliothèque nationale in Paris. In all likelihood, the collection was once the property of Jean Dupuy, French consul in İzmir between 1626 and 1651.10 The “defter” in question contains 36 Ottoman legal certificates (temessük or hüccet). The importance of the material lies in the fact that twenty-seven of the thirty-six documents were issued by four different Ottoman kadıs held captive on Malta. The certificates contain acknowledgments on the part of Ottoman subjects (made on their release from captivity) that they had taken out loans with the French consul in İzmir or with his agents living in Malta in order to pay off their ransoms. Since to my knowledge such documentary evidence has not surfaced before, we may now for the first time have an insight into the operation of a very interesting and hitherto little known ransom organisation. However, before addressing this issue, I should like to evoke a number of important features of Maltese piracy.11 In the 1620s Maltese piracy was operating as a well-established system. Within this system a clear distinction was made between ventures of the knights connected with their compulsory military service and the sea voyages that were organised by private entrepreneurs with the permission of the Order and flying under its flag. At first sight the Order’s fleet seems to have been astonishingly small: in the sixteenth century, it comprised just four galleys. This number rose to five in 1596, to six in 1628, to seven in 1651 and to eight in 1685.12 On the other hand, these ships were among the best equipped in the whole of the Mediterranean and they also had the most determined crews.13 The primary tasks of this maritime force were to guard the Christian waters: every 9  Fontenay, “La place de la course”, p. 1325, 1343, notes 7, 8, 10, and 12, 1345; id., “The Mediterranean”, p. 61, 77. 10  Bibliothèque nationale de Paris, Département des manuscrits, manuscrits orientaux, fonds turc, 1234. 11  On this, see mainly the pioneering work by Earle, op. cit., p. 97–191, and the studies of M. Fontenay who supplements Earle’s description in many respects. 12  Fontenay, “Corsaires de la foi”, p. 365. 13  For the famous carrack of the Knigths, see H. J. A. Sire, The Knights of Malta, New HavenLondon, 1996, p. 87–88.

884

Fodor

year for several weeks the galleys patrolled the straits of Malta, the Tyrrenean Sea, the area around Sicily, and the entrance to the Adriatic. Whenever the European powers established a crusader naval force, the Maltese always joined it. Nevertheless, in the seventeenth century the Knights Hospitaller considered piracy to be the most effective form of fighting against the Muslims. About 500 members of the Order living on Malta took part in a number of different ways in this strange war. In order to rise within the Order’s hierarchy, the knights had to complete four so-called full caravans, that is, they had to fulfil four terms of service, each lasting six months, at sea on one of the Order’s galleys. Thus, 25–30 knights were always to be found on board the various galleys who mainly spent their tour of service looting and pillaging Levantine shipping.14 The prize was due to the Order itself, and thus for most of the knights these actions brought no more than fame and honour. With regard to the material or financial rewards of the corso, we are very much in the dark, for there is little data that can be statistically evaluated. Although there were no doubt a few large strikes, it would seem that the proceeds were far outweighed by the costs of such ventures. In 1630, for example, when the costs of maintaining the Maltese fleet amounted to 150,000–200,000 piastre, the Order’s annual income from piracy (excluding galley slaves) was estimated to be 9,000 piastre (12,000 écu).15 In circa 1660 the income may have been a mere 15,000 piastre (20,000 écu).16 In contrast, the maritime voyages of the private corsairs sailing from Malta were probably highly profitable. The activities of the privateers were regulated by a statute issued in 1605. Before setting sail for Barbary or for the Levant, a pirate ship needed to obtain the licence of the so-called Tribunale degli Armamente and the letters patent of the Grand Master. In addition, such vessels were required to fly the flag of the Order or of the Grand Master. The owner(s) and the captain of a ship had to undertake to refrain from causing damage to Christians, to submit a tenth of the prize to the Grand Master and to divide the rest of the prize according to the exact rules among other claimants of the Order, the captain, the crew, and the investors, and to submit themselves to the justice of the Tribunale in case of dispute.17 Between 1600 and 1624 the Order issued licences for 280 corsos (350 ships), which corresponds to 14  The number of compulsory caravans were three in the sixteenth as well as in the eighteenth century. Cf. Fontenay, “Corsaires de la foi”, p. 372; Earle, op. cit., p. 106; Sire, op. cit., p. 83. 15  At least according to Fontenay, “Corsaires de la foi”, p. 378; id., “La place de la course”, p. 1332. 16  Fontenay, “La place de la course”, p. 1337. The figures quoted are very uncertain. 17  Earle, op. cit., p. 108 ff.

Piracy, Ransom Slavery And Trade

885

an average of 14 ships annually. This number rose to between 20 and 25 in the period from 1660 until 1675, which we may consider the golden age of Maltese privateering.18 Often the ventures were financed by investment companies of changing composition, in which a decisive role was played by a group of about 50 men (most of whom were or had been knights). Usually they provided the technical knowledge and capital needed for the corso, and where a greater investment was necessary, equity holders and bondholders were sought, who then received payment from the prize according to the rules mentioned above. We may conclude from a list of sold prize that around 1660 the Maltese corso brought in an annual income of about 120,000–150,000 piastre.19 Organised along mainly capitalist lines, until the mid-1670s these ventures were dominated by Frenchmen, Italians and other “foreigners” (many of the knights were also of French descent). However, an order issued by King Louis XIV in 1679 compelled the French to give up piracy under Maltese flags.20 Although local entrepreneurs soon took their place, the Maltese corso began to decline. The island of Malta, the capital of piracy, became in the course of the eighteenth century a centre of trade and services and an important base for French commerce in the Levant. Captives were considered to be the most valuable prize of the Maltese corsairs. We have few data concerning their exact numbers, but between 1660 and 1662 about 200 new captives were brought to the island annually; most of these men were Muslim and Ottoman subjects. Between 1655 and 1674, on average 125 Levantine slaves were registered every year in the Maltese quarantine but with the captives taken by the Order’s ships the figure could have amounted to 200–250.21 The number of men captured was probably substantially higher, for some of them were turned over to their relatives or local community on the spot in return for a high price. The others were transported to Malta, where they were first placed in quarantine and then—once the tithe had been paid to the Grand Master—offered for sale at open auctions. A large number of the captives were purchased by the Order itself and many of them were immediately put on the benches of the galleys. In 1632, on the six galleys of the Order, 1,284 galley-slaves were pulling on the oars.22 The others were sent to work on fortress walls, or in workshops, convents and the palace of the Grand Master. Many captives were bought by Frenchmen, Spaniards, Neapolitans, and the 18  Fontenay, “Corsaires de la foi”, p. 367. 19  Fontenay, “La place de la course”, p. 1336–1337. 20  Fontenay, “Corsaires de la foi”, p. 383; Sire, op. cit., p. 91. 21  Fontenay, “L’Empire ottoman et le risque corsaire”, p. 195–196. 22  Earle, op. cit., p. 169.

886

Fodor

Papal court as galley slaves. The knights and commoners living on the island also acquired large numbers of captives with a view to working them on their ships or in their shops and houses, or in order to rent them out.23 However, most of those who acquired slaves did so with the intention of receiving a ransom at a later date. As in all the pirate port-cities around the Mediterranean, here too the greatest profits were to be made in ransom slavery. The various forms of ransoming corresponded almost exactly with the customs prevailing in North Africa, Central Europe (on the Hungarian-Ottoman border), and elsewhere, although there were a number of differences.24 One of these differences was that in the Muslim world there were no institutions (or more exactly, they ceased to exist after the thirteenth century) dealing officially with redemption (like, for example, the Redemptionist Fathers in the Christian world).25 The other difference was that the sultan’s court provided assistance only to the socalled “principal captives” or to those with influential relatives or acquaintances.26 This is what happened, for instance, in November 1590 when it promised to free six Maltese knights in return for the release of 29

23  On the life of slaves on Malta, see Earle, op. cit., p. 168–178. Cf. İsmet ParmakSizoǧlu, “Bir Türk kadısının esaret hatıraları”, Tarih Dergisi, V, 1953, p. 77–84; Walter Schmucker, “Die maltesischen Gefangenschaftserinnerungen eines türkischen Kadı vor 1599”, Archivum Ottomanicum, II, 1970, p. 191–251; Cemil Çıftçı, Macuncuzade Mustafa Efendi: Malta esirleri, Istanbul, 1996. 24  For an excellent summary of the ransom slavery in Hungary, see Géza Palffy, “Trade in Slaves and Ransom Slavery along the Ottoman-Hungarian Frontier in the SixteenthSeventeenth Centuries”, in Pál Fodor and Géza Dávid (eds), Capturing Slaves and Ransom Slavery along the Ottoman Borders (forthcoming). 25  Bono, op. cit., p. 267–349; Stephen Clissold, The Barbary Slaves, London, 1977, esp. p. 12–16, 102–129; Earle, op. cit., p. 86–88; Rieger, op. cit., p. 331–334. A remarkably useful study is Michel Fontenay, “Le Maghreb barbaresque et l’esclavage méditerranéen aux XVe–XVIIe siècles”, Les Cahiers de Tunisie, XLIV/157–158, 1991, p. 7–43. Cf. Bartolomé Bennassar and Lucile Bennassar, Les Chrétiens d’Allah. L’histoire extraordinaire des renégats, XVI e et XVII e siècles, Paris, 1989. 26  The Ottoman government’s attitude towards their subjects taken into captivity in Christian countries was studied in some detail by Nicolas Vatın, “Note sur l’attitude des sultans ottomans et de leurs sujets face à la captivité de leurs en terre chrétienne (fin XVe–XVIe)”, Wiener Zeitschrift für die Kunde des Morgenlandes, LXXXII, 1992, p. 375–395. Cf. id., “Deux documents sur la libération de musulmans captifs chez les Francs (1573)”, Wiener Zeitschrift für die Kunde des Morgenlandes, LXXXIII, 1993, p. 223–230. To my knowledge, the notion of “principal captive or slave” is only applied in the Hungarian scholarly literature.

Piracy, Ransom Slavery And Trade

887

Ottoman slaves in Malta.27 Through the intervention of the valide sultan (the mother of the sultan), the governor of Morea paid the considerable ransom, amounting to 500 guldens, for the release of Macuncuzade Mustafa Çelebi, who became famous through his accounts of his captivity in Malta.28 Under these circumstances Turks and Ottoman Jews taken to Malta could mainly rely on their relatives, acquaintances and businessmen specialising in redeeming slaves, who helped them make contact with their beloved and advanced the ransom if they saw any chance of getting their money back.29 From the documents mentioned above, one can gain understanding of a similar, so far unknown, redeeming business partnership. The leader of the enterprise was probably Jean Dupuy, the French consul at İzmir, who assumed his office in 1626 and represented the interests of the French traders in the new, rapidly developing trading centre of the Levant until 1651.30 The partners of the consul lived on Malta and were traders (tacir) and captains (kapudan), at least according to the terms used by the kadıs. One of them was called Caki Viyal (probably Giacomo Vialli), the other was known as Covan Menzelet (possibly Giovanni Menzelet). The consul and his two Maltese trading partners acted as each other’s legal representatives with full powers (vekil-i şeriyye ve kaymakam). Within three and a half years (from June 7, 1625 to November 19, 1628) they assisted 28 Ottoman captives in their endeavours to go home. An additional seven captives received money from their relatives through them. The majority of the redeemed slaves were from west and south-west Anatolia and the rest came from some islands of the Aegean (Rhodes and İstanköy/ Kos). If the places of residence of those people whose names figured among the witnesses (either released in another way or still being in captivity), and the places where the kadıs had previously worked are added to the list of the above settlements, one can gain an impression of the areas from where Ottoman subjects were taken into captivity by the Maltese during these years (see map).

27  İstanbul, Başbakanlık Osmanlı Arşivi, Mühimme defterleri 67, 73/20; 69, 280/140, 292/146, 297/149, 373/186. 28  Schmucker, op. cit., p. 247. 29  Some entrepreneurs acting on Malta, too, are mentioned by Earle, op. cit., p. 89, 91, 168– 169, 172, 174. Cf. Vatın, art. cit., p. 225–227, 229–230. It is well known that in this branch of business the Jews of Leghorn established the biggest network. In late seventeenthcentury Tunis, for example, in cooperation with their brethren, they ransomed the 44 per cent of the Christian slaves. Cf. Fontenay, “Le Maghreb barbaresque”, p. 26–29. 30  Daniel Goffman, Izmir and the Levantine World, 1550–1650, Seattle-London, 1990, p. 120.

888

Fodor

Piracy, Ransom Slavery And Trade

889

As everywhere, the amount of ransom to be paid depended on the family background and financial means of the captive. Except for one case, the sums were given in riyal guruş (Spanish 8-real) to the prisoners. The three business associates lent altogether 12,736.5 guruş to the 28 prisoners, which is an enormous sum and indicates that the creditors were not afraid of taking high risks. The highest amount was given to Mustafa Çelebi ibn Elhac Emrullah, resident of Bursa, who paid 1,091.5 guruş towards his release ; the lowest sum was paid by Kavlı bin Elhac Musa, whose release cost 191 1/4 guruş.31 The Ottoman captives paid 442 guruş on average for their freedom. This amount included three items : the ransom (baha), the so-called kapu hakkı (it was probably the price of the safe-conduct written out in their name), and the hirage charge (gemi or sefine kirası). The breakdown of costs is exemplified by only one case : while Receb bin Zülfikar from Taraklı Yenicesi paid a 200 gulden ransom (approximately 300 guruş), his letter of safe-conduct and the seafare cost him an additional 54 guruş.32 The ransom agreed and the additional costs were advanced sometimes by the consul, sometimes by the Maltese merchants, but it was always the latter two who handed the money to the slaves in the presence of a kadı held in captivity and many witnesses. In most cases the captives recognised the amount received as deyn-i şeri, that is, a loan sanctioned by Muslim religious law, and undertook to pay the entire amount back to the French consul as soon as they arrived at İzmir. Captives released concomitantly on a joint loan sometimes acted as surety for one another, but on the whole bailsmen were seldom named in the transactions. It is surprising that there was only one case where the redeeming company asked for interest to be paid. Four captives from Geyve to be repatriated pledged to pay 1,425 guruş back instead of the 950 guruş received jointly to the French consul, which is a 50 per cent interest. Assuming that the ransom dealers operated in the hope of gaining high profits rather than for humanitarian reasons33, one may conclude that in all other cases the captives acknowledged the receipt of sums that were higher than their actual debts and costs and thus the interest payment remained hidden. This assumption is supported by a certificate in which Bostan bin Yunus declared that he had received 470 guruş from the mentioned merchants “at an interest prevailing in the Muslim realm” (diyar-i İslam muamelesi üzere) and with 31  Bibliothèque nationale de Paris, Département des manuscrits, manuscrits orientaux, fonds turc. 1234, p. 28 and 5. 32  Ibid., p. 24. 33  The ransom dealers in Tunis usually demanded 20–22 per cent interest, see Fontenay, “Le Maghreb barbaresque”, p. 27.

890

Fodor

this money he had liberated himself from the hand of the unbelievers. Because he undertook to repay precisely the same amount of money, it is obvious that his debt of 470 guruş comprised the interest as well.34 The same conclusion can be drawn from a deal between a Turk from Karamürsel, a certain Mehmed reis, and the French consul in İzmir. In mid-May 1626 Mehmed handed 300 guruş to the consul for the release of Hüseyin bin Hasan, a native of his town. On top of this he paid 30 guruş as ücret-i kadem.35 The consul undertook to advance for the safe-conduct and the hirage, which would subsequently be paid back by Mehmed reis. One month later the captive in question was given 525 1/4 riyal guruş by the Maltese business associates of the consul. At first sight this would mean that he received an extra 225 1/4 guruş on top of his ransom.36 If, in the only known case, 54 guruş were sufficient to cover the price of the safe-conduct and the hirage, then 225 guruş would seem exorbitant for the same purpose. It can therefore be presumed that both in this particular case and in the other ones, the consul and his partners demanded high interest for their assistance, which was concealed, and the captives at their mercy were compelled to acknowledge the receipt of sums containing the interest as well.37 This would also explain why the amount of ransom was so high.38 34  Bibliothèque nationale de Paris, Département des manuscrits, manuscrits orientaux, fonds turc. 1234, p. 18. 35  The meaning of this term is not clear. If we read it in this form, it may be translated as “step fee” (possibly “good luck fee”). If we interpret the second word as kıdem, than it may be rendered as “seniority fee”. Whichever is correct, it can be taken for certain that it denotes the mediator’s commission due to the consul. 36  Bibliothèque nationale de Paris, Département des manuscrits, manuscrits orientaux, fonds turc. 1234, p. 12, 22, 36. 37  This practice seems a little strange. Though Islamic law in principle prohibited the taking of interest, in the Ottoman Empire even the religious establishment regarded it as a routine affair to lend money at an annual interest of 20 per cent. Thus one may suspect the business partnership of having demanded a much higher interest than was usual in the Empire. For the Islamic prescriptions concerning interest, see Joseph Schacht, Introduction au droit musulman (traduit de l’anglais par Paul Kemp et Abdel Magid Turki), Paris, 1999, p. 124. On the Ottoman practice, see Ronald C. Jennings, “Loans and Credit in Early 17th century Ottoman Judicial Records. The Sharia Court of Anatolian Kayseri”, Journal of the Economic and Social History of the Orient, XVI/2–3, 1973, p. 168–216, esp. p. 183–187. 38  For the prices on the Mediterranean slave markets and the amounts of ransom, see Jean Matthieux, “Trafic et prix de l’homme en Méditerranée aux XVIIe et XVIIIe siècles”, Annales, IX/2, 1954, p. 157–164; Maurice Aymard, “Chiourmes et galères dans la Méditerranée du XVIe siècle”, in Mélanges en l’honneur de Fernand Braudel. Vol. I. Histoire économique et sociale du monde méditerranéen, 1450–1650, Toulouse, 1973, p. 57–60;

Piracy, Ransom Slavery And Trade

891

The primary motive of the French consul’s engagement in redemption was evidently to acquire financial gains. His first documented transaction was made in 1625, that is, before he was appointed consul. This indicates that he first started bringing Ottoman captives home as part of his business undertaking. On the other hand, as consul he was in need of money from whatever source as he bought his office at an open auction for a significant sum from the French state. In addition, he received no regular salary so his expenses and investments were financed by duties, known as consulage (konsolosluk hakkı) he levied on the French traders.39 However, apart from personal interests, he may have been driven by national, political and commercial motives as well. By the 1610s İzmir had become the major transit harbour on the east-west trade route, where the Venetians and French, who had previously dominated the Levantine trade, were in sharp competition with the English and Dutch for the raw material and markets of western Anatolia. In the 1620s the development of the European buying up systems in the hinterland of İzmir was in full swing. Those who could maintain good contact with the local authorities could really become successful. Those, like the Venetians, who primarily expected protection from the central government, were facing increasing difficulties as from the beginning of the seventeenth century the court’s power considerably weakened in the provinces. In these circumstances, it was in the best interests of the French consul, who was constantly at loggerheads with his own merchants and the French ambassador to Istanbul because of the levy and its distribution, to win the confidence and support of the local authorities. To this end, like other foreigners, he often lent money to the holders of taxfarms of the commercial and customs district of İzmir-Chios. As soon as he assumed his office, he lent 2,510 guruş to the Jewish tax-farmer Elkas Avraham so that the latter could send to Istanbul the amount demanded for the salary of the court mercenaries and for the yearly payment (salyane) of a bey.40 In 1629 he lent 1,500 guruş to an Armenian tax-farmer called Bedr, who, similarly to

Fontenay, “L’Empire ottoman et le risque corsaire”, p. 196–197, 199; id., “Le Maghreb barbaresque”, p. 25–26, 34. 39   M asson, op. cit., p. 51–54, 77–95; Gaston Rambert (ed.), Histoire du commerce de Marseille, Tome IV, Louis Bergasse, De 1599 à 1660, Gaston Rambert, De 1660 à 1789 Paris, 1954, p. 56– 73. Niels Steensgaard, “Consuls and Nations in the Levant from 1570 to 1650”, Scandinavian Economic History Review, XV/1–2, 1967, p. 30–31; Goffman, op. cit., p. 97–98, 120. 40  Bibliothèque nationale de Paris, Département des manuscrits, manuscrits orientaux, fonds turc. 1234, p. 10. Dupuy lent Avraham money later on as well, but after the latter’s death (1635) he had difficulties in recovering his loan. Cf. Goffman, op. cit., p. 127–128.

892

Fodor

Avraham, promised to repay the sum by conveying a fixed part of the exportimport duties to the consul.41 In the competition for the patronage of the authorities, the French consul evidently gained additional moral and political prestige as a result of his assistance in the release of Ottoman captives. It is hardly surprising that the Maltese transactions occurred at the beginning of his term of office. He probably wanted to gain the goodwill of the municipal authorities and the people and thus consolidate his position. By doing this, he not only increased his freedom of activity, but improved the reputation of the French nation as a whole as the Ottomans were aware that the French played a significant role in the Maltese kidnapings (probably this was the reason for the French ambassador to Istanbul’s participation in the Maltese-Ottoman exchange of captives in 1590–1592). Therefore the entire redemptionist activity of the French consul had an element of irony as the situation which he turned to his own and his merchants’ financial and moral advantage was mainly the result of the banditry of his fellow countrymen on the Mediterranean. Appendix List of Ottoman slaves who were either released from Maltese captivity or received funds with the assistance of the French consul in İzmir and two Maltese merchants and the amount of their ransom, 7 June 1625–19 November 1628. (In the case of slaves marked with an asterick, the merchants only forwarded the money sent by relatives and friends.)

Name and place of origin

Ransom (baha)

Gate-duty (kapu hakkı) and hirage charge (gemi or sefine kirası)

1. Bostan bin Yunus, Elmalı 2. Receb bin Zülfikar, Taraklı Yenicesi 3. Musli bin Abdurrahman, İstanköy 4. Mehmed bin Yusuf, İzmir

470 riyal guruş 200 sikke-i hasene

Included in the ransom 54 riyal guruş

292 riyal guruş

Included in the ransom

230 3/4 riyal guruş

Included in the ransom

41  Bibliothèque nationale de Paris, Département des manuscrits, manuscrits orientaux, fonds turc. 1234. p. 6.

893

Piracy, Ransom Slavery And Trade

Name and place of origin

5. * Musli Dede bin Memi, (?) 6. *Hasan bin Veli,(?) 7. * Bali bin Süleyman, (?) 8. *Abdurrahman bin Mehmed, (?) 9. Emrullah bin Ali, Geyve

10. Elhac İbrahim bin Şaban, Geyve 11. Abdulfettah bin Mahmud, Geyve 12. Hasan bin Ahmed, İstanköy 13. Hüseyin bin Hasan, Karamürsel 13a.  Elhac Hüseyin bin Kurşun Hasan, Karamürsel 14. Mehmed bin Bali, İznik 15. Elhac Abdi bin Yusuf, Tire 16. Mehmed bin Turmış, İstanköy 17. Kavlı bin Elhac Musa, Manisa (?) 18. Abd reis bin Mustafa, Üsküdar 19. Mehmed, bin Abd reis 20. Ahmed, nephew of Abd reis

Ransom (baha)

Gate-duty (kapu hakkı) and hirage charge (gemi or sefine kirası)

348 riyal guruş

Included in the ransom

560.5 riyal guruş 348 riyal guruş 163 3/4 riyal guruş

Included in the ransom Included in the ransom Included in the ransom

950 riyal guruş (for the 3 slaves nos. 9, 10, 11 making a total of 1, 425 guruş to be repaid) ..

Included in the ransom

.. 275 3/4 riyal guruş

Included in the ransom

300 riyal guruş + 30 guruş for the consul as ücret-i kadem 525 1/4 riyal guruş

Separate, but not known

329.5 riyal guruş 780.5 riyal guruş

Included in the ransom Included in the ransom

304 riyal guruş

Included in the ransom

191 1/4 riyal guruş

Included in the ransom

Included in the ransom (probably 2251 /4 guruş)

Included in the ransom 2,000 3/4 riyal guruş (for the 3 slaves nos. 18, 19, 20 together) .. ..

894

Fodor

Appendix (cont.) Name and place of origin

Ransom (baha)

21. Elhac Mustafa bin 519.5 riyal guruş Demür, İzmir 22. Musli bin Mahmud, İzmir 1,205.5 riyal-i kebir (for the 3 slaves nos. 22, 23 and 24 together) 23. Ali bin Mahmud, İzmir .. 24. İbrahim bin Nebi, İzmir .. 25. Mustafa Çelebi ibn 1,091.5 riyal-i kebir Elhac Emrullah, Bursa 26. Osman bin Kara 438 3/4 riyal-i kebir Mehmed, Ula 27. Ali bin Șaban, Rhodes 1,727.5 riyal-i kebir (for the 4 slaves nos. 27, 28, 29, 30 together. They stood surety for each other). 28. Ivaz bin Seyami çavuş, .. Rhodes 29. Ali bin Ramazan, Rhodes .. 30. Hasan bin Mahmud, .. Rhodes 31. Elhac Mustafa bin Elhac 597 1/4 riyal-i kebir Mahmud, the village of Durbalı (Anatolia) 407.5 riyal-i kebir 32. İbrahim bin Elhac Hasan, Sıǧacık 33. * The Jew İlyas veled-i 100 riyal guruş (sent by Meyir, İzmir his mother) 34. * The Jew Simo veled-i 100 riyal guruş (sent by Yasef, Rhodes a Jew from Tire) 35. * Ali Çelebi bin 20 riyal guruş (sent by Abdullatif, imam, Bursa his mother)

Gate-duty (kapu hakkı) and hirage charge (gemi or sefine kirası)

Included in the ransom Included in the ransom

Included in the ransom Included in the ransom Included in the ransom

Included in the ransom

Included in the ransom (not yet released) (not yet released) (not yet released)

Shifting Patterns of Ottoman Enslavement in the Early Modern Period1 Ehud R. Toledano

The Ottoman Empire and the Enslaved2

The Ottoman Empire was the last and greatest Islamic power of the early modern and modern eras. In many ways, the history of the Middle East between 1517 and 1918 is a chapter in Ottoman history, and the Ottoman heritage lingered in the eastern Mediterranean many decades after the demise of the Empire. While often viewed in the West as the paragon of conservatism and stagnation, the last two decades of intense research have shown that the Ottoman Empire was, through many periods of its long history, a complex and fascinating entity, dynamic and adaptable, pragmatic and resilient, tolerant and accommodating. There were of course periods in which it lived up—in many ways—to the negative view it has acquired, but the overall account of the teeming and diversified social web under its rule certainly defies that image. The ‘decline S ource: Toledano, Ehud R., “Shifting Patterns of Ottoman Enslavement in the Early Modern Period,” in Eyal Ginio and Elie Podeh (eds.), The Ottoman Middle East: Studies in Honor of Amnon Cohen, Leiden and Boston: Brill, 2014, 201–220. 1  I chose the present article for this volume because it seems to me that it touches on a number of themes that Professor Amnon Cohen, my teacher at the Hebrew University in the 1970s, has treated occasionally in his work. These include the history of the Mediterranean in the 16th to the 18th centuries, with special reference to economic and social aspects of that history. His work on the port cities of Palestine, and the use he made of the Marseilles Chambre de Commerce archive, provided inspiration for some of the discussion in the following pages. His thorough research on, and use of, the Shari˒a court records have contributed to my own efforts, here and elsewhere, in exploiting the Nizami court records of the 19th century. An earlier, and differently structured, version of this article was published in Istituto internazionale di storia economica F. Datini, Settimana di studi (38th: 2006, Prato, Italy), Relazioni economiche tra Europa e mondo islamico, secc. XIII–XVIII: atti della “trentottesima settimana di studi” 1–5 maggio 2006 [Europe’s economic relations with the Islamic world, 13th– 18th centuries], Simonetta Cavaciocchi (ed.), (Le Monnier: Grassina, 2007), vol. ii, 699–718. 2  Although it deals mainly with Ottoman enslavement in the “long 19th century” and not with the early modern period discussed in this article, I have occasionally drawn here on pertinent points made in my As If Silent and Absent: Bonds of Enslavement in the Islamic Middle East (New Haven and London: Yale University Press, 2007).

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_030

896

Toledano

paradigm’ of Ottoman history refers to the early modern period, or more precisely to the 17th and 18th centuries. The dramatic transformation that the Empire underwent during that period also had a profound impact upon the social institution of enslavement. Enslavement was a universal phenomenon, present in almost every known human society and culture. The Ottomans were no exception, and in many ways they followed in the footsteps of earlier Islamic states. The legal essence of enslavement derives from Islamic law, although various Muslim societies developed their own brand of enslavement. Thus, the Ottomans had inherited the basic practice of military-administrative bondage from the caliphate of Al-Muʿtasim (r. 833–842) and later the Mamluk Sultanate (1258–1516/7), which they had defeated and replaced in the eastern Mediterranean. From fairly early on in their history, the kul, or the slaves of the sultan recruited by child levy in the Balkans, formed the backbone of the imperial army and government. The female cognate institution was known as harem slavery, and for much of the early modern period, it was the mainstay of the imperial and elite household network. But other types of bondage co-existed with the kul/harem system, namely agricultural, domestic, and menial enslavement. The main source of enslaved persons in the Ottoman Empire was the large pool of captives taken as spoils of war on the various fronts, including those in Europe. From the 15th to the 18th centuries, European slaves, mostly from the north-eastern shores of the Mediterranean, i.e., Greece and the Balkans, were forcibly transported into the Ottoman Empire and employed there in different occupations. While many of the male captives were used on agricultural estates, others were employed on smaller farms, gardens, vineyards, and orchards. A large number of women reached the harems of the imperial elite and were engaged in domestic service or—in smaller numbers than fantasized by European travelers—used as concubines. The enslaved tended to livestock and fisheries, and worked in bakeries, cotton mills, shipyards and ports. Some of the servants were trusted with trade and finances, but most of those who labored in an urban setting were simply servants.3 In general, the Ottoman system did not favor agricultural enslavement, and the advent of Ottoman rule, as in the case of Cyprus, normally meant the extinction of land cultivation by enslaved persons.4 One of the few scholars to 3  Yvonne Seng, “A Liminal State: Slavery in Sixteenth-Century Istanbul,” in Slavery in the Islamic Middle East, ed. Shaun E. Marmon (Princeton: Markus Wiener Publishers, 1999), 25–42 (more specifically for our purposes here—30). 4  Ronald C. Jennings, Christians and Muslims in Ottoman Cyprus and the Mediterranean world, 1571–1640 (New York: New York University Press, 1993), 240–241.

Shifting Patterns of Ottoman Enslavement

897

have studied Ottoman agricultural enslavement is Halil Inalcik, who showed that enslaved captives were employed on large Ottoman estates up to the 16th century, being mainly used in rice fields and other cash crop farms.5 But with the major transformations that the Ottoman Empire underwent in the 17th century, agricultural production—alongside the restructuring of other aspects of government, society, and the economy—was reorganized into smaller units, abandoning the great estates and their large-scale methods of labor exploitation. In the evolving realities of the Ottoman countryside, agricultural labor in most regions was now organized around a free peasantry, cultivating small tracts of land under several tax-regimes. Much of the newly acquired territories was parceled out to low and high-level military officers in return for taxes and military service as part of the timar system. In the early modern period, that system evolved into short-term tax farming by auction (iltizam) and later—larger long-term and lifetime tax farms (malikâne).6 Agricultural enslavement was to be seen again in the Ottoman Empire only in the second half of the 19th century, when it was introduced from the outside either as an imported formation or as a local response to an external market stimulant. The first instance occurred when the Circassians were deported—in today’s parlance, ethnically cleansed—from the Caucasus by Russia during the 1850s and 1860s. The refugees were allowed to bring with them to the Ottoman Empire their enserfed farm laborers, who entered as families and were settled on land provided by the government. The other case of agricultural slavery was the temporary rise in the use of enslaved Africans, mostly Sudanese, on the cotton fields in Egypt during the cotton boom of the 1860s, caused by the American Civil War. But these two cases lie well beyond the chronological scope of this paper, as they occurred in the second half of the 19th century.7 5  Halil Inalcik, “Servile Labor in the Ottoman Empire,” in The Mutual Effects of the Islamic and Judeo-Christian Worlds: the East European Pattern, eds. Abraham Ascher, Tibor Halasi-Kun, and Bela K. Kiraly (New York: Brooklyn College Press, 1979), 25–52, especially 30–35; and idem, “Rice Cultivation and the Çeltukci-Re’âyâ System in the Ottoman Empire,” Turcica 14 (1982): 69–141, especially 88–94. 6  For an introduction-level survey of the timar system, see Halil Inalcik, The Ottoman Empire, the Classical Age, 1300–1600 (New Rochelle, NY: A. D. Caratzas, 1989); for a monographic treatment of the transformation into iltizam and malikâne, see Dina Rizk Khoury, State and provincial society in the Ottoman empire: Mosul, 1540–1834 (Cambridge: Cambridge University Press, 1997), and the references to Ariel Salzmann’s articles contained therein. 7  For agricultural enslavement among the Circassians in the Empire, see Toledano, Slavery and Abolition, Chapter 3. For agricultural enslavement in Egypt, see Gabriel Baer, “Slavery and Its Abolition,” in idem, Studies in the social history of modern Egypt (Chicago: University of Chicago Press, 1969), 161–189. For a related discussion, see also Ehud R. Toledano, “Where

898

Toledano

Thus, the other forms of enslavement in the Empire towards the early modern period included domestic service in urban elite households, largely performed by women who were either captured or purchased from southeastern Europe, lands lying north of the Black Sea, and increasingly in the 18th century—from the regions between the Caspian and the Black Seas, i.e., Georgia and Circassia. Enslaved men from all those regions performed menial tasks such as mining and occasional public works. This variety of functions performed by enslaved persons in Ottoman societies, coupled with the equally varied places of origin from whence the enslaved were wrenched, constitute the fabric of Ottoman enslavement as a subject of research and study. Rather than think of these as unrelated, disjointed types, we should rather see them all inhabiting a continuum, with varying origins, cultures, functions, and statuses.8 Indeed, the complexity of the practice makes it necessary to look for a differentiated approach that can accommodate its internal contradictions and seeming intractability: even to a trained eye in non-Ottoman, non-Islamic forms of enslavement, this mixture of high and low status, as of honor and shame, must look perplexing. To better understand the spectrum of enslavement, we may examine the experience of enslaved persons in the Empire from six angles, which in turn affected their treatment and shaoped their fortunes:

• the tasks the enslaved performed—whether domestic, agricultural, menial, or kul/harem; the • stratum of the slavers—whether an urban elite, rural notability, smallhold cultivators, artisans, or merchants; location—whether in the core or the peripheral areas; • type of habitat—whether urban, village, or nomad; • • gender—whether male, female, or eunuch; • ethnicity—whether European or Caucasian (and later African). Have All the Egyptian Fallahin Gone to? Labor in Mersin and Çukurova during the Second Half of the Nineteenth Century,” in Mersin University, Center for Urban Studies, Mersin, the Mediterranean, and Modernity: Heritage of the Long-Nineteenth Century (Mersin: Mersin University Press, 2002), 21–28. 8  A model for understanding Ottoman enslavement is suggested in my article “The Concept of Slavery in Ottoman and Other Muslim Societies: Dichotomy or Continuum?,” in Slave Elites in the Middle East and Africa: A Comparative Study, eds. Miura Toru and John Edward Philips (London and NY: Kegan Paul International, 2000), 159–176.

Shifting Patterns of Ottoman Enslavement

899

For the late 18th century and after, the following impression clearly emerges from the sources:

• enslaved domestic workers in urban elite households were better treated than enslaved people in other settings and predicaments; • the farther from the core, the lower on the strata scale, and the less densely-

populated the habitat, the greater the chances the enslaved had to receive worse treatment.

Before we move on, however, we need to address briefly the question of whether the category of kul/harem should be discussed as a legitimate form of enslavement. Leading Ottomanists have suggested alternative terms to describe the predicament of people in that group, feeling that they cannot properly be lumped together with members of the less privileged groups of domestic and agricultural slaves in Ottoman societies. Metin Kunt refers to the kul as “the sultan’s servants,” whereas Suraiya Faroqhi prefers to call them “servitors.”9 As against that, in her recent book Morality Tales, Leslie Peirce firmly asserts that “the privileges of elite slavery were temporary.” For one thing, she adds, elite slaves were not allowed to bequeath their wealth—nor status, I would add—to their offspring, and their wealth reverted to the treasury upon their death (to an extent a loophole was available to them through the mechanism of charitable endowment known as vakıf/waqf). Just as the sultan “controlled his enslaved servants’ religious and cultural identity and their material environment,” Peirce argues, he also “controlled their right to life, taking it if they were judged to have violated their bond of servitude.” She then defines what in her view is “a paradox at the heart of the Ottoman system—that ordinary subjects enjoyed rights denied to those by whom they were governed. One of their rights was immunity from the sultan’s direct power of life and death.”10 Despite the fact that, over the centuries of Ottoman imperial rule, certain aspects of kul servitude were gradually being mitigated in practice, Peirce is certainly correct in her observations. In fact, the main changes in the status of kul/harem slaves came only in the 19th century, whereas in the early modern 9   See Metin Kunt, The Sultan’s Servants: the Transformation of Ottoman Provincial Government 1550–1650 (NY: Columbia University Press, 1983); Suraiya Faroqhi, “The Ruling Elite between Politics and ‘the Economy’,” in An Economic and Social History of the Ottoman Empire 1300–1914, eds. Halil Inalcik and Donald Quataert (Cambridge: Cambridge University Press, 1994), 564 ff. 10  Leslie Peirce, Morality Tales: Law and Gender in the Ottoman Court of Aintab (Berkeley, CL: University of California Press, 2003). All quotes are from page 315.

900

Toledano

period we can safely assume their bondage to have been real in the terms described above. As in previous works, here too, my view is that all legally bonded subjects of the sultan should be treated as enslaved persons for the purpose of social analysis, also. This is an integrated, inclusive position, i.e., that there was no difference of kind between kul/harem slaves and other types of Ottoman slaves, although there certainly was a difference of degree of enslavement between them within the category of Ottoman slavery. Finally, we should say a few words about the number of persons enslaved within the Ottoman Empire. Clearly, we still need further research before the picture becomes full and clear. Madeline Zilfi rightly observes for the end of the 19th century that the size of the enslaved population hovered around five percent, and enslavement was “the practice of a small, privileged minority and as such scarcely reflected the experience of the majority.”11 The overwhelming number of families, she adds, were monogamous, and did not own slaves nor employ free servants. This was probably true also for the early modern period, though we still do not possess reliable data for the volume and precise nature of Ottoman enslavement during that time. However, it does seem plausible to assume that the demise of agricultural enslavement reduced the overall number of enslaved persons in the Empire, whereas urban household bondage continued to remain on the same level. At the same time, households became much more important from the 17th century onward, both in the metropolitan center and the provinces, forming the backbone of political-social-economic life in Ottoman societies.12 It is also very difficult to project backwards the 19th-century figures given for the volume of the slave trade into the Ottoman Empire or the size of the enslaved population within its borders. Much of the work for the last hundred years of the Empire has been done and published, and despite some legitimate differences, the picture is fairly clear. Scattered data and reasonable extrapolations regarding the volume of the slave trade from Africa to the Ottoman Empire yield an estimated number of approximately 16,000 to 18,000 men and women who were being transported into the Empire per annum during much of the 19th century. Ralph Austen’s estimates for the total volume of

11  Madeline C. Zilfi, “Servants, Slaves, and the Domestic Order in the Ottoman Middle East,” Hawwa 2/1 (2004), 29. 12  On that, see Ehud R. Toledano, “The Emergence of Ottoman-Local Elites (1700–1800): A Framework for Research,” in Middle Eastern Politics and Ideas: A History from within, eds. I. Pappé and M. Maʿoz (London and New York: I.B. Tauris, 1997), 145–162, and the literature on Ottoman-local elites contained therein.

Shifting Patterns of Ottoman Enslavement

901

coerced migration from Africa into Ottoman territories are as follows:13 from the Swahili coasts to the Ottoman Middle East and India—313,000, across the Red Sea and the Gulf of Aden—492,000, into Ottoman Egypt—362,000, and into Ottoman North Africa (Algeria, Tunisia, and Libya)—350,000. If we exclude the numbers going to India, a rough estimate of this mass population movement would amount to more than 1.3 million people. During the middle decades of the 19th century, the shrinking Atlantic traffic swelled the numbers of enslaved Africans coerced into domestic African markets, as well as into Ottoman ones. Although the regions whence enslaved persons were being captured and sold into the Ottoman Empire had changed dramatically, as we shall see below, it might be reasonable to argue that—allowing for the expected population growth—overall demand remained fairly steady. With the end of agricultural enslavement in the 16th century, the internal market restructured itself around stable demand for unfree labor in domestic, menial, and household service. Following the significant rise during the 16th century in the number of kul required by the imperial government, and later the changing structure and functions of the imperial army in the 17th, much of the force was no longer servile. Increasingly, the recruitment and socialization of actual kul devolved from the sultan’s household-court to leading members—kul and non-kul—in both his central and provincial administration. All in all, the size of the kul/ harem group was shrinking towards the end of the early modern period. So we may perhaps venture an educated guess that already in the 18th century, kul/harem slaves no longer formed a significant part of the enslaved population in the Empire. While many of the leading officeholders in the military and administration were still kul, the bulk of both the army and the bureaucracy consisted of free men. If we also allow for a 17th to 18th centuries lull in agricultural enslavement, then the rest of the unfree labor market should have remained the same going into the 19th century as well. Many of the women in urban elite households were still harem-enslaved, but they would now be taken from different regions. Thus, we may have at least a rudimentary method of reckoning backward the numbers of enslaved persons that were being transported into Ottoman territories during the early modern period; we could then 13  The following figures are derived from his two articles: “The 19th Century Islamic Slave Trade from East Africa (Swahili and Red Sea Coasts): A Tentative Census,” in The Economics of the Indian Ocean Slave Trade in the Nineteenth Century, ed. William Gervase ClarenceSmith, Special Issue of Slavery and Abolition, 9/3 (1988): 21–44, and “The Mediterranean Islamic Slave Trade out of Africa: A Tentative Census,” Slavery and Abolition 13/1 (1992): 214–248.

902

Toledano

attempt to estimate the size of the enslaved population. However, until reliable figures can be obtained from archival sources of the early modern period, any estimates will have to begin with the figures and estimates we already have for the 19th century. As for the ethnic composition of the enslaved population, Yvonne Seng’s work is especially useful as she maps the ethnic landscape in Üsküdar, what is now the Asian part of Istanbul.14 She writes that at the beginning of the 16th century, approximately fifteen percent of that town’s 30,000 inhabitants were non-Muslim, and that in public places and commercial areas, the languages often heard included “Greek, Armenian, Turkish, Farsi, Kurdish, Slavic, and Central Asian dialects.” Greek and Slavic obviously represented people who came from south-eastern Europe, but whereas we cannot assume all of them to have been enslaved, it stands to reason that not a few were. Most of the enslaved persons who appear in court records—mainly because they had absconded—point out that they had been captured in the Ottoman campaigns in the Balkans, but also in the Crimean Khanate’s incursions into Russia and Poland. Origins of captives listed in court records include Russian (39 percent), Croatian (31 percent), and Bosnian (11 percent), with the remaining 19 percent coming from Hungary, Walachia, and Bulgaria; enslaved Greeks, Circassians, Albanians, and Africans were rare at the time.

The Pattern Shift of the Early Modern Period

As far as enslavement in the Ottoman Empire was concerned, the decades bridging the 18th and the 19th centuries witnessed a major shift in slave trading patterns. What was a mostly European traffic gradually turned into a mostly African phenomenon, and a pool of enslaved Europeans could no longer be replenished by conquest and prisoner taking. Trade replaced the spoils of war, and the Ottomans, like the Europeans and Americans before them, turned to Africa, where internal strife and poverty placed enslaved women and men on the markets in large numbers. Whereas the Atlantic world was mostly interested in the male population for agricultural labor, the Ottoman world, to a large extent like the Indian Ocean world, looked for females for domestic service in urban households. There is still precious little research on Ottoman enslavement in the period preceding the 19th century. Whereas the past two decades have seen an impressive growth in the literature dealing with enslavement and the slave trade 14  Seng, “A Liminal State,” 27–28.

Shifting Patterns of Ottoman Enslavement

903

in the Empire during the last century of Ottoman rule, only very few studies, all in article format, have appeared on the early modern period. Here we have to rely on the pioneering work of scholars like Halil Sahillioğlu, Halil Inalcik, Ronald Jennings, Alan Fisher, Jane Hathaway, Yvonne Seng, Suraiya Faroqhi, and Madeline Zilfi.15 Much ground still needs to be covered before a satisfactory picture can emerge, but the good news is that there is no shortage of first-hand and first-rate sources—from court records and government correspondence to narrative accounts. For the early modern period—contrary to the 19th century—these have been only very partially tapped: one need only count the small number of court cases adduced in all these studies together in order to realize the task ahead and the promising potential for future research. I would like to begin this section with three stories that reflect the dramatic change over time we wish to explore here. These stories come from the contact regions between south-eastern Europe and the Ottoman Empire, and they are all culled from Ottoman court records dating from the late 16th to the early 19th century. It is far from my intention to lift them out of their historical context in order to draw an imaginary line of development; rather, these have been selected only in order to put a human face on the shift in trade patterns that

15  For full references to these works, see notes 6–9, and 14. See also: Halil Sahillioğlu, “Slaves in the Social and Economic Life of Bursa in the Late 15th and Early 16th Centuries,” Turcica 17 (1985): 7–42; Alan Fisher, “The Sale of Slaves in the Ottoman Empire: Markets and State Taxes on Slave Sales,” Boğaziçi University Journal 6 (1978): 149–174; idem, “Studies in Ottoman Slavery and Slave Trade, II: Manumission,” Journal of Turkish Studies 4 (1980): 49–56; R. Jennings, “Black Slaves and Free Slaves in Ottoman Cyprus, 1590–1640,” JESHO 30/3 (1987): 286–302; M. Zilfi, “Goods in the Mahalle: Distributional Encounters in Eighteenth-Century Istanbul,” in Consumption Studies and the History of the Ottoman Empire, 1550–1922: An Introduction, ed. Donald Quataert (Albany, NY: State University of New York Press, 2000), 289–311; John Hunwick, “Islamic Law and Polemics Over Race and Slavery in North and West Africa (16th–19th Century),” in Slavery in the Islamic Middle East 43–68; Yvonne J. Seng, “Fugitives and Factotums: Slaves in Early Sixteenth-Century Istanbul,” JESHO 39/2 (1996): 136–167; Hasan Ferit Ertuğ, “Musahib-i Sani-i Hazret-i ŞehrYari Nadir Ağa’nın Hatıratı-I,” Toplumsal Tarih 49 (October 1998): 7–15; Suraiya Faroqhi, Stories of Ottoman Men and Women (Istanbul: Eren, 2002) (this is a collection of previously published articles, of which nos 4, 6, and 13 (published between 1997 and 2001) are relevant to our discussion; Erdem (full citation in note 22 below) also has some interesting observations about pre-19th century enslavement. On kul/harem enslavement in Ottoman Egypt, see Jane Hathaway, The Politics of Households in Ottoman Egypt: the Rise of the Qazdaglis (Cambridge: Cambridge University Press, 1997), and idem., A Tale of Two Factions: Myth, Memory, and Identity in Ottoman Egypt and Yemen (Albany, NY: State University of New York, 2003).

904

Toledano

we have already noticed. The stories will first be told, followed by an attempt to put them in the relevant context. The first story comes from the Şeriat court in Ottoman Cyprus in mid-1594, and the main protagonist was an enslaved white woman, most probably of European descent, named Rahime bint Abdüllah, from her name clearly a convert to Islam.16 The slaver who purchased Rahime was a district governor on the island, and he obviously held her in great esteem and confidence since she served as his treasurer or accountant. So much so, that Rahime was entrusted with delivering four loads of jewels, goods, and cash, all valued at the large amount of 15,000 gold coins, from the district to the provincial treasury in Nicosia (Ottoman Lefkoşa). But, following the death of the governor, and scheming with one of the province’s high financial officers, Rahime and the latter, named Mustafa, absconded with the cargo and were never found. For us, the story indicates the fact that enslaved Europeans were commonly present in Ottoman Mediterranean societies, and that some of them enjoyed a high degree of trust; at times, as here, this trust was violated. The second story takes us to Istanbul towards the end of the 18th century.17 On 17 June 1786, Ahmet Hasbi, the kadi of Bursa, signed a legal notice reporting the arrival in the city of two unmarried female slave dealers, said to have their respective origins in Bursa and in Syria. The head of the slave dealers’ guild and police officers alleged that the two had engaged in activities that contravened the acceptable norms of the trade, as enshrined in the imperial edict and regulations governing the purchase and sale of enslaved persons. The women were banished from Istanbul as an example to the rest, and in order to protect the code of the slave dealers’ guild, explained the order which was delivered to the kadi by a special courier of the Imperial Council. Although this was not spelled out in the kadi’s memo, it seems safe to assume that the kind of transgression attributed to the two was of a sexual nature, i.e., exposing enslaved females to male clients under the guise of displaying them in a private house as if for sale. Finally, the third story is situated in the Bulgarian port city of Varna, on the western shores of the Black Sea. On 14 August 1827, the registrar of the local Şeriat court recorded the arrival of seven female African lodge leaders, called in Ottoman Turkish kolbaşıs (see further below for details) who had been exiled

16  For the text of the court record, see Jennings, Christians and Muslims, 240. The case was reviewed by the Şeriat court in Nicosia in Şaban 1002, which fell between 22 April and 21 May 1594. 17  BOA/Cevdet/Zaptiye/4327.

Shifting Patterns of Ottoman Enslavement

905

from Istanbul.18 The seven women, he wrote, were accused of conducting in their lodges, where they lived, what was termed “African wedding” events. On those occasions, lodge members assembled, musical instruments were played, fire tricks were displayed, and other kinds of “abominations” and unacceptable actions were performed as part of a healing process for demon-possessed Africans. Echoing the expulsion edict, the report stated that the women were banished to Varna in order to save their communities from that evil and harm. The seven, all mentioned by name and address, were collected from different quarters of the city and the sultan’s chief halberdier appointed one of his officers to accompany them from Istanbul to Varna. They were to remain under detention in Varna and not allowed “to take [even] one step” towards another location until a further edict was issued in their matter. What, then, do these stories tell us? What connects Cyprus, Istanbul, and Varna? How came African women to end up in Bulgaria? What course of history do the stories chart for us? Do they reflect a pattern of enslavement that connected Europe and the Ottoman Empire? In order to answer these questions, we need to situate the stories in their longue durée historical context, which must begin with a brief account of Ottoman enslavement in the early modern period. In the first, Rahime, an enslaved European woman in Ottoman Cyprus, appears very much as an integral part of Ottoman Cypriot society at the end of the 16th century. There was nothing unusual about her being there, about her enjoying the trust of the district governor, and about her agility in plotting with yet another official to embezzle the money she was supposed to bring to the Nicosia treasury. Thus, for our purposes here, both the trust and its violation represent the routinization of a European enslaved presence in Ottoman Mediterranean societies; they both indicate the fact that enslavement was an integral part of a European-Ottoman relationship, which of course was far more multi-faceted and complex than just that. Indeed, it also reflected the power relations that made the presence of Rahime in Cyprus possible. The second story represents an interim position: two slave dealers were banished from Istanbul because they were alleged to have been engaged in prostituting enslaved women in private houses. The two women were from Bursa and Syria, not from European territories, but the enslaved women they were exposing to clients would now come from a mixture of sources—south-east European, Caucasian—i.e., from the Caucasus—and African. That exploited population represents the shifting patterns of trade in enslaved persons transported into the Ottoman Empire during the second half of the 18th century. 18  BOA/Cevdet/Zaptiye/Dahiliye/92.

906

Toledano

It was in that period, as mentioned above, that European sources of enslaveable persons were dwindling to the point of drying up, only to be replaced by African and Caucasian ones. The story may also serve to point out that, as we approach the modern period, while enslavement was still sanctioned by the Ottoman state, the government was already intervening to check the level of abuse. This trend would become a major feature of 19th-century reforms, in which the Tanzimat-state (1830s–1880s) entered decidedly into the enslaverenslaved relationship in order to limit the enslavers’ entitlement to the labor and body of the enslaved.19 It not only turned its attention to the enslaved, but sought also to limit abuse of the powerless in general, and protect the weak and needy in society. When we use the term “state” in the Ottoman context of the early modern period, we do not mean a well-integrated, modern entity, much in the way we think of present-day, or even late 19th-century European, states. Rather, the Ottoman variant was a “compound” polity, made up of a coalition of the interest groups that formed its imperial elite. That elite was mostly male and Muslim, multi-ethnic, kul/harem and freeborn, military-administrative-legallearned, urban and rural, officeholding and propertied, Ottoman-imperial and Ottoman-local. The center of that “composite” polity moved with changing political circumstances from the palace to the seat of the grand vezir and back, with shifting coalitions forming, collapsing, and re-forming among interest groups and leading elite members. This composite polity is reminiscent of the notion of a “classical tributary empire,” i.e., “segmented, loosely integrated, and partly overlapping forms of power and authority.”20 In a similar manner to other “societies with slaves” and “slave societies,” the Ottoman state upheld the rights of enslavers and refrained as much as possible from intervening in their relationships with the enslaved. When it intervened, this was in most cases to help enslavers recover their absconding slaves, or, conversely, to liberate severely-abused enslaved persons from abusive enslavers. Until 1845, the Tanzimat-state was also reluctant to impose its criminal system upon enslaved persons, leaving the responsibility in the hands of enslavers. However, that changed as part of the growing role the state assumed in criminal matters in general. In August of that year, while reviewing a theft case in which an enslaved African male was accused of stealing from his master, a 19  On this, see Toledano, As If Silent and Absent, Chapter Three: Turning to the “Patron State” for Redress, 108 ff. 20  See conference on “Royal Courts and Capitals,” Sabanci University, Istanbul, 14–16 October 2005, part of the COST Action “Tributary Empires Compared: Romans, Mughals and Ottomans in the Pre-Industrial World From Antiquity Till the Transition to Modernity.”

Shifting Patterns of Ottoman Enslavement

907

certain Mustafa Bey, the Council of Ministers decided that unfree offenders should be treated as free ones.21 They, too, were henceforth to be handled by the state—not by the enslavers as stipulated in the Şeriat—and receive the same penalties as free offenders. Having served their term in prison, such enslaved convicts should be returned to their enslavers, the Council ruled. The third story completes our account of the transformation that shifted the patterns of Ottoman enslavement during the early modern period. Although it took place in 1827, the trend is noticeable already during the second half of the 18th century and the case was chosen because it provides more details than earlier ones. The arrival of the seven African heads of lodges in Varna is strikingly emblematic in several senses. The women represent not merely the wellestablished presence of enslaved Africans in the Ottoman Empire, but their having deep roots in actual Ottoman societies. That communities of enslaved Africans were so entrenched in the Ottoman urban landscape by the close of the 18th century is attested to by the fact that they had established a series of lodges to care for the communal needs of both the enslaved and the freed. This was done with government support, and for the most part, lodges were allowed to exist even when various un-Islamic practices were taking root in them. The banishing of those African community leaders would happen only when orthodox pressure from Islamic clerics became persistent. Full exploitation of the available archival material remains yet to be accomplished, and the evidence for the existence of lodges involving freed and enslaved Africans comes mainly from 19th-century Istanbul, and is drawn from travel accounts and contemporary informants. That literature was studied and interpreted by Y. Hakan Erdem,22 whose summary of accounts composed by Z. Duckett Ferriman, Lucy Garnett, Richard Davey, Leyla Saz, George Young, Halide Edip, and Emine Foat Tugay yields an intriguing description of a loosely defined network of lodges, located mainly in Istanbul and other large cities. Each of these lodges was run by a female African freed slave, called kolbaşı or sometimes godya/godiya. Erdem aptly describes her functions as those of both “union leader and priestess of a religious cult.” The kolbaşı, who served in her capacity for life, was charged with caring for sick and unemployed freed Africans and placing them in gainful jobs. At times, the kolbaşı would purchase 21  İrade/Meclis-i Vala/1280/Mazbata, 23.7.1845, the Grand Vezir to the Sultan, 5.8.1845. 22   Slavery in the Ottoman Empire and Its Demise, 1800–1909 (New York: St Martin’s Press, 1996), 173–176. For a similar account of these Istanbul lodges (which ignores Erdem’s work but adds some untapped sources), see John Hunwick, “The Religious Practices of Black Slaves in the Mediterranean Islamic World,” in Slavery on the Frontiers of Islam, ed. Paul E. Lovejoy (Princeton, NJ: Markus Wiener Publishers, 2004), 160–162.

908

Toledano

the freedom of enslaved persons in special need of rescue. An Ottoman document cited by Erdem suggests that the organization was actually yet another instrument of government service and control, similar to guilds, Sufi orders, urban quarters, and other intermediary formations. For most of the time, the state turned a blind eye on Zar-Bori healing rituals, which served as an adjustment mechanism for the uprooted and re-acculturated enslaved communities in the Empire. The fascinating thing about these African cultural retentions in Ottoman societies is the way in which they were transformed by, and absorbed into, local Islamic practices. In the Ottoman core regions, the ritual performed by African lodge members consisted of dance, gestures, music, loud vocal utterances, periodic use of incense, and object worship.23 Dancing was normally done in a circle with one or two persons dancing in the middle, leading and stimulating the others round about. The instruments played were made of materials that could be found in the new environment into which the enslaved were interjected, but followed the models seen and heard in their places of origin. It appears that while such rituals resembled scenes from Sufi zikr rituals, going into individual trance was a more common occurrence in the African synthesized version. Not infrequently, participants would clean themselves before Zar rituals as before prayer (Arabic, wudu’). Following the initial stage of dancing and trance, the kudya would calm down the women, then lead the participants in reciting the Fatiha, i.e., the opening chapter (Arabic, sura) of the Koran. As the ritual progressed, the Fatiha would be recited many times, on each occasion in honor of another Islamic personage, beginning with the Prophet, his family, and various leading and local Sufi saints and other public figures. This infusion of Islamic elements into the Zar-Bori ritual is reminiscent of what Kathryn McKnight calls “African-Catholic syncretism.”24 In the African-Ottoman ZarBori ritual, invocations of the Fatiha were followed by singing and playing of 23  Summed up by Erdem, 175, and described by Boratav, 87–88. This is also stated in a court record at BOA/Cev Dahiliye/92. 24  Kathryn Joy McKnight, “‘En su tierra lo aprendió’: An African Curandero’s Defense before the Cartagena Inquisition,” Colonial Latin American Review 12/1 (June 2003), 79. Referring to 17th-century Cartagena de Indias, she describes a common phenomenon of injecting prayer components into indigenous healing practices. Specifically in the case of Mateo de Arará, brought before the Inquisition Tribunal in mid-century, the man was accused of performing a ritual considered to have been diabolical. The action was well grounded in a shamanic African ritual intended to extract from a group of people a yerbatero (a person using herbs to cause harm). This was done by uttering words and the movement of a manipulated little broom, but during the ritual Mateo invoked “the Virgin Mary and Our Lord Jesus Christ.”

Shifting Patterns of Ottoman Enslavement

909

musical instruments, to please and pacify the possessing spirits. These spirits were of several types and categories, but most of them were male. The most intriguing category for us here is the one that included contemporary figures such as various Ottoman officeholders—governors, ministers, men of religion.25 Despite all that, syncretism did not win the African lodge leaders acceptance by the official, state Islamic doctrine, hence they occasionally had to face censoring and exile, as in our third story. The circle described in this paper was thus complete with the arrival on European soil, in our case the Bulgarian town of Varna, of seven African women banished from Istanbul for having practiced Zar-Bori infused with popular Islamic rituals. By the end of the early modern period, Africans constituted the vast majority of the enslaved population in the Ottoman Empire. Women were the overwhelming majority among the enslaved, and most of them served in urban elite households as domestic servants. Within about a century and a half, African women—and to a lesser extent, also Circassian and Georgian ones—replaced European men in the Ottoman servile workforce, and agricultural enslavement gave way to household and menial servitude. All that is left for us to do now is to explain how and why it all happened. As already mentioned above, the main source for enslaved persons at the beginning of the early modern period was war captives. Both the state and individual officers on the battlefield appropriated the defeated enemy’s manpower accessible to them, and then redistributed it on the domestic unfree market. Most of the captives came from south-east European powers, including present-day Russia and Ukraine, but on the eastern fronts, Iranian and Caucasian captives also fell victim to war-related enslavement. Owing to the end of Ottoman expansion in Europe towards the beginning of the 18th century—except for few and brief interludes—the supply of enslaveable European men and women dried up. Similarly on the Iranian front, the 1736 treaty ending the Ottoman-Iranian war also provided for the release of war captives by the Ottomans and the repatriation of enslaved Iranian subjects (which proved quite difficult to enforce internally).26 With occasional Crimean Tatar raids into Russian and Ukrainian territories, and with Ottoman activities to stabilize control over the western Caucasus, those areas now replaced the former sources for unfree labor in the Empire. This changing trade pattern can also be seen in the geographic origins of the slave dealers who carried out the traffic. Faroqhi argues that in the second half of the 18th century, many of them came from central and east Anatolian towns, 25  Sengers, 69–71, 259–263. 26  On this, see also Faroqhi, Stories of Ottoman, 101.

910

Toledano

indicating that the importation areas had shifted to the Caucasus and the regions north of the Black Sea.27 But the most significant transformation of the trade patterns in enslaved persons reaching southwards from European and eastern sources—it was the rise, gradual at first, of African lands as the main enslaveable human reservoir for Ottoman markets. This trend peaked towards the middle decades of the 19th century, but it was already in evidence during the last quarter of the 18th.28 As our third story clearly shows, communities of enslaved Africans existed already then on Ottoman soil. The interesting point, however, is why free labor had not replaced servile, in contrast to what had happened when agricultural enslavement disintegrated in the 16th century. The reason lies, for the most part, in the structure of the Ottoman-imperial and Ottoman-local elites, at whose heart lay the Ottoman elite household.

Household, Status, and Demand: Economic or Socio-Cultural?

During the 17th and 18th centuries, the household emerged as the basic unit of belonging or attachment throughout the Ottoman lands. Although households surely existed before that period, they nonetheless came to play a distinct role in Ottoman societies as a result of the large-scale transformation that took place in the Empire from the end of the 16th century onwards. We need not go into the various reasons that caused the transformation, as they have been amply discussed in the literature.29 Suffice it here to note that a dual process of 27  Ibid., 259–260. 28  See, for example, Dr Frank’s eye-witness account of the trans-Saharan slave trade into Egypt at the close of the 18th century: Michel Le Gall, “Translation of Louis Frank’s Mémoire sur le commerce des nègres au Kaire, et sur les maladies auxquelles ils sont sujets en y arrivant (1802),” in Slavery in the Islamic World, 69–88. 29  The main contributors to the debate over the transformation of the Empire’s governance in that period are Islamoğlu and Keyder, “Agenda for Ottoman History,” in The Ottoman Empire and the World Economy, ed. Huri Islamoğlu-Inan (Cambridge: Cambridge University Press, 1987), 42–62; Roger Owen, The Middle East and the World Economy, 1800–1914, rev. ed. (London and New York: I.B. Tauris, 1993) (first ed., 1981), “Introduction: The Middle East economy in the period of so-called ‘decline’, 1500–1800,” 1–23, 294–299 (notes); Metin Kunt, All the Sultan’s Servants: the Transformation of Ottoman Provincial Government 1550–1650 (New York: Columbia University Press, 1983); Suraiya Faroqhi, “The Ruling Elite between Politics and ‘the Economy’,” in An Economic and Social History, 545–575 (especially 552–556); Rifaat Ali Abou-El-Haj, Formation of the Modern State: the Ottoman Empire, Sixteenth to Eighteenth Centuries (Albany, NY: State University of New York Press, 1991); Jane Hathaway, The Politics, 1, 14, 24 (and throughout the book); idem,

Shifting Patterns of Ottoman Enslavement

911

localization and Ottomanization was taking hold in the provinces, producing Ottoman-local elites throughout the Empire.30 In this process, the Ottoman imperial elite was becoming less mobile, with posts being assigned within limited regions, so that specialization according to needs of specific provincial “clusters” were developing within the military and the bureaucracy. The local elites developed strong ties to the local economy, society, and culture, and linked their and their children’s future to one province, often to one city. At the same time, local elites—urban and rural notables, ulema, and merchants— were seeking to become part of the imperial administration, trying to attain government offices and being Ottomanized in the process. The localizing imperial elite and the Ottomanizing local elites gradually merged into Ottomanlocal elites, which better served the interests of both sides. Of major importance in this process was the household, or rather more specifically in our case, the Ottoman-local household, which served as the social, economic, political, and even cultural unit that facilitated and promoted Ottoman-local integration. In the 17th century, households were being created around leading officeholders in the bureaucracy and within the military. While forming initially around the nuclear and extended family of the founder, from the outset they relied on patronage relationships between the head of the household and a broad array of clients. The essential components of any household were the founder’s retainers, who were a sort of militia force, often small, armed, and protecting the interests of the household. A concomitant component was the producers of the household wealth, which enabled its expansion through recruitment and networking. Marriage among the various households was another essential element for forming inter-household alliances to promote common causes and take over income-producing economic assets. Conjugal arrangements provided the socio-political cement that bonded household coalitions, or factions as they were often called, and made networking possible. Although essentially a constructed notion, households also had a physical dimension, located in estates, often one or several complexes A Tale of Two Factions, 4–6; and Oktay Özel, “Population Changes in Ottoman Anatolia During the 16th and 17th Centuries: The ‘Demographic Crisis’ Reconsidered,” IJMES 36/2 (May 2004): 183–205, on demographic and economic pressures during the 16th and 17th centuries. This debate has consequently helped to revise the ‘decline paradigm,’ which is now virtually defunct in Ottomanist discourse, though unfortunately still quite alive outside the field of Middle East Studies, and even to some extent within sections thereof. 30  The arguments put forth in the following paragraphs are fully developed in my forthcoming book on Ottoman-local elites, but in part have already been published in Toledano, “The Emergence of Ottoman-Local Elites,” 145–162.

912

Toledano

that housed dozens of individuals, hundreds in the larger ones, who performed a wide variety of functions including command and control, enforcement, finances, service, and trade. Household heads vied for resources and coalesced locally, usually in a provincial town, where the local governor resided, but they soon realized that it was essential to build a network that would transcend subdistrict, district, and even provincial bounds. Truly successful household coalitions had to be also connected to the imperial capital, where officeholders were consolidating their patronage networks through that new-type household. In the 17th century, the crucial stepping-stone towards household consolidation—i.e., its social reproduction—was the ability to survive the founder’s demise, that is to entrench a multi-generational structure. Until the latter part of the century, many households disintegrated at that stage, leaving the provincial scene to new households and new factions. However, gradually, some households and some factions proved more resilient, better adapted to the changing circumstances at the center and the provinces, and more capable of sustaining the incessant competition over resources. By the end of the first quarter of the 18th century, in provinces throughout the Empire, a single household, or rather faction of households, usually emerged as hegemonic, securing for its leader and his lieutenants near-full control of the body politic and the economy. The main offices of state, hence also access to, and appropriation of, the main incomegenerating assets, would fall into the hands of members of that household. All this occurred in a world of intense political struggles that were led and directed—through active networking and by skillful deployment of balancing acts—by members of the imperial and Ottoman-local elites: men and women, both kul/harem and freeborn, both in Istanbul and in the provinces. Among the most famous of these households we may mention the Kazdağlıs of Egypt, the Eyübizades of Iraq (mainly in Bagdad and Basra), the Azms of Syria, the Husaynis of Tunis, and the Karamanlıs of Libya. In some cases, hegemonic households would turn in the 19th century into local dynasties (Egypt, Tunisia), in others the Tanzimat-state would remove them and take their place (Syria, Iraq). Ottoman political culture was heavily influenced by patterns that evolved during the last two and a half centuries of imperial rule, leaving their mark on political interaction in the successor states of the Middle East for decades after the Empire’s demise. The strong link between political and economic interaction, the belief in diversification through placement of family members in competing networks to minimize risk and increase security, the overwhelming impact of patronage politics, the lingering effect of “grandee families,” and the presence of both formal and informal dynastic orders are some of the salient features that the Ottoman system bequeathed to modern Middle Eastern and North African societies.

Shifting Patterns of Ottoman Enslavement

913

For our purposes here, it is important to examine the ways in which households recruited and socialized new members. The purchase of enslaved persons for various roles was one of the four most important channels of recruitment to imperial-center and Ottoman-local households. The other three modes of recruitment-cum-bonding to a household were biological-kin relationships, marriage, and voluntary offer of loyalty and services in return for patronage. Less prevalent were adoption and suckling relationships, but the sources occasionally do mention them, too. Bonding ensured that loyalty and patronage would flow from top to bottom and from the bottom up in households across Ottoman societies, linking people from various elites to non-elite groups and individuals. In that way, society was cohesively undergirded both vertically (within a household) and horizontally (alliances among households). Not infrequently, individuals were bonded to a household through more than one of these ties, as, for example, when one was the purchased kul-type (Arabic, mamluk) retainer of the household head and was also married to his daughter. Attachment to a household gave an individual protection, employment, and social status. But not less significantly, it gave household members (kapı halkı) a sense of belonging and an identity, both social and political. Thus, this system, which encompassed Ottoman political, social, and economic life, kept alive the demand for enslaved persons, whether as its kul/harem leadership or its household workforce.31 To be sure, already during the early modern period, much of the labor in elite households was free, not bonded, but the remaining servility was still so engrained in the system and so indispensable culturally, i.e., mainly as status symbol, that it ensured the continuation of demand for enslaved persons in the Ottoman Empire well into the 19th century. Despite the major reforms of the Tanzimat, and the prohibition of the slave trade—enforced in earnest only in the last quarter of the century—slavery was not legally abolished until the demise of the Ottoman Empire and the rise of the Turkish Republic.32 So, we end up with an open question: what was the value for Ottoman enslavers of sustaining that constant demand for enslaved labor in urban elite households? It is doubtful that the real reason was economic: there was a ready and active market for free labor in domestic service, in fact it constituted the lion’s share of the labor market as a whole. As previously observed, in the Ottoman social structure of the early modern and modern periods, the basic building 31  For a similar view, though cast in somewhat different terms, see Faroqhi, Stories of Ottoman, 149. 32  On the last phase see Erdem, and on the suppression of the traffic, see Toledano, Suppression.

914

Toledano

block of social, political, and economic networking was the elite household, both at the imperial center and in the provinces. Within that system, elite households emulated the sultan’s household-court as much as they could afford to. As long as an essential component of the sultan’s household-court continued to include kul/harem and domestic enslavement, so did—albeit to a lesser degree—the households of his military-administrative officeholding elite, which served him and his composite state. As long as the size of harems, the presence of eunuchs, and the pattern of socio-political marriages within the elite persisted, the demand for enslaved persons remained stable. In that sense, we can say that what perpetuated Ottoman enslavement throughout the early modern period and into the modern era was its socio-cultural value, not its economic worth.

Ransom Slavery along the Ottoman–Hungarian Frontier in the Sixteenth and Seventeenth Centuries Géza Pálffy Introduction In 1682 the military court (sedes bellica) of the border fortress of Fülek convened to discuss a peculiar matter. Twelve prisoners who had been released on suretyship from the Turkish dungeons in Várad, had requested the professional opinion of the court. As their petition stated, while the twelve men had been begging together their ransoms, their fellow prisoners—who were their guarantors—had dug themselves through the walls of the prison and made their escape. Some of them had never been caught; but others had been recaptured by the Ottomans and returned to their dungeon cells. The twelve prisoners on release were now asking the military court whether or not the recaptured prisoners continued to be their “bailsmen”. Ever since the Treaty of Adrianople (Edirne) in 1568, the Ottoman–Hungarian peace treaties had provided that neither party was required to return prisoners freed with the help of God. However, in this particular case the military court decided that the recaptured Christian prisoners were not new prisoners but old ones who had been returned to their cells. Thus, from this point on, only the recaptured persons could be considered to be the guarantors of the twelve prisoners on release. When issuing its ruling, the military court pointed out that it had deliberated on the matter “by referring to examples of some ordinary things and by recalling and complying with the old custom existing between the Hungarian and Turkish border fortresses.”1

S ource: Pálffy, Géza, “Ransom Slavery along the Ottoman–Hungarian Frontier in the Sixteenth and Seventeenth Centuries,” in Géza Dávid and Pál Fodor (eds.), Ransom Slavery along the Ottoman Borders: Early Fifteenth – Early Eighteenth Centuries, Leiden: Brill, 2007, 35–83. 1  M OL Esterházy cs. lvt. [Archives of the Esterházy family] P 125, Pál nádor [Palatine Paul Esterházy] No. 11.378. The affair is mentioned in Géza Pálffy, Katonai igazságszolgáltatás a királyi Magyarországon a XVI–XVII. században [Military Jurisdiction in Habsburg–Hungary in the Sixteenth and Seventeenth Centuries]. Győr, 1995, 144.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_031

916

Pálffy

Although research into the Habsburg–Hungarian system of border fortresses and Ottoman–Hungarian relations has produced some fine results, a comprehensive and detailed elaboration of the customs in the marches (Grenzbräuche),2 including an exposition of the old order existing between the Hungarian and Turkish border castles cited above, is still lacking. Even the Hungarian historian Sándor Takáts, who knew much of the frontier warriors and the ways in which Ottomans and Hungarians coexisted, failed to treat the subject in a systematic manner.3 Thus, while his works contain valuable and useful data, they have never been presented in monographic form. Furthermore, his essays on the subject were greatly influenced by the proTurkish and anti-Habsburg attitudes that largely determined—and sometimes even distorted—Hungarian historiography for many years. In defence of more recent researchers, we can state that they faced the difficult task of reconstructing a system of common law of the border fortresses which—unlike the customary law of the nobility as formulated in István Werbőczy’s Tripartitum—was never codified in writing. Nevertheless, a reconstruction of these customs can be instructive, for these traditions existed only temporarily (for a period of 300 years) and were soon forgotten after Ottoman rule ended in Hungary in the late seventeenth century. This article does not attempt a systematic survey of all the related issues, it merely tries to bring together some important elements of Ottoman–Hungarian common law and customs along the borders, including the characteristic features of slavery and the trade in captives, based on new documentary evidence from the Budapest and Viennese archives.4 2  For example in his decree of November 17, 1652, King Ferdinand III (1637–1657) ordered Ádám Forgách, border fortress captain-general of Érsekújvár, to appoint Miklós Révay to the office of vice-captain dem Gräniczbrauch gemeß. MOL Forgách cs. lvt. [Archives of the Forgách family] P 287, Series II. Fasc. HH. (42. cs.) fols. 288–289. 3  Sándor Takáts, “Török–magyar szokások a végekben [Ottoman–Hungarian Customs in the Marches]”, in Idem, Rajzok a török világból [Sketches from the Turkish World]. II. Budapest, 1915, 213–238. 4  For previous literature, see Néda Relković, “Embervásár a török időkben [Slave Market in the Ottoman Period],” Századok 44 (1910) 113–121. Sándor Takáts, “Magyar rabok, magyar bilincsek [Hungarian Captives, Hungarian Shackles],” Századok 41 (1907) 415–435, 518–540. Idem, “A török és a magyar raboskodás [Ottoman and Hungarian Captivity],” in Idem, Rajzok a török világból [Sketches from the Turkish World]. I. Budapest, 1915, 160–303 (this is a study abundant in valuable data, but it does not provide a systematic description of the ransom taking system, and its evaluations sometimes seem to be too naive). Aleksandar Solovjev, “Trgovanje bosanskim pobljem do god. 1661.,” Glasnik Državnog muzeja u Sarajevu, Društvene nauke. Nova serija 1 (1946) 139–162. József [László] Kovács, “Török rabszolgák Sopronban,

Ransom Slavery Along the Ottoman–Hungarian Frontier

917

soproniak török rabságban a 17. század folyamán [Ottoman Slaves in Sopron—Citziens from Sopron in Ottoman Captivity in the Seventeenth Century],” Soproni Szemle 13:1 (1959) 6–11. Edit Izsépy, “Rablevelek a váradi török börtönből [Letters of Captives from the Turkish Prison of Várad],” Az Egyetemi Könyvtár Évkönyve 5 (1970) 315–327. Gábor Dobos, “Török– magyar rabok a nyugat-dunántúli végeken [Ottoman–Hungarian Captives in the Western Transdanubian Marches],” Studium II. Acta Juvenum Universitatis Debreceniensis de Ludovico Kossuth Nominatae. Debrecen, 1971, 63–73. Sergij Vilfan, “Die wirtschaftlichen Auswirkungen der Türkenkriege aus der Sicht der Ranzionierung, der Steuern und Preisbewegung,” in Die wirtschaftlichen Auswirkungen der Türkenkriege. Die Vorträge des 1. Internationalen Grazer Symposions zur Wirtschafts- und Sozialgeschichte Südosteuropas (5. bis 10. Oktober 1970). (Grazer Forschungen zur Wirtschafts- und Sozialgeschichte, 1.) Ed. by Othmar Pickl. Graz, 1971, 177–199. Edit Izsépy, “Az egri törökök fogságába esett magyar rabok kiváltásának és szállításának problémái [Problems of Liberation and Transport of Hungarian Captives Captured by the Turks of Eger],” Agria: Annales Musei Agriensis 11–12 (1974) 159–169. Pavo Živković, “Mletačka trgovina bosanskim robljem u srednjem vijeku,” Godišnjak Društva Istoričara Bosne i Hercegovine 21–27 (1976) 51–58. Danilo Klen, “Pokrštavanje »turske« djece u Rijeci u XVI i XVII stoljeću,” Historijski zbornik 29–30 (1976–1977) 203–207. Bogumil Hrabak, “Skopskiot pazar na robje vo XV i XVI vek,” Glasnik. Institut za nacionalna istorija 24 (1980) 151–161. László Fenyvesi, “Az igali portya és a körmendi kótyavetye balkáni tanulságai. (Adalék a hódoltsági rác-vlach-iflák-vojnik problematikához, 1641) [The Balkan Aspects of the Hungarian Raid on Village Igal and the Auction in Körmend (Some Data to the Problem of Serb, Vlach, Eflak, and Voynuk Population in Ottoman Hungary)],” in Magyar és török végvárak (1663–1684) [Hungarian and Ottoman Border Fortresses (1663–1684)]. (Studia Agriensia, 5.) Ed. by Sándor Bodó—Jolán Szabó. Eger, 1985, 199–218. János J. Varga, “Rabtartás és rabkereskedelem a 16–17. századi Batthyány-nagybirtokon [Keeping of and Trade in Captives on the Batthyány Estates in the Sixteenth and Seventeenth Centuries],” in Unger Mátyás Emlékkönyv [Festschrift for Mátyás Unger]. Ed. by E. Péter Kovács – János Kalmár – László V. Molnár. Budapest, 1991, 121–133. Idem, “Gefangenenhaltung und Gefangenenhandel auf dem Batthyány-Grundbesitz im 16.–17. Jahrhundert,” Burgenländische Heimatblätter 4 (1995) 145–162. Hajnalka Tóth, “Török rabok Batthyány I. Ádám uradalmaiban [Turkish Captives on the Estates of Ádám Batthyány I],” Aetas 2002/1, 136–153.—For the most important captive letters, songs, and diaries, as well as other documents relating to the keeping of captives, see Fr[anjo] Rački, “Dopisi izmedju krajiških turskih i hrvatskih častnika,” Starine 11 (1879) 76–152; 12 (1880) 1–41. Pál Jedlicska, “XVI. századi török–magyar levelek Pálffy Miklóshoz [Ottoman–Hungarian Letters to Miklós Pálffy in the Sixteenth Century],” Történelmi Tár (1881) 691–705. Farkas Deák, “Okiratok a török–tatár rabok történetéhez [Documents on the History of Turkish and Tatar Captives],” Történelmi Tár (1886) 110–126. Géza Dongó Gy[árfás], “Magyar rabok és rabnők török–tatár fogságban [Hungarian Captives in Turkish and Tatar Captivity],” Adalékok Zemplénvármegye történetéhez 16 (1910) 87–91, 152–155, 281–284, 331–333; 17 (1911) 78–81, 179–181, 314–315; 18 (1912) 88–91, 230–231, 289–291; 19 (1913) 76–77, 162–163, 249–251, 333–335; 20 (1914) 62–64, 169–171, 263–266; 21 (1915) 81–83, 146–148, 263–266; 22 (1916) 50–52. Auer János Ferdinánd pozsonyi nemes polgárnak héttoronyi fogságában írt naplója 1664 [The Diary of Johann Ferdinand Auer, Citizen of Pozsony Written during his Captivity in the Yedikule 1664]. (Fontes historiae Hungaricae aevi Turcici, 1.) Ed. by Imre Lukinich. Budapest, 1923.

918

Pálffy

Customs of Slavery along the Frontier during the Ottoman Rule

Trade in slaves constitutes a particularly interesting and vivid chapter in the history of Ottoman–Hungarian relations. Two main types of historical periods may be distinguished in the course of a development that lasted for centuries. Periods belonging to the first type were characterised by military offensives, while more peaceful years belonging to the second type were characterised by raids and skirmishes into enemy territory. During periods of the first type, e.g. the sixteenth-century military offensives of Sultan Süleyman the Magnificent (1522–1566) or the great war against the Ottomans in the late seventeenth century (1683–1699), both sides captured large numbers of men— who were then sold off at distant slave markets (Sarajevo, Istanbul) or purchased by Austrian noblemen and Italian merchants (Venetians, Florentines, etc). During periods of the second type, comprising much of the history of Ottoman–Hungarian relations (e.g. 1466–1520, 1568–1591, 1606–1663), the two sides traded captives for ransoms. This trade was a direct consequence of the coexistence of the two powers in the border zone and was closely connected to the incursions (perpetrated by both sides) into enemy territory. This phenomenon was not unique to Hungary. A similar development may be observed in other frontier areas of the Ottoman Empire such as North Africa and the Polish–Russian region, as well as in the Mediterranean.5 Ransom slavery along the Ottoman–Hungarian frontier reached its peak during the long period of peace (1606–1663). Indeed, it would seem that this method of acquiring and keeping captives became a profitable business amid the chaotic conditions created by the Fifteen Years’ War (1591/93–1606). A similar development can be observed during the earlier Turkish wars of the sixteenth century (1526–1566) and the Christian war of re-conquest in the late seventeenth century (1683–1699). Given the incessant warfare and large numbers of uprooted people, including inhabitants of the “militarised” market towns (oppida) and settlements as well as the destitute soldiers of the border fortresses, it was not difficult to find armed men who would take captives during raids or kidnap men, women, and children for money. This is well demonstrated by a special contract drawn up in the summer of 1655 between Márton Toldy, juryman (iuratus assessor) of Veszprém county, and the warriors of the border fortress of Veszprém which stipulated if “God grants them Wathay Ferenc énekes könyve [Songbook of Ferenc Wathay]. I–II. Ed. by Lajos Nagy. Facsimile edition. Budapest, 1976 and Rabok, követek, kalmárok az oszmán birodalomról [Captives, Envoys, and Merchants on the Ottoman Empire]. Published by Lajos Tardy. Budapest, 1977. 5  See with more literature Mária Ivanics’s and Pál Fodor’s studies in the present volume.

Ransom Slavery Along the Ottoman–Hungarian Frontier

919

fortune and they bring from my property (an estate in Tolna county under Ottoman rule), which I have given to them, people, cattle, horses or cash, half shall belong to the warriors, and half shall be mine. And if the warriors shall feel so inclined, they may receive money for their service, and the captives shall be mine. Given the first condition, they shall not select the best of the captives, but shall bring them here, where they will be divided into two.”6 On the Ottoman side such kidnappers were mainly the martaloses7 “assisted” in ever growing number by the so-called pribeks.8 By this time, members of the two groups were involved less in the supply of slaves to distant markets than in the kidnapping of men for ransom. Even though the Hungarian border captains had ordered that men who settled on enemy territory in order to make a living from kidnapping should be severely punished, more than a few men did chose this way of enriching themselves. If they subsequently fell into Christian hands, two possible fates awaited them. Where border captains applied the rules strictly, on the basis of chapter 15 of part I of István Werbőczy’s Tripartitum9—see, for example, the case of the nobleman Pál Soltész, who was sentenced by the military court (sedes bellica) of Upper Hungary in 1665 and who “of his own freewill and under no duress, while abandoning his Christian faith, had become a Turkish pribek … (and then) 6  M OL Fördős cs. lvt. [Archives of the Fördős family] P 1754, Tétel 59. fol. 5. An excellent indication of the extent of the kidnappings for ransoms is provided by a report of damages stating that between 1633 and 1649 the Ottoman border fortress soldiers captured or killed 4,207 persons in the Kanizsa district alone, while they also took 4,760 cattle. At the same time, however, they burnt “only” 66 houses to the ground. MOL Batthyány cs. lvt. [Archives of the Batthyány family] P 1313, Török vonatkozású iratok [Documents pertaining to the Turks] No. 230. fol. 173. 7  With more literature, see Milan Vasić, Martolosi u jugoslovenskim zemljama pod turskom vladavinom. (Akademija nauka i umjetnosti Bosne i Hercegovine, Djela XXIX; Odjeljenje istorijsko-filoloških nauka, 17.) Sarajevo, 1967. 8  The word is of South Slavic origin (cf. pribegnuti = refuge) and means “refugee”. Contrary to the view of Sándor Takáts (“A pribékek [Pribeks],” in Idem, Rajzok, I. 310–311) and in agreement with the findings of Ferenc Szakály (Mezőváros és reformáció. Tanulmányok a korai magyar polgárosodás kérdéséhez [Market Town and Reformation. Studies in the Early Hungarian Burgher Development]. [Humanizmus és reformáció, 23.] Budapest, 1995, 260: note 227), we should consider the henchmen to have been first and foremost kidnappers and marauders rather then “professional” spies. 9  This chapter of the Tripartitum provided for the hanging of thieves and the implement or breaking on the wheel robbers. The Customary Law of the Renowned Kingdom of Hungary: A Work in Three Parts Rendered by Stephen Werbőczy (The “Tripartitum”) (Decreta regni mediaevalis Hungariae, Series I, vol. 5.) Ed. by János M. Bak – Péter Banyó – Martyn Rady. Budapest, 2005, 68–69.

920

Pálffy

along with other henchmen had taken Christians captives, bringing them to the Turkish fortress Eger”—such brigands could expect to be mercilessly tortured and then impaled, which was the usual punishment for such crime.10 Sometimes, however, commanders might choose to spare a captured informer for themselves if he promised to refrain from kidnapping activity and to work instead as an informer, or more accurately as a double agent. Although the captains had been promised by the ruler a reward of 24–30 florins for each pribek executed,11 they would show mercy if there was some hope of securing even greater booty in future raids based on the information gained. The upsurge in ransoming and kidnapping activity in the seventeenth century clearly indicates the potential of the frontier slave trade. The beginnings of this peculiar form of commerce go back to the fifteenth and sixteenth centuries and its significance is exemplified by the fact that members of the illustrious Batthyány family levied ransoms on their Turkish captives amounting in total to several tens of thousands of florins, although we know that these ransoms were rarely or only partially remitted.12 Indeed, later on, the provision of board and lodging for the many captives had become a serious problem for the prison warders of the fort owned by the Batthyánys in Németújvár, even though several other rooms had been arranged for the men in buildings close to the dungeons.13 The following table indicating the number of Christians held captive in the largest prison of Ottoman Hungary, the so called Csonka Torony, or “Stump Tower” in Buda14 demonstrates the culmination of this development in the middle of the seventeenth century.

10  MOL Csáky cs. lvt. [Archives of the Csáky family] P 71, Fasc. 264. Tétel 5. Sedes bellica, October 19, 1665, Kassa. Cf. also Dongó, “Magyar rabok,” (1915), 263–266: No. XLIX. 11  Takáts, “A pribékek,” 310–311. 12  Varga, “Rabtartás és rabkereskedelem,” 126. Some similar and mainly Slavonian examples: Vilfan, “Die wirtschaftlichen Auswirkungen,” 183–192. Ransom slave trade developed into a very profitable business for the Ottomans, too. For example, between 1644 and 1647, the cash ransoms of the 178 Christian soldiers captured from the border fortresses in the Kanizsa district amounted to more than 64,600 florins. Varga, op. cit., 133: note 40. 13  MOL Batthyány cs. lvt. P 1314, Missiles Nos. 41.223, 41.225 and 41.258. Cf. also Ödön Kárffy, “Hírek a kanizsai török rabokról [News about the Ottoman Captives in Kanizsa],” Hadtörténelmi Közlemények 13 (1912) 473: No. I (1563). 14  Figuring as carcer Budimensis in Latin documents issued by the Ottomans, as turris, quem curtam et intectam vocant in Latin, as gestutzter Turm in German and as kula in Ottoman texts; cf. MOL Batthyány cs. lvt. P 1314, Missiles No. 9294. Szakály, op. cit., 277: note 283. Auer János naplója, 120 and MOL Batthyány cs. lvt. P 1314, Missiles Nos. 9290, 9305, 26.681, 31.526 and 48.556.

Ransom Slavery Along the Ottoman–Hungarian Frontier Table 1

921

The number of Christian captives in the Stump Tower in Budaa

Year

Number of captives

1636 1650 September 8, 1652 1652 June 14, 1657 October 4, 1657 1658 1661

57 107 45 40 224 220 180 235

a  For the sources of this table, see: MOL Batthyány cs. lvt. P 1314, Missiles Nos. 9277, 9282, 9284, 50.740, 7763, 9289, 9292, 9297–9298, 9302, 9305 and Imre Nagy, “A budai csonkatorony pecséte [The Seal of the Stump Tower of Buda],” Századok 2 (1868) 661–662 and Lajos Fekete, Budapest a törökkorban [Budapest during the Ottoman Rule]. (Budapest története, 3.) Budapest, 1944, 161. In 1687, there were 300 Christian prisoners in the fortress Eger, while in 1679 70 captives were being held at Székesfehérvár. Izsépy, “Az egri törökök,” 159 and Károly Jenei, “Iratok Fejér megye török hódoltságkori történetéhez [Documents on the History of Fejér County during the Ottoman Rule],” Fejér Megyei Történeti Évkönyv 6 (1972) 206: No. 34/a.



The Acquisition and Public Auction of Captives

In times of peace the greatest opportunities for the acquiring of captives were provided by the raids into enemy territories, the aims of which also included the taxation and plundering of the land, the stealing of local cattle and horse stocks. We know much about such practices,15 but a special type of action mentioned in the contemporary sources does require some explanation, if only because it seems to have been one element of the wider system of customs connected with the raids. Such actions are designated in Latin by expansis (explicatis) vexillis or in Turkish by the less accurate phrase bayraklar 15  Sándor Takáts, “A török portya és a magyar portya [The Ottoman Raid and the Hungarian Raid],” in Idem, Rajzok, I. 336–358 and recently Lajos Gecsényi, “A végvári harcok taktikája. (Török lesvetés Győr alatt 1577–ben) [Tactics of the Frontier Wars (Ottoman Raid on Győr in 1577)],” in Scripta manent. Ünnepi tanulmányok a 60. életévét betöltött Gerics József professzor tiszteletére [Studies in Honour of József Gerics on the Occasion of his 60th Birthday]. Ed. by István Draskóczy. Budapest, 1994, 165–175.

922

Pálffy

ile or ve bayrak ve borular ile; they were attacks carried out with “unfurled flags” which proved very lucrative in terms of the numbers of captives taken and the booty acquired.16 In the course of the attacks, the raiders—Ottomans and Hungarians alike—set out openly, without disguising their plans, for enemy territory in great numbers. Carrying their respective flags, they proceeded to plunder the land. Such actions went beyond the scope of ordinary raids; they could have been interpreted as open violations of the peace treaties.17 Christian and Ottoman subjects (soldiers and civilians) who were captured during the raids were taken to prisons located on the territory of the opposing empire. A captive’s fate was primarily determined by his family background or by his rank and importance in the military hierarchy.18 The so-called “major captives” from the Hungarian side were first and foremost members of the greater, middle, and sometimes lesser nobility (e.g. Ferenc Bebek or Ferenc Wathay) or military officers of senior rank (the border fortress generals, captains, and senior officers, e.g. János Krusics, captain-general of the border fortresses guarding the mining towns along the river Garam, and Mihály Sárközy, captain of Ajnácskő) as well as “political prisoners” (e.g. Bálint Török or István Majláth). As the “Emperor’s slaves” and gifts to the sultan, they were mainly taken to Constantinople, or sometimes to the dungeons of one of the governors (beylerbeyi) in Hungary (e.g. the Stump Tower in Buda). During the first century of the Ottoman rule, Hungarian prisoners transported to the sultan’s capital would be enjailed in the so-called Black Tower of Galata in the northern part of the imperial city (known in Turkish as Galata kulesi and in Latin as turris Maris Nigri and described in Hungarian as galatai nagy torony, Konstantinápoly ellenébe, or az Fekete-tenger mellett az Fekete Torony, Konstantinápolyon kívül, Galata városában),19 or in the Rumeli hisarı, which had been built by Mehmed 16  Damásd, 1640: MOL Magyar Kancelláriai Levéltár [Archives of Hungarian Chancellery], A 97, Hungarica et Transylvanica, Rácz Károly halálával átvett iratok [Legacy of Károly Rácz], 22. cs. fol. 116. Examples concerning the Ottomans: Ludwig Fekete, Türkische Schriften aus dem Archive des Palatins Nikolaus Esterházy 1606–1645. Budapest, 1932, 138– 139: No. 45 and Kiskomárom, 1644: MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 101. 17  The letter of Sahin Hasan, beylerbeyi of Kanizsa, to Ádám Batthyány: Ibid., No. 75/c. 18  In addition to Takáts, “A török és a magyar raboskodás,” passim, see also Rabok, követek, kalmárok, 27–28. 19  References in chronological order: İslâm Ansiklopedisi. V/2. Istanbul, 1993, 1214/120 and 151; 1556: MOL Magyar Kincstári Levéltárak [Hungarian Treasury Archives], MKA Nádasdy cs. lvt. [Archives of the Nádasdy family] E 185, Letter of György Sennyey to Tamás Nádasdy, October 18, 1556, Constantinople; 1604: Wathay Ferenc énekes könyve, I. 4 and 1562: Takáts, op. cit.,” 176.

Ransom Slavery Along the Ottoman–Hungarian Frontier

923

the Conqueror (1451–1481) to control the Bosphorus. Christian prisoners referred to this place in their letters as Újvár i.e. New Castle. (In Turkish it was known as Yeni hisar/Jaynyzzar or as Bogaz kesen, in Hungarian as Az Bokoz keze nevű torony az Fekete-tenger parton, and in Latin as castellum novum non procul a Constantinopoli, novum castrum penes Nykra mare or turris Bokoz-Kezy penes Nykra mare).20 In addition, prisoners were also kept in St Paul’s prison also located in Galata (des Turckhischen Khaysers Gefenckhnuss zu Sanndt Pauls, zu Galattha),21 while some “political prisoners” were incarcerated in the “Seven Towers” located on the south-western outskirts of Istanbul (Jedikule/Jedikula, known in German as Schloss Jedi Kula or Sibentürme/Siebenthürm)22 as early as the sixteenth century. However, at that time, this latter building served primarily to house the state treasury; it was only in the course of the following one hundred years that it became a notorious prison for Christian captives. At the same time, Hungarian prisoners were also put in the imperial arsenal prison (known in Turkish as Baba Cafer zindanı, in Italian as Bagno).23 As against their comrades in jails of Ottoman Hungary, prisoners held here were unlikely to be granted freedom, owing to their distinguished backgrounds or else for political reasons. Less valuable captives who ended up in Istanbul were rarely

20  For the various Turkish names: İslâm Ansiklopedisi, V/2. 1214/44; the distorted form of the year 1550: Georgius Pray, Epistolae procerum regni Hungariae. II. Pozsony, 1806, 203–204: No. 89; 1563: MOL Zichy cs. lvt. [Archives of the Zichy family] P 707, Missiles No. 8375 (an unpublished letter of György Bebek from his prison cell); 1545: Takáts, “A török és a magyar raboskodás,” 171: note 3. 1562: Takáts, “Magyar rabok, magyar bilincsek,” 432: note 4 and 1563: Ibid. 21  In 1563 Wolfgang Schreiber was imprisoned here: Documente privitóre la istoria Românilor. II/I. 1451–1575. Cu portretul lui Iacob Heraclid despot voevod. Ed. by Eudoxiu de Hurmuzaki. Bucuresci, 1891, 468–172: No. CCCCXXIX. For Schreiber’s imprisonment, see recently: Szakály, op. cit., 87–88. 22   Auer János naplója, 199. According to Hans Dernschwam, in the mid-sixteenth century Bálint Török, István Majláth and László Móré were held captive here, although at the same time many Christian prisoners lived in Galata, in the tower by the sea. Hans Dernschwam, Erdély – Besztercebánya – Törökországi útinapló [Diary of Travels in Transylvania, Besztercebánya and the Ottoman Empire]. Ed. and translated by Lajos Tardy. Budapest, 1984, 202, 244. 23   Sein wir untterhalb des Arsenals ausgesetzet und in die aldaige kays. gemeine Gefängnus, Bagno genandt, welches in 2 Theil getheilt, nämlich der Russen, und Frenken Bagno, worin über 2000 und mehrers Gefangene sih befinden, geführet (1663). Auer János naplója, 187–189 and 209–210: No. 1. Cf. Takáts, “Magyar rabok, magyar bilincsek,” 418.

924

Pálffy

held with those for whom large ransoms were expected. Instead they were sold immediately as ordinary or galley slaves.24 It is not surprising therefore that Christian prisoners greatly feared being sent to the Ottoman capital city. One such prisoner, Gyurka Horvát from Magasi in Vas county, who was held in the Stump Tower in Buda, even spread rumours of his own death and pretended to be someone else in order to avoid shipment to Constantinople.25 The prison warders were fully aware of this and would often mention to a prisoner the possibility of his transport to Istanbul in order to increase the amount of ransom offered.26 Things were little different in prisons located on Christian territory, where if a gaoler was dissatisfied with the amount of ransom offered by a prisoner, he would threaten to send him to Vienna.27 Once he was in Vienna, a prisoner had little chance of ever being able to return to the Ottoman Empire, for Turkish prisoners taken to the city usually ended up as servants to the court of one of the imperial dignitaries or on the banks of an Italian commercial galley. Like the Turkish “Emperor’s captives”, high-ranking Ottoman officers were taken to the prison of the court of the Hungarian kings. From the 1560s, border fortress captain-generals paid by the ruler found themselves under the following strict requirement: “By God’s permission, if the warriors should get some prey during battles with the enemy … the captain-general is bound to divide up the booty among the warriors evenly, satisfying himself with the gift he is due, and keeping for us the pashas, beys, and ağas, if such can be caught.”28 It 24  Sándor Takáts, Komáromi daliák a XVI. században [Valiant Soldiers of Komárom in the Sixteenth Century]. Budapest, 1909, 42–43. 25  In his letters to Ádám Batthyány, however, he calls himself by his proper name. This all tends to suggest that the letters of the captives were not always “censured” by the Ottoman gaolers. MOL Batthyány cs. lvt. P 1314, Missiles Nos. 20.020, 20.021. Cf. also: A Turkish prisoner in Németújvár in 1646 contrived in the same way, as “he calls himself sometimes İbrahim and sometimes Ömer in his vacillating speech.” Ibid., No. 41.244. 26  Ibid., No. 23.882. Kanizsa (1634) At the fortress of Eger in 1658, after escape of one of the prisoners on release, several of his fellow prisoners were taken to the Porte, so that higher ransoms and sureties could be obtained from the others. Izsépy, “Az egri törökök,” 162. 27  E.g. MOL Batthyány cs. lvt. P 1314, Missiles No. 41.219. 28  This customs was also known in Transylvania. See Sándor Takáts, “A hadi kótyavetye a török világban [War-Auction in the Turkish World],” in Idem, Rajzok a török világból [Sketches from the Turkish World]. III. Budapest, 1917, 147. The earliest evidence in Latin known by me comes from the instruction of Simon Forgách, border fortress captaingeneral of the mining district, issued on October 26, 1569: Proinde supremus noster capitaneus milites suis manubiis (Passis tamen, Zansakis et Beghis Turcicis exceptis, quos nobis reservamus) pro militari in Hungaria diu iam observata consuetudine frui sinat et

Ransom Slavery Along the Ottoman–Hungarian Frontier

925

was under these provisions that in 1583 Ali bey of Koppány, captured by the soldiers of Veszprém and Palota, was transferred to Emperor and King Rudolf II (1576–1608), who donated him immediately to his devotee, Obersthofmeister Leonhard IV. von Harrach.29 Both the Hungarian and Ottoman border fortress soldiers were unhappy that their respective emperors demanded the best captives for themselves (as some kind of de facto prisoner tax). Indeed, there were cases—for instance in 1587 after both the Koppány and the Kacorlak actions— in which the soldiers refused to comply with the ruler’s orders and lodged complaints in connection with the more valuable captives’ being sent up to Vienna.30 However, their requests were never met, and thus they were left with the option of concealing the capture of high-ranking Ottoman officers— a tactic that was not necessarily more successful, given that the Ottoman leadership was inclined to approach the Court in Vienna or the Aulic War Council when seeking freedom for distinguished pashas or beys. Consequently, for soldiers on both sides, the real prize became the so-called “ordinary captives”. The Christians considered as such the lesser noblemen possessing only limited wealth, while on both sides ordinary captives included the lower rank officers of the border fortresses (vice-captains, secondlieutenants, voivodes, corporals, or alaybeyis, odabaşıs, and ağas) and the common soldiers. Such men were taken to the two imperial capitals only in exceptional cases. Usually they were imprisoned in the neighbouring vilayet or sancak centres, or in the larger Hungarian fortresses, whence after many years of captivity they would return to their native lands, thus enriching their owners considerably. As the above table concerning the number of Christian prisoners in the Stump Tower shows, they were far more numerous than the major captives. By the seventeenth century the ransom slave trade had developed into a major area of business. The third large group of prisoners in the Hungarian theatre of war (on both sides) comprised peasants and ordinary soldiers of peasant origin serving in the border fortresses. Great numbers of such men were taken captive by martaloses and Tatars in the first half of the sixteenth century and by soldiers of the imperial forces during the war of re-conquest. The aim of their capture humaniter eos tractet, contentus nimirum, si ei honestum munus de qualibet praeda ab hostibus reportata impartietur. ÖStA HKA Hoffinanz Ungarn rote Nr. 21. 1570. Aug. fol. 106. 29  See Ferenc Szakály’s study in the present volume. 30  The ruler denies the handing over of Receb bey of Koppány: Ibid. and Sándor Takáts, “Berenhidai Huszár Péter [Péter Huszár from Berenhida],” in Idem, Régi magyar kapitányok és generálisok [Old Hungarian Captains and Generals]. Budapest, [19282], 314–317.

926

Pálffy

was not ransom income but their sale at the slave markets of Constantinople or to German and Italian merchants. Once they had been taken to the seraglios, the courts of the nobles, or the galleys, their only hope for freedom lay in divine intervention. Nevertheless there were some who managed to return to their native lands. Such men included the Christian galley slaves who made their escape during the battle of Lepanto in 1571,31 or learned men (e.g. Georgius de Hungaria, Bartholomaeus Georgievits or György Huszti from Raszinya) who, having won their freedom, enriched contemporary Europe’s knowledge of the east with fantastic “accounts of their adventures”.32 People captured during the raids were divided up by the soldiers of the border fortresses in various ways. In some cases—as mentioned in connection with the warriors of Veszprém—the division was made on the basis of prior (or sometimes subsequent) agreements drawn up with the Hungarian lords of the plundered areas or with the border captains.33 If the raiders came from various border fortresses and the booty or number of captives was substantial, then, similarly to a modern auction, the military booty, goods, livestock, and captives (the latter being regarded as material objects) were sold off and the proceeds divided among the soldiers proportionately to their participation. This process was designated in Hungarian by the phrase kótyavetye of South Slavic origin (Latin: auctio, German: Beutverkaufung or Austeilung).34 Registration and assessment of the booty and the captives, their sale, and the subsequent division of proceeds (known as kótyavetyepénz or Beutgeld) were all performed by commissioners elected by the soldiers (known in Hungarian as kótyavetyések, in German as Beutmeister, and in Latin as licitator or persona auctionem curans), who were assisted by a number of scribes. In a manner similar to the oath of a judge ( juramentum), the commissioners had sworn to 31  Ferenc Szakály, “L’espansione turca in Europa centrale dagli inizi alla fine del secolo XVI,” in I Turchi il Mediterraneo e l’Europa. Ed. by Giovanna Motta. Milano, 1998, 133–151. 32   Rabok, követek, kalmárok, passim. 33  In 1647 the foot soldiers of Kiskomárom “began bargaining with my lord captain”, during the course of which they managed to persuade him to be satisfied with 60 florins from the prey (instead of 100 florins). Béla Iványi, Végvári élet a Dunántúlon a XVII. században [Life in the Transdanubian Marches in the Seventeenth Century]. Budapest, 1958, 47; cf. MTA Kt. Ms. 5301/7. 34  More rarely the booty would be divided among the border soldiers according to the “number of horses or swords”, i.e. proportionately but without an auction. János J. Varga, Szervitorok katonai szolgálata a XVI–XVII. századi dunántúli nagybirtokon [The Military Service of the servitors on the Large Transdanubian Estates]. (Értekezések a történeti tudományok köréből, Új sorozat 94.) Budapest, 1981, 109: note 85 and Takáts, “Magyar rabok, magyar bilincsek,” 533.

Ransom Slavery Along the Ottoman–Hungarian Frontier

927

carry out their task in an impartial manner.35 The soldiers of the border fortresses also gave their word—as they did after the raid on Igal in 1641:36 “God help me, Father, Son and Holy Spirit, the Holy Trinity, one certain God. Given that I left my home beside Lord Ádám Batthyány against the Turkish enemy, if I found something, gained or drew it, be that from a Christian, Turk or Serb, I would give it up. If I got information of some profit in the hands of someone else, I would tell it.”37 At the auctions, in accordance with the ruler’s provisions cited above, the border fortress troops granted each border captain one valuable captive; indeed, after particularly successful actions, prisoners were also granted to the border fortress captain-generals and the ruler himself. The size of the “honest gift” (munus condecens or honestum)38 to be made—which was at least as large as that due to a common soldier39—was then determined by the border fortress captains. All this could be traced back to a custom that had become established among the troops of the big landowners by the middle of the sixteenth century. According to the custom, the head and keeper of the soldiers, that is the lord, was entitled to receive a third of all captives.40 This amount could not be withheld by the fighting troops. However, this element of customary law developed in a different manner in the royal border fortresses. The ruler and the border fortress captains realised in the middle of the sixteenth century the value of incorporating the rule of the third into the system of customs under development in the royal border fortresses. During the period of Ottoman occupation, they made various attempts to introduce such a “tax”—which would have been similar to the one-fifth (pencik) levied by the Ottomans. However, their efforts were vigorously and repeatedly opposed by the fortress warriors, who regarded exemption from seigneurial or state taxes 35  MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 51. fol. 10. For the contemporary judges’ oath, see Ferenc Kovács, A magyar jogi terminológia kialakulása [Formation of the Hungarian Legal Terminology]. (Nyelvészeti tanulmányok, 6.) Budapest, 1964, 95–120. For the duties of the auctioneers, see also Ábel Ődöngő [Sándor Takáts], “Nyelvtörténeti adatok. Kótyavetyés [Data on the History of the Hungarian Language: The Auctioneer],” Magyar Nyelv 6 (1910) 132. 36  For a detailed account of the lessons of the Igal raid, see Fenyvesi, “Az igali portya”. 37  MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 51. fol. 10. 38   The first expression is used in the instructions of Imre Bornemissza, captain of Krasznahorka, issued in 1573: MOL MKA Diversae instructiones, E 136, Tétel 7. fol. 540, the second in the cited instructions of 1569 by Simon Forgách, captain-general of the mining district: ÖStA HKA Hoffinanz Ungarn rote Nr. 21. 1570. Aug. fol. 106. 39  Iványi, op. cit., 49. Cf. Varga, Szervitorok, 108–109. 40  Takáts, “A hadi kótyavetye,” 146. Cf. also from 1566: Kárffy, op. cit., 473–474: Nos. II–III.

928

Pálffy

as an important aspect of their collective rights and privileges (which also included an independent administration of justice).41 Seeking to avoid a drawnout conflict, the War Council finally refrained from introducing the tax in the second half of the sixteenth century, for it was unable to assure the proper remuneration of the soldiers. Subsequently, the military command in Vienna had to suffice with the receipt of the major captives (in accordance with the captains’ instructions), while the commanders of the border fortresses received only gifts. On the other hand, among both the private soldiers of the seigneurs and the inhabitants of the so-called haiduck settlements, the giving of a third was still a custom even in the second century of Ottoman era. Even so, among the haiducks, the giving of a third was required only where military service was performed for money rather than for privileges.42 Thus, throughout the sixteenth and seventeenth centuries, the division of military booty and of captives often gave rise to disputes. The significance of the issue was well demonstrated at a military court (sedes bellica) convened in Körmend in July 1647 by Ádám Batthyány, captain-general of the border fortress district facing Kanizsa. The court debated the problem and issued various directives in connection with it.43 Referring to “established customs”, the council confirmed that the soldiers of the royal border fortresses—in accordance with the size of their spoil—owed simply “reasonable gifts” (munus concedens) to their captains and were not required to grant them a third. This statement was needed because some fortress commanders in an abuse of their power, had even resorted to the use of force in order to obtain the third from salaried soldiers of the ruler. At the same time, in connection with the private soldiers, the judges of the military court reaffirmed the earlier custom and upheld the requirement to give a third. In addition, they determined that in the course of an auction, military booty was to be divided in proportion to the numbers of guards of the various border fortresses, and that those obtaining booty after “hearing word from others” were to receive only half the amount per man. They also declared that soldiers who brought back a Turkish head or a captive from the raid were to receive an additional sum of three florins over and above

41  Pálffy, Katonai igazságszolgáltatás, 71–72. 42  Hernádnémeti, 1630: Imre Dankó, A Sajó–Hernád-melléki hajdútelepek [The Haiduck Settlements Along the Rivers Sajó and Hernád]. Sátoraljaújhely, 1991, 22 and István Szendrey, Hajdú-szabadságlevelek [Letters of Privileges for the Haiducks]. Debrecen, 1971, 233: No. 6; Fehértó and Bekény, 1632: Ibid., 239–240: No. 9; Lúc, 1642: Ibid., 252: No. 18. 43  MOL Batthyány cs. lvt. P 1313, Katonai iratok 252. cs. fols. 622–625.

Ransom Slavery Along the Ottoman–Hungarian Frontier

929

the amount due to them from the proceeds of the auction.44 Those, however, who did not heed the calls of their fellows and refrained from participating in the raids, or who purposely delayed their departures, were deserving of punishment rather than of gain. An auction was always to be held in the border fortress whose soldiers who had participated in greatest numbers in the incursion, but only after the fortress captain or the border fortress captain-general had been informed.45 Subsequently, the border fortress captain-generals and the commanders of the fortresses attempted to use this latter decree, which was based on the old customs, to persuade the royal soldiers of castles under their command to recognise their (non-existent) right of pre-emption in connection with the distinguished captives. This practice can likewise be traced back to the customs of the soldiers of the private landowners. Such soldiers were only allowed to sell their captives once the dominus had given his permission and made his purchases. An excellent formulation of this demand can be found in a statute of the landlords of Keszthely issued in 1661 where it is stated that “people and warriors living in Keszthely, whoever’s servants they may be, who win something with weapons: captive, cattle, and good weapon, should sell it to the lords at a good price, so that anyone of them may buy it.”46 The three lords even attempted to introduce the right of pre-emption (binding only the private soldiers earlier) among soldiers of the border fortresses in the pay of the ruler. Thus, when in 1652 the army of Kiskomárom (near Lake Balaton) protested against this practice they did so in vain, for Ádám Batthyány informed them in no uncertain terms that “if they bring a good captive, it is our due alone.”47 As regards the purchase and sale of captives, another noteworthy legal regulation became established by the end of the seventeenth century: the vendor was required to give a certain “warranty” for the captive being sold. If, for instance, the captive died within fifteen days of sale, the vendor was required to pay back the full price to the purchaser. This custom came into being because many of the captives that were taken prisoner were injured or fell ill while they were in prison. Many owners sought to sell them on as quickly as possible. 44  This decree followed very old traditions that continued to survive after the ending of the border fortress systems at the time of the independence movement of Ferenc Rákóczi II: “A soldier who brings back a captive from among the enemy shall receive a clear three florins.” Takáts, “A hadi kótyavetye,” 179: note 1. 45  For this practice, see also Varga, op. cit., 110. 46  Sándor Kőszeghy, “Keszthelyi rendtartás 1661–ből [The Statute of Keszthely from 1661],” Magyar Gazdaságtörténelmi Szemle 2 (1895) 64: article 6. 47  Varga, “Rabtartás és rabkereskedelem,” 124.

930

Pálffy

Despite the practice of a two-week term of warranty, disputes did arise, especially where there was no written record of the sale. In 1654, for instance, Ferenc Csáky, captain-general of Veszprém, and the soldiers of the nearby fortress of Tihany became involved in such a dispute. The quarrel began when the captain-general bought an injured Turkish captive from the soldiers for 500 florins and the man died just fifteen days later. Csáky attempted to recoup his money by (illegally) arresting the soldiers of Tihany (who were not under his jurisdiction) and demanding his money back in his own judicial forum, the military court (sedes bellica) of Veszprém. At the court, the border fortress soldiers referred to the “customs of the border fortress soldiers”, but they did so in vain for the captain-general denied that there had been any word of a twoweek expiry date in their original verbal agreement. Acting under pressure, the military court judged in favour of Csáky. Subsequently, the soldiers of Tihany lodged an appeal at the military court of the captain-general of the border fortress district of Győr. At this court the ruling was made in favour of the soldiers of Tihany; the judges invalidated the Veszprém court’s decision, and even reprimanded Csáky for seriously abusing his authority as captain.48

Opportunities for Freedom: Escape, Conversion, Payment of Ransom

Captives sold at the auctions were usually bought by the generals of the border fortress districts or by affluent fortress captains and officers. Common soldiers on both Ottoman and Christian sides kept no prisoners because they could not afford to pay the gaolers (Latin: custos carceris, German: Tömlizer, Kerkermeister) or the castellans (Hungarian: porkoláb, Latin: castellanus, German: Burggraf ) for their upkeep. The fee was called the “prison ransom” (tömlöcváltság, Tömnizgeld); it was to be paid by the owner of a captive on or before his release from prison.49 Problems concerning payment of this 48  MOL Csáky cs. lvt. P 71, Fasc. 273. Köteg I. fols. 27–31. For a detailed account of the affair, see Varga, Szervitorok, 158–159 and Pálffy, Katonai igazságszolgáltatás, 122–123. 49  MOL MKA Acta diversarum familiarum, E 200, Tétel 22: Szigeti Horváth cs. lvt. [Archives of the Horváth family from Sziget], Letter of István Henyes and Boldizsár Ányos to Márk Horváth. March 16, 1551, Sümeg; cf. Ferenc Szakály, “Egy végvári kapitány hétköznapjai. (Horváth Márk szigetvári kapitány levelezése Nádasdy Tamás nádorral és szervitoraival, 1556–1561) [Everyday Life of a Fortress Commander (Correspondence of Márk Horváth, Commander of Szigetvár with Palatine Tamás Nádasdy and his Servitors, 1556–1561],” Somogy Megye Múltjából (Levéltári Évkönyv) 18 (1987) 57; further data: MOL Batthyány cs. lvt. P 1314, Missiles 34.785, 41.253.

Ransom Slavery Along the Ottoman–Hungarian Frontier

931

sum sometimes arose when a Turkish slave was exchanged for a Christian one. (In one such case, Ádám Batthyány forgot to pay off his prison warders of Németújvár.)50 Still, gaolers on both the Turkish and Christian sides were unlikely to destitution, given that by the middle of the seventeenth century— owing to their not always “peaceful” practices—it had become customary on the borders for prisoners to pay a “gratuity” to them on release. Thus, in 1660, as both sides were attempting to regulate (unsuccessfully as it turned out) some customs concerning the keeping of slaves, it was solemnly pronounced that “castellans and gaolers, both Turks and Hungarians, should not take gifts from the poor captives, but should be satisfied with their due [namely the prison ransom].”51 Captives that passed into the ownership of a senior officer or captain by way of purchase, exchange, or as a gift could win their freedom in a number of ways. Rarely, this occurred through a “miraculous act of God and without paying a ransom”—as in the case of Pál Farkas, a man who escaped from Szigetvár after he had murdered his gaoler.52 Under the terms of the Treaty of Adrianople (1568) and according to customary law, neither party could demand back such fugitives. Nevertheless, those who were caught while attempting their escape were punished severely. For example, in the late autumn of 1652, thirty-eight Turkish captives from Szalónak in West Hungary revolted, but were brought back to their cells by armed guards. The vengeance of the gaolers was merciless. Apart from the injured, all those who had attempted to escape were thrashed (300, 350 or 380 strikes), which proved fatal for many.53 News of the rebellion, which caused a considerable loss to Ádám Batthyány, spread quickly. In January 1653 the captives of the Stump Tower in Buda learnt of it, too. (Indeed they also suffered its consequences for the Ottomans introduced stricter rules concerning the return of the captives who had been released from the Batthyány fortresses on bail.)54 Since a primary element of the Ottoman–Hungarian customary law was the principle of reciprocity— according to which if either party contravened the traditions (e.g. forbade 50  Ibid., No. 41.234 (1650). 51  Ibid., Nos. 9299 and 7761. 52  Kálmán Szily, “Farkas Pál és Farkas Ádám följegyzései 1638–tól 1694–ig [The Records of Pál Farkas and Ádám Farkas from 1638 to 1694],” Történelmi Tár (1884) 88–89. 53  Janissary Hüseyn of Buda “was beaten on December 4 and 5, 1652 because he had participated in the capture of our fortress Szalónak and he had escaped; on December 8, 1652 he died.” MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 49. p. 72. 54  MOL Batthyány cs. lvt. P 1314, Missiles No. 9286. A year earlier they had received word of the battle of Vezekény, and mourned their fallen fellows in their cells. Ibid., No. 9284.

932

Pálffy

begging or tortured the prisoners excessively) the other party was entitled to treat its own prisoners in the same way55—the liberation of the Christian captives of Buda was hindered for quite a long time. Even prisoners who failed to escape could win their freedom without paying ransoms in a number of other ways. For instance, they could accept the faith of their warders, that is they could become Christians or Muslims.56 (However, such conversion was more often the result of “persuasion”.)57 In such cases, prisoners became subjects of the Ottoman sultan or of the Hungarian king, and they also became the servants of their owners. At the same time, under the 1609 decrees of Egerszeg no one was to be persuaded “through the making of fair promises or by force”.58 A captive could also be freed in the course of an exchange of prisoners of equal value. Such actions took place on the basis of mutual agreements between the two states or the frontier authorities, as well as during important legations. The agreements provided for the exchange of prisoners of equal ransom value and they also covered the time and place of 55  “Should the bailsmen in Hungary receive a single trash, he [the bey of Szolnok] swore on his word of honour to deal the Hungarian captives 500 strikes each in return for every single trash” (1656); Valentinus Bujdosó [Kálmán Thaly], “Szólnoki rabság levele, Koháry Istvánhoz [The Letter of Captives in Szolnok to István Koháry],” Századok 5 (1871) 215; “When we were severely beaten or we were poorly supplied with provisions, we wrote a letter to Ónod, and the Hungarians started to treat the Turkish captives in the same way”; Magyar Simplicissimus [Hungarian Simplicissimus]. (Aurora, 4.) Ed. by József TurócziTrostler. Budapest, 1956, 191. Cf. also the letter of the prisoners in the Stump Tower to the gaolers of Németújvár: MOL Batthyány cs. lvt. P 1314, Missiles No. 9307 (undated). 56  In 1596 Şaban alias Benedek and Ahmed alias Jancsi, after almost ten years’ imprisonment converted to the Christian faith, which—as they stated—“we wish to keep until our deaths and to die in the Christian faith”. At the same time, in a letter to the Hungarian Chamber, they begged for at least a small donation in their support, whereupon one florin was paid out to them. Gábor Mátray, “Magyar rabok megváltási módja török fogságból a XVI. században [The Practice of Redemption of Hungarian Captives from Ottoman Captivity in the Sixteenth Century],” Divatcsarnok 2:9 (15.02.1854) 210: No. IV. After 1686 some of the Turks who were still in Buda—like their compatriots in Eger—avoided eternal slavery in the same manner. In the 1690s, many of the (mainly female) Ottoman prisoners that had been purchased by the citizens of Sopron, converted to Christianity. Lajos Némethy, “Az 1686. évi visszafoglalás után Budán maradt törökök [The Turks Remaining in Buda after the Reoccupation in 1686],” Századok 11 (1877) 141–148 and Kovács, “Török rabszolgák Sopronban,” 8. Cf. the excellent data concerning the baptism of Turkish children in the (still intact) Fiume registry of births: Klen, “Pokrštavanje »turske” djece”. 57  We know of one such experience of a Greek merchant from the 1660s: Dongó, “Magyar rabok,” (1915), 263–266: No. XLIX. 58  MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 9.

Ransom Slavery Along the Ottoman–Hungarian Frontier

933

the exchange. Cross-border exchanges were most likely to occur on the signing of a peace treaty or following the visit of an important envoy.59 Finally, a Christian captive could also gain his freedom by agreeing to obtain the “withdrawal” of a Turkish prisoner in return for his ransom—“a head for a head”.60 This, however, brings us to the most essential element of the keeping of captives, namely the determining and obtaining of the ransom. The ransom (German: Schätzung, Ranzion, Hungarian: sarc, Turkish: baha, fidye) was offered by the captives themselves to their masters. If an owner was not satisfied with the amount—as was often the case—a process of bargaining would begin. The owner would attempt to push the price of ransom as high as possible, and would state his demands.61 Before the ransom amount was fixed, the owner would try to ascertain, with the help of his spies, just how large a sum the relatives of the prisoner were able to pay.62 If a captive was unwilling to pledge the amount demanded, various appliances could be employed to persuade him to change his mind, including the strappado (Latin: trochlea, funis oblungus vulgo strappado, Hungarian: csiga), the stocks (Latin: cyppus, catasta, German: Geige, Fiedel, Hungarian: kaloda)63 or the so-called “young 59  In late 1628 the Aulic War Council, after receiving word that the pasha of Buda was willing to exchange 6 Christian prisoners, ordered István Pálffy, captain-general of the mining district, to ensure—in return for a ransom to be paid by the Hungarian Chamber—the release of six Ottoman prisoners of similar ransom value. Although Pálffy fulfilled the Council’s request, the Chamber was slow to transfer the amount promised. Pál Jedlicska, Eredeti részletek gróf Pálffy-család okmánytárához 1401–1653 s gróf Pálffyak életrajzi vázlatai [Supplementary Details to the Historical Documents on the Count Family Pálffy 1401– 1653 and Biographical Sketches of the Count Pálffys]. Budapest, 1910, 62–63: Nos. 111, 113, 71: No. 127. In 1570, on the nomination of Karl Rhym as the Habsburg ambassador to the Porte, two Ottoman prisoners in Komárom were exchanged for two gun-boatmen held captive in the Stump Tower. Takáts, Komáromi daliák, 42. Cf. also 1666: [Paulus Tafferner], Caesarea legatio, quam Mandante Augustissimo Rom[anorum] Imperatore Leopoldo I. Ad Portam Ottoma[n]nicam suscepit, perfecitq[ue] excellentissimvs dominvs, dominvs Walterus S. R. I. Comes de Leslie. Viennae, 1668, 171–172 and ÖStA KA Alte Feldakten 1665/13/2, as well as Ibid., 1666/12/1, passim; 1668–1674: Auer János naplója, 20–36 and the many Ottoman data from the seventeenth century concerning requests for the release of captives at the time of the peace treaties: Fekete, Türkische Schriften, passim. 60  In 1650, İbrahim of Fehérvár “who wished to liberate a head for a head, undertook the lowering of Máté Beszprémi Balog’s ransom, and besides he promised us 1,200 cubes of rock-salt”: MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 49. p. 16. 61  Ibid., Nos. 107, 110, 131, 144, 152, 154 and 188. 62  Varga, “Rabtartás és rabkereskedelem,” 126–127. 63  For a description of the strappado, see: Kivégzés, tortúra és megszégyenítés a régi Magyarországon (Kiállítási katalógus) [Execution, Torture, and Shaming in Old Hungary

934

Pálffy

wife”.64 Of course, for their part, the captives would try to reduce the amount to be paid. They would often request their pashas or captain-generals to intervene on their behalf, who would then write something similar to the following: “although he was the sancakbeyi of His Majesty the Emperor, but he received the office not for gold or treasure, but for his service to the faith. Believe me, he was so poor that (when he was taken prisoner) the mighty Emperor’s treasurer seized all his fortune for the Emperor and sold it.”65 Those captives who refused to put up the ransom fee despite torture remained “unransomed” and were sentenced to perpetual captivity. Only the above mentioned pribeks could expect a fate worse than this. According to the customary law of the border fortresses, they could not buy their freedom for money: for their kidnapping activity they were sentenced to death preceded by torture. A contract was issued to both the captive and his keeper concerning the agreed ransom fee.66 Such a decree would establish the composition of the ransom and the method and timing of payment. Ransoms usually comprised amounts of cash as well as horses, oxen, and a wide range of other goods (including banned weapons, carpets, coffee, etc).67 In 1660 the Ottoman side urged the introduction of a rule whereby weapons and other goods would not be requested for the captives, but that they would be exchanged for cash or for other prisoners in the following way: “a lord for a lord, a leader for a leader, a sipahi for a nobleman, a beşli for a cavalryman, a haiduck for a Turkish footsoldier.”68 But none of this was ever realised. Indeed, in western Transdanubia, in addition to many other goods, rock-salt was demanded from the Turkish captives. It seems from the amount of salt acquired in this way (many thousands (Exhibition-Catalogue)]. Ed. by Attila Pandula and Péter Havassy. Eger, 1989, 19. For the stocks, see: Sándor Takáts, “Adatok nyelvünk történetéhez. Kaloda [Data on the History of the Hungarian Language: Stock],” Magyar Nyelv 2 (1906) 271. Takáts, “Magyar rabok, magyar bilincsek,” 535–540. Károly Vajna, Hazai régi büntetések [Hungarian Old Punishments]. II. Budapest, 1907, 99–112 and Ferenc Temesváry, Büntető eszközök a régi Magyarországon [Criminal Vehicles in Old Hungary]. (A Jánosházai Múzeum Közleményei, 1.) Szombathely, 1970, 29–32. 64  The “young wife” in the Stump Tower in Buda: MOL Batthyány cs. lvt. P 1314, Missiles No. 9307 (undated). Cf. Dongó, “Magyar rabok,” (1914), 62: No. XXVIII. 65  The letter of Ferhad, pasha of Buda, to Miklós Pálffy, dated June 6, 1589: Jedlicska, “XVI. századi török–magyar levelek,” 694–696: No. IX; cf. also Ibid., 691–692: No. III. 66  Ágoston Szalay, Négyszáz magyar levél a XVI. századból. 1504–1560 [400 Hungarian Letters from the Sixteenth Century. 1504–1560]. (Magyar Leveles Tár, 1.) Pest, 1861, 282: No. CCXCIV (1558). 67  MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 43. Cf. Szily, op. cit., 88–89. 68  MOL Batthyány cs. lvt. P 1314, Missiles No. 9299.

Ransom Slavery Along the Ottoman–Hungarian Frontier

935

of cubes annually)69 that the Batthyánys were able to fulfil the needs of their estates in this western area of the country where Transylvanian salt was so hard to obtain. Meanwhile the Turks bargained with their Transylvanian captives and obtained many hundreds or even thousands of salt cubes.70

Difficulties Concerning the Collection of the Ransom

Once both Hungarians and Ottomans had recognised that it was the captives themselves who were most capable of collecting their ransoms (or at least the outstanding amounts), they elaborated various schemes for this to happen, while ensuring their return through a system of suretyship. Generally, a captive would only be released after his relatives had sent a certain part of the ransom to the owner. In Buda “it was the custom that if half the ransom had been paid, the captive would be released.”71 The smallest guarantee that could be made on the release of a captive was a “charter” (German: Glaubensbrief, Hungarian: hitlevél) in which a Hungarian nobleman or border fortress captain, or indeed a high-ranking Ottoman official, vouched for the return of the captive and pledged payment of a certain sum or of the ransom fee itself. On receipt of the charter, the keeper of the captive fixed the date (terminus) by which the outstanding amount of the ransom fee had to be paid. In a few cases the term indicated would be just one or two weeks, but more often it would extend over several months, and sometimes (although rarely) over a period of years.72 If payment was not made by the given date, the captive was to be returned to his prison cell. However, since often the full amount of the ransom could not be collected in time, it became general practice for captives to request extensions to their terms.73 Although such extensions depended on the good will of the keepers, they were usually granted. Thus, captives generally paid off their often considerable ransoms in 69  A special record was made of the use of rock-salt obtained from the captives: MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok Nos. 141, 303–304, 306, etc (1646–1652). 70  Dongó, “Magyar rabok,” (1915), 146–147: No. XLVI. 71  MOL Batthyány cs. lvt. P 1314, Missiles No. 9282. Cf. also Ferenc Szakály’s study in the present volume. 72  In 1683, lady Zsuzsanna and her daughter Katus, who had been taken from Kislehotka (Nyitra county) and imprisoned in Buda, were given 8 years by Ahmed pasha of Buda in which to collect their ransom of 220 Hungarian florins. Pál Horánszky, “Ahmed az utolsó budai basa magyar levele [Hungarian Letter of Ahmed, the Last Pasha of Buda]”, Magyar Családtörténeti Szemle 3 (1937) 191. 73  MOL Batthyány cs. lvt. P 1314, Missiles No. 9294 (1657).

936

Pálffy

small instalments. Releases made on the basis of such charters were not universally popular. Indeed, they were even banned after representatives of the Christian side attempted to mislead the Ottoman gaolers with fake charters.74 But the beys of the border fortresses also often broke their pledges and would even threaten the royal border fortress soldiers with the capture of their major prisoners. Acting out of revenge, in 1611 the Hungarian chronicle writer Gergely Pethő threatened Mustafa ağa of Szigetvár in the following manner: “I tore up your letter of safe conduct on a pig-tail in front of many Christian and Muslim warriors, let these noble people see your humanity.”75 For the money-hungry owners of captives, a more satisfactory solution was offered by the “system of pledge”, as it was called. This meant that wealthy relatives of a captive would generally send a number of servants to be held in prison for the duration of the temporary release of the captive. There were also exemplary cases of self-sacrifice in which family members or fellow soldiers were ready to accept the miserable prison life—sometimes for months—while the captive tried to collect his ransom. This practice, however, also proved unpopular, for it could be abused. Very often the person or persons held in lieu of the captive would not be set free on payment of the ransom. Instead, they would be kept as a security for another captive who was on release. And although—as Ahmed, the alaybeyi of Kanizsa expressed so clearly in 1641—“there was no law in either Turkey or Hungary allowing for someone to be held as a guarantee for the ransom of another person,”76 the captains and beys often sought to insure themselves by employing this illegal method. The most easiest, safest, most effective and consequently most popular means of obtaining the release of captives was the institution of surety or bail (German: Bürgschaft, Latin: sponsio, Hungarian: kezesség, Turkish: kefalet). The system in question was widely employed in the jurisdiction of the border fortresses77 and thus rapidly became one of the most interesting elements of Ottoman–Hungarian customary law in the frontier zone. However, while Christian soldiers acting as bailsmen for fellows facing military court cases would deposit sums of money or goods, in the case of the captives the role of bailsman was very different. The guarantors (German: Bürge, Latin: sponsor, fideiussor, obses, Hungarian: kezes, Turkish: kefil) themselves were a very di-

74  Ibid., No. 15.615. 75  MOL Festetics család keszthelyi levéltára [Archives of the Festetics family in Keszthely], Gersei Pethő cs. lvt. [Archives of the Pethő family] P 235, Antiqua miscellanea No. 950. 76  MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 75/d. 77  Pálffy, op. cit., 154–155.

Ransom Slavery Along the Ottoman–Hungarian Frontier

937

verse group. According to the regulations, nobody could be forced into becoming a bailsman. However, in reality, the practice was just the opposite.78 Very often—as has been demonstrated in the examples above—fellow prisoners of a released captive would be forced into becoming his guarantors. The number of bailsmen ranged from several to as many as fifty. And in some cases the whole prisoner population of the Stump Tower in Buda was forced into standing surety.79 In letters of bail (German: Bürgschaftsbriefe, Latin: litterae fideiussionales, Hungarian: kezeslevél, Turkish: kefaletname), which contained details of the ransoms due from the released prisoners and indicated the timing of payment, bailsmen undertook to pay the agreed ransom fees if the released prisoners failed to return. In other instances, under the terms of bail, guarantors could expect to pay further sums of money or double the original ransom fees, or even to lose parts of their bodies (eyes, noses, ears, fingers or teeth) and be subjected to thrashings.80 Sometimes gaolers had no qualms about carrying out these draconian punishments and amputations.81 The prisoners often attempted to forestall such punishments by offering up further amounts of money or other parts of the body in order to extend the deadline by which the released captive was to return. However, in 1634, further bail terms did not help the prisoners of Kanizsa, for the released prisoner “begging for his ransom in the towns and villages, spent the money on drink.”82 Where a prisoner acting as bailsman denied or failed to recognise his responsibilities, he and the released captive were brought together in one of the border fortresses and forced to give evidence before a group of compatriots.83 The best bailsmen were not other destitute prisoners but the wealthier market towns and settlements of the occupied territories or border zone. The 78  See the 1647 letter of the captives of Buda: MOL Batthyány cs. lvt. P 1314, Missiles No. 7761 and MOL Zichy cs. lvt. P 707, Missiles No. 2102. Published in Monumenta Hungariae Judaica. XII. 1414–1748. Ed. by Sándor Schreiber. Budapest, 1969, 75–76: No. 11/3 (1577). 79  MOL Batthyány cs. lvt. P 1314, Missiles No. 9295 (1657). 80  For individual examples of the above cases, see: MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 4. fol. 94, Nos. 105, 139 and 148. 81  MOL Batthyány cs. lvt. P 1314, Missiles No. 9307. Cf. also with the vivid description offered by Hungarian Simplicissimus of the forcing of the ransom payment (Magyar Simplicissimus, 190–191): “We were laid down, our feet were put in the stocks, one on the other; and they started to beat slowly but without mercy our soles, and generally demanded with ever increasing beating the ransom … they cut a cross into my soles, pressed out the blood and put diluted horse-dung on them to make it drain the pus.” 82  MOL Batthyány cs. lvt. P 1314, Missiles No. 23.882. 83  Béla Vay, “Menheth bég levele Ibrányi Lászlóhoz [The Letter of Mehmed Bey to László Ibrányi]”, Magyar Családtörténeti Szemle 2:5 (1936) 14.

938

Pálffy

Ottomans were quick to make use of the opportunities at hand. For instance, in 1584, ignoring the consequences for commercial trade, the Ottomans forced six oppida (Ráckeve, Nagymaros, Vác, Nagykőrös, Cegléd, and Tolna) in the occupied areas to be guarantors of the ransom fee of Ali, bey of Koppány which amounted to more than 30,000 golden florins. A quarter of a century earlier, three border settlements (Szikszó, Debrecen, and Rimaszombat) undertook a guarantee for the payment of Gáspár Mágochy’s ransom of 14,000 florins.84 In the Dalmatian territories we know of cases from as early as the fifteenth century in which the citizens of Ragusa stood bail.85 These customs remained in place in the Kanizsa border zone area even as late as the seventeenth century. At the request of their captives, the Ottomans would often call judges from villages in the vicinity of Kanizsa, in the counties Zala and Vas, and require them (on behalf of their villages) to undertake payment of bail for Hungarian captives.86 Indeed, we know of one case in which a person who turned to the Ottomans in another matter was forcibly retained as a bailsman, even though he was unrelated to the captive.87 Captives sent out to collect their ransoms initially had to wear heavy iron chains around each foot. Released captives were therefore called “iron captives” (in Latin: captivus ferreus, in Hungarian: vasas rab).88 Later, after they had paid off the larger part of their ransom, they wore a chain on just one of their legs. The chains were not worn by more important slaves whose former superiors or lords paid off the so-called chain ransoms in advance. However, at the request of the captives, even wealthier relatives often refrained from paying this sum. There were two main reasons for this. First, people were more likely to be merciful to those who begged in chains and were less likely to think that the beggar was a fake. Second, the iron captives had various special rights. Since their movement was impaired by the chains, on both sides of the border they were commonly transported from one village to another in a carriage provided by local inhabitants until they reached the imaginary frontier of the two states. Thus, the captives of the Stump Tower could travel westwards from 84  See Ferenc Szakály’s study in the present volume. 85  Vilfan, op. cit., 183. 86  MOL Batthyány cs. lvt. P 1314, Missiles No. 16.401 and MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok Nos. 75/d–75/e. 87  Jenő Förster, “Egy korompai fiu török rabságban [A Boy of Korompa in Ottoman Captivity],” Közlemények Szepes vármegye múltjából 10 (1918) 153–154. 88  “They took the shackles from one of my feet, leaving on the other. The other end of it was fastened to my belt, and so I had to set out with two other Christians to Ónod with a Turkish safe conduct.” Magyar Simplicissimus, 193.

Ransom Slavery Along the Ottoman–Hungarian Frontier

939

Buda towards Tata as far as Bicske.89 Local inhabitants were also expected to provide for the released captives’ board and lodging. Proof that this practice soon became an established custom is offered by the many passes issued to the captives (German: Paßbrief, Latin: passus, salvus conductus, Hungarian: menlevél, útlevél) in which village magistrates were warned that everywhere they should offer a carriage to “the ransomed captives on their journeys” or the “beggar women captives” and provide for their food and beverage.90 The iron captives were also often given so-called begging letters, which—to prevent the captives from using them later on—included the date by which the ransom had to be paid. The introduction of these begging letters became necessary following an increase in the number of false captives. Such tricksters would even place chains around their legs and take up position close to a church of a village or town. Iron captives were also protected by customary law, the unwritten rules of which forbade their being whipped or bound. The strength of this custom is demonstrated by a 1680 decree of the military court of Fülek that was issued in connection with a case brought by a Hungarian captive on release against the castellan of the fortress, an interesting case per se. The captive had been travelling on a coach along a dirt road when the castellan had attacked him without reason and beaten him. Thus he had violated the system of customs established between the Ottoman–Hungarian border fortresses, and therefore the captive was requesting that since “he had been freed from paying his ransom and bail because he had been beaten by the castellan … according to the old custom, the latter should pay the ransom.” Even though the castellan denied that he had committed the violent deed, following evidence from witnesses called to the case, the military court ruled against him. However, as the captive’s body had not turned blue despite the

89  MOL Batthyány cs. lvt. P 1314, Missiles No. 9293. 90  Varga, “Rabtartás és rabkereskedelem,” 133: note 4. MOL MKA Rákóczi cs. lvt. [Archives of the Rákóczi family] E 190, Hadügy, Mai 23, 1668, Munkács: Ferenc I. Rákóczi’s letter of safe conduct for the prisoner György Kerczeghy; MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 215 (1649), No. 351 (1643), No. 392 (1657), No. 396 (1658); Letters of Ali, ağa of the cavalryman of Párkány for Miklós Kossuth from 1651 and Ferdinand III’s “passport” (also given to him) from 1652: Kálmán Thaly, “Kossuth Lajos három vitéz őse és a család czímerlevele [Three Valiant Ancestors of Lajos Kossuth and the Coats-of-Arms of the Kossuth Family],” Turul 12 (1894) 156–157; see further the begging letter (in Hungarian on the one, and in Turkish on the other side of it) issued by Ahmed pasha of Buda for lady Zsuzsanna and her daughter, who had been taken from Kislehotka (Nyitra county) from 1683: Horánszky, “Ahmed az utolsó budai basa,” 191.

940

Pálffy

beating, the court ruled that the castellan should pay just part of the ransom— an amount equal to half the so-called blood money, or 20 florins.91 Another interesting case occurred in 1647. A Turkish captive released on bail from the dungeon of Kapuvár had gone to the house of his Hungarian owner, who had then bound him for a night. Subsequently, the captive had travelled to Buda and lodged a complaint with the pasha. In co-operation with the gaolers, the pasha proceeded to serve justice, much in the same way as the military court in Fülek had done. Thus, in the spirit of the customary law of the border zone, the pasha ruled that the status of the captive no longer applied and that the captive should be released. In addition, the bailsmen were also freed of any obligation.92 When issuing its ruling, the Ottoman court repeatedly requested the opinion of the Christian prisoners in the Stump Tower. The Hungarian prisoners stated that in their view the master of the captive did not respect the institution of bail, and that as a result the captive should be given his freedom. In connection with captives released on bail, another unwritten rule reappears time and again in contemporary Ottoman–Hungarian correspondence. It often happened that shortly before his ransom was due to be paid, an iron captive would either fall victim to robbers or die in an epidemic. In such cases the bailsmen or the party requesting the release of the captive had to prove that the captive in question really had died, and that the ransom fee was no longer payable. Thus, if a captive died before his ransom was due, his body “according to the old custom of the border zone” had to be sent to his keepers, who would then receive no further payment.93 If, on the other hand, he died after this date, his bailsmen were required to pay off his ransom. The sending of the corpse was important proof of the death of the captive for both parties and prevented unfortunate happenings like the following one in Sopron in 1699. Here, a citizen of the town was about to marry the widow of a former inhabitant who had “died” in a Turkish prison, when the latter suddenly gave word that he was still alive.94 In the middle of the seventeenth century, the unruly behaviour of an increasing number of soldiers rendered conditions along the border rather chaotic. The captives also played their part in this. Not only was there a rapid

91  MOL Esterházy cs. lvt. P 125, Pál nádor No. 11.378. Cf. Pálffy, op. cit., 142–143 and MOL Batthyány cs. lvt. P 1314, Missiles No. 27.485. 92  Ibid., No. 7762. 93  Jedlicska, “XVI. századi török–magyar levelek,” 705: No. XXIX (1594). MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 4. fol. 94 (1601) and Ibid., No. 158 (1647). 94  Kovács, op. cit., 9: note 2.

Ransom Slavery Along the Ottoman–Hungarian Frontier

941

increase in the number of “fake” beggars on Ottoman territories,95 but even the “true captives” began to abuse such prerogatives as the right of the “iron captives” to transportation by carriage or the right to food provisions. In their desire to collect the ransom fee as rapidly as possible, prisoners on release would sometimes form into groups of bandits and force wealthier settlements in the occupied territories to hand over a carriage or provide for their ransom fees. They would also lease market duties and, as commissioners, would extract money from traders, and pilfer and sell the must of serfs during the vintage, etc. By the 1650s such abuses had become so common that both the Turkish and Hungarian authorities began to stipulate that only iron captives receive transport and provisions. Indeed, both sides ordered the pursuit and capture of the bandits. In 1654, Heves county even passed a decree stipulating on which roads released captives might travel from Eger to Fülek.96 To prevent such violations and to speed up the collection of ransoms, in the middle of the seventeenth century the captives began to be set free in a different way. Released captives were accompanied by one or two fellow prisoners. Such men were referred to as “postmen” (in Hungarian posták), and their task was to assist captives in collecting their ransoms and to check up on them.97 They were called “postmen” (hereinafter I will refer to them as “messenger”) because they were the ones who kept in contact with the bailsmen and the keepers; they forwarded the ransom instalments and delivered any letters written in connection with this. Beggar captives, freed of having to take their instalments to their remote keepers, could thus concentrate their energies on collecting the ransoms.98 95  Sometimes the “true” captives themselves requested to be commissioned by one of the counties to catch and present false captives. Dongó, “Magyar rabok,” (1914), 332: No. XL. 96  Numerous examples from Gyöngyös in the middle and second half of the century are mentioned in Lajos Fekete, “Gyöngyös város levéltárának török iratai [The Ottoman Documents of the Archives of the Town Gyöngyös],” Levéltári Közlemények 10 (1932) 289, 316: No. 72 and Ibid., 11 (1933) 93–140, passim. From Jászberény: Klára Hegyi, “Jászberény török oklevelei [The Ottoman Documents of the Town Jászberény],” Szolnok megyei levéltári füzetek 11 (1988) 5–177, passim and János Botka, “Latin és magyar nyelvű források a Jászság XVI–XVII. századi történetéhez [Latin and Hungarian Sources on the History of Jászság in the Sixteenth and Seventeenth Century],” Szolnok megyei levéltári füzetek 11 (1988) 179–358, passim. 97  MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 148 (1647). 98  In the sixteenth century, the ransoms of captives on release were often brought in by Ottoman or Hungarian border fortress soldiers. This practice, however, did not become firmly rooted because in such cases the opposing side had to issue letters of safe conduct for the soldiers, so that they could travel unhindered in the enemy zone. Jedlicska, “XVI.

942

Pálffy

In the case of the messengers, too, bailsmen were found among the remaining fellow prisoners. Their debts increased even further if the released captives and their messengers made their escape. If this happened and the gaolers could be persuaded, further messengers were sent out (again on bail) to determine the escapees’ whereabouts. If the prisoners agreed to being thrashed, they could even beg for a messenger who returned with less than the amount of his instalment to be allowed out once again.99 If a messenger died before payment of the ransom was due, the other messenger was obliged to bring in his body. If the body was not brought in, or if the messenger died after the payment date, prisoners who were serving as bailsmen had to pay his ransom, too.100 We even know of one case from the 1660s in which an owner of a messenger attempted to collect his ransom fee from the bailsmen, even though the messenger in question had died before the deadline. The Turks in the council (divan) rejected the owner’s arguments, whereupon he turned to the kadı who had just arrived from the distant interior of the Empire and who gave a sentence in his favour. However, both owner and kadı had finally to accept the binding nature of established customary law. Captain-general Ferenc Csáky’s military court in Upper Hungary stated clearly that they should refrain from making “their unusual request, since it is a breach of the ancient law of the marches, if the body has been brought in before the deadline.”101 Finally, where a captive did pay off the remaining amount of the ransom as well as the prison and chain fees to his owner, on his final release he received a letter of discharge confirming that all payments had been made. At such time, his begging letter, the letter of safe conduct, and the letters of bail of his fellow prisoners were destroyed by his keepers. At the same time, he was required to state solemnly that he would not take part in any offensive actions for one whole year.102 Having returned home, many former captives continued to serve in the border fortresses; they also set about paying off their debts incurred through the collection of the ransom fee. századi török–magyar levelek,” 698: No. XIV (1589). In the second half of the seventeenth century, on the other hand, it was not unknown for Ottoman captives to force the inhabitants of Gyöngyös to collect their ransoms. Fekete, “Gyöngyös város,” (1933) 108: No. 134. 99  In 1657 the Christian captives in Buda submitted themselves to 1,000 thrashes in order that a messenger would be released once again. MOL Batthyány cs. lvt. P 1314, Missiles No. 26.646. 100  An unusual example from circa 1650: “He died when he was a messenger in Turkey. His corpse was brought out by the older Ali from Vál”; MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok Nos. 49, 65, 94 and 242. 101  MOL Csáky cs. lvt. P 71, Fasc. 264. Tétel 5. Sedes bellica. October 19, 1665, Kassa. 102   Magyar Simplicissimus, 199.

Ransom Slavery Along the Ottoman–Hungarian Frontier

943

Various methods were used by captives released on bail (or their relatives) to collect the ransom fee. Wealthier nobles would sell off their belongings and then request loans from relatives. Finally, if all else had failed, they would sell off or mortgage their properties.103 The common soldiers of the border fortresses, on the other hand, sometimes had no choice but to set off on long journeys as beggars. They travelled from the Transylvanian border to Laibach in Carniola. Indeed, sometimes they could even be found wandering across territories belonging to the Holy Roman Empire.104 Meanwhile, they would request the Hungarian Chamber in Pozsony or the Aulic War Council to continue paying their salaries for the duration of their imprisonment (or release on bail) in view of their great poverty.105 Often the begging serfs or the poor common soldiers of the border fortresses would turn to neighbouring counties and request “charitable subsidy” (subsidium charitativum).106 Sometimes they would simply request the magistrates of a town to permit them to beg a part of their ransom in front of the local church.107 Indeed, in 1683 even one member of the affluent Kisfaludy family, a man called László, who subsequently became 103  Samu Borovszky, Borsod vármegye története a legrégibb időktől a jelenkorig [History of Borsod County from the Oldest Times to the Modern Age]. I. Budapest, 1909, 294–295. 104  The travels of the captives from Hungary in the Holy Roman Empire are borne witness to by the dozens of passes and patents issued to them by the Chancellery of the Imperial Court Council in the sixteenth and seventeenth centuries. ÖStA HHStA Reichshofkanzlei Paßbriefe Karton 1–18: passim and Ibid., Patentes und Steckbriefe, Karton 1–4: passim. For begging of captives from Upper Hungary in Laibach, see Vilfan, op. cit., 189–190. 105  Petitions of border-fortress soldiers from Győr, Eger, and Komárom from 1570: Mátray, op. cit., 209–210: Nos. I–II. In the mid-seventeenth century, border fortress captain-general Ádám Batthyány intervened at the Aulic War Council in the matter of the salary petition of Gergely Soós and György Felső, commanders of the small border-castle Kányavár in Zala county. MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 32. 106  Thaly Kálmán, “Kossuth Lajos,” 153–154 and Szerémi [Arthur Odeschalchi], “Emlékek Barsvármegye hajdanából. (1439–1711.) [Documents on the History of Bars County (1439– 1711)], Part V,” Történelmi Tár (1892) 539–540: No. CL. Similar begging letters from the seventeenth century to Bars county: Ibid., 538–542: Nos. CXLVIII–CLIII; to Abaúj county: Dongó, “Magyar rabok,” (1910–1916) passim, and to Borsod county: Borovszky, op. cit., 295–296. 107  The captives at Eger wrote in 1663 to the chief justice of Kassa: “Please kindly consider permitting the two messengers presenting this letter [to beg] in front of the church, so that Christians may grant them godly alms.” Lajos Kemény ifj., “Kassa város levéltárából [From the Archives of the Town Kassa],” Történelmi Tár (1897) 573–574. Cf. also the German-language petition of two Croatian soldiers to the town of Laibach (1596): Vilfan, op. cit., 185, as well as the petition of pastor Balázs Maráczi Nagy to the town of Sopron: Deák, “Okiratok,” 110–111: No. I.

944

Pálffy

vice-general of the border fortress district around Győr, had no choice but to accept the charity of the Hungarian bootmakers’ guild in order that he might win back his liberty.108 The counties and the treasury generally gave one or two florins to captives who turned to them with letters beseeching their mercy. In 1570, on hearing of the increasing numbers of such petitions, King Rudolf even permitted the Hungarian Chamber to set aside an annual sum of 300 Hungarian florins for the support of the captives.109 Meanwhile, private individuals, officers and soldiers also demonstrated their humanity in the form of charitable actions. They knew very well that they too might one day share the miserable fate of the poor captives.110 In 1649 the garrison troops in Kiskomárom (near Lake Balaton) donated one whole month’s salary in order to assist the fortress’s vice-captain, who was held captive in Kanizsa.111 Indeed, in the late seventeenth century the humanist Kristóf Lackner (mayor of Sopron) provided in his will for the establishment of a special foundation to support fellow citizens imprisoned by the eternal enemy of Christendom (Erbfeind des Christlichen Glaubens). Lackner’s foundation was able to offer great sums in support of the prisoners. Often sums of 50–100 florins were transferred, as against the very modest subsidies of the counties and the Hungarian Chamber, which normally consisted of just a few florins.112 At the turn of the sixteenth and seventeenth centuries, the city of Laibach rendered similar financial assistance to its imprisoned inhabitants and also permitted captives released on bail to beg in the city.113 In the seventeenth century, a number of redemption societies were established in territories belonging to the Holy Roman Empire. Finally, during the wars of re-conquest against the Ottomans (1683–1699), an active role in the redemption of captives in Hungary was also played by a holy order that had been established with the aim of liberating the Christian captives, the Ordo Sanctissimae Trinitatis.114

108   Alfréd Lengyel – Elemér Lovas, “Kisfaludy László, győri alkapitány levelezése [Correspondence of László Kisfaludy Vice-general of Győr],” Győri Szemle 10:2 (1939) 90: No. 18. 109  Mátray, op. cit., 209: No. I. 110  Dongó, “Magyar rabok,” (1914), 169: No. XXXII. 111  Varga, “Rabtartás és rabkereskedelem,” 130. Cf. some similar examples in Takáts, “Magyar rabok, magyar bilincsek,” 522. 112  Kovács, op. cit., 6–11. 113  Vilfan, op. cit., 184–192. 114  Ferenc Fallenbüchl, A rabkiváltó trinitárius szerzetesek Magyarországon [The Monks of the Ordo Sanctissimae Trinitatis in Hungary]. (A Szent István Akadémia Történelem-, Jog- és Társadalomtudományi Osztályának értekezései, II/9.) Budapest, 1940, 9–14, 32–41.

Ransom Slavery Along the Ottoman–Hungarian Frontier

945

Apart from these unselfish acts of charity, there were also cases in which contracts were signed by lords with captive private or common soldiers (or with their relatives). Under the terms of such contracts, seigneurs would agree to pay the captives’ ransoms, if the captives or their families pledged to enter into their service on release.115 Of course some seigneurs subsequently took advantage of this situation. Breaching the contracts, they forced the former captives into permanent and unending service.

The Representation of Interests in the Prisons

From the second half of the sixteenth century, the customs relating to the ransom fees, primarily the institution of bail, and the awful conditions prevailing in most prisons required that prisoners take a common stand. This process of communal development was speeded up by an increase in the number of prisoners in the first half of the seventeenth century; an increase that is well demonstrated by prison population figures for the Stump Tower in Buda. Whereas prisoners in the Constantinople prison of the Seven Towers (Yedikule) came from various enemy states and thus formed very distinct groups (Venetians, Germans, Hungarians, etc),116 captives in the prisons of Ottoman Hungary and the Kingdom of Hungary almost without exception shared the same nationality and religion. Such homogeneity facilitated the taking of a united stand; prisoners were more capable of expressing their concerns either to their keepers or to the pashas and captain-generals who were based just some miles away. The prisoner rebellions of the period—such as the aforementioned rebellion at the Batthyány castle in Szalónak in 1652 or the Várad rebellion 30 years later—proved to the owners of the captives that the best way of securing payment of the ransoms was to permit the formation of organisations representing the interests of prisoners. In addition, the disturbances of the seventeenth century served to accelerate the process by which smaller settlements and communities (including the prisoner communities) joined together to defend their own interests. The spokesman of the prisoner community was the so-called “prisonsteward” (Hungarian: rabgazda, German: Wirt). As the diary of Johann Ferdinand Auer from Pozsony states, in accordance with the custom of the border 115  For examples from the seventeenth century: Varga, “Rabtartás és rabkereskedelem,” 131 and Jedlicska, Eredeti részletek, 112: No. 209. Sometimes, however, the captives were willing to pledge themselves to service until death. Borovszky, Borsod vármegye, 296. 116   Auer János naplója, 191–196.

946

Pálffy

fortresses, the gaolers chose the steward from among the more senior captives.117 (In doing so, they probably acted with the consent of the prisoners.) These spokesmen were generally required to represent and defend the interests of the captives and to communicate with the Ottoman owners and gaolers, who did not always speak Hungarian. In the first half of the 1660s, for instance, the office of steward was held in Buda by István, a cavalryman from the fortress Érsekújvár who had been a captive for nine years. In Várad and in Eger it became customary for two stewards to be elected. At the Stump Tower in Buda, which was the largest prison in the occupied territories, a clerk was elected alongside the steward. The clerk was required to manage prisoner correspondence and to issue the various documents (petitions, letters of bail, pledges, etc). This post was held in the early 1650s by Péter Rácz and from 1656 until 1657 by Géci Sentei.118 Furthermore, according to a piece of evidence from 1661, prisoners in the Stump Tower even had their own priest, a man called György Szolnai.119 Thus, the prisoners’ interests were represented in a manner that resembled the decision-making process of a village council. The primary task of the prison steward was to communicate and maintain contact with the local Ottoman dignitaries (pashas, beys, and kaimakams, etc) and the prison keepers (dizdars and gaolers).120 They had to represent the interests of the captives in a diplomatic manner that was satisfactory to both parties. In the centres of the vilayets and sancaks it became customary for the pashas and beys to hold special councils (divan) in which they debated the captives’ ransoms and other issues.121 Stewards and clerks (accompanied by other prisoners) were able to take part in these discussions, too. It was during the divan that the owners of captives received their ransoms and that captives were forced into becoming bailsmen. Moreover it was here that letters 117  Ibid., 122: Überall in denen Kranitzhäussern der Gebrauch, das einer untter den Gefangenen von den Tömlitzer zum Haubt über die andern, so hernach der Wirth genennet wirdt, gemacht, welcher Sorg und Aufsiht vor die andern tragen muess. 118  For Rácz: MOL Batthyány cs. lvt. P 1314, Missiles Nos. 15.615, 15.617, 20.020, 26.681 and 27.310. According to the captives’ register of 1654 he was a soldier from Egerszeg, whose ransom had been determined by the Ottomans at 900 thalers. Ibid., Nos. 7763 and 27.310. For Sentei: Ibid., Nos. 9278, 9289–9296 and 42.841. 119  It is hard to say whether we should consider the priest mentioned in the published document as caring for the captives’ spiritual lives. Nagy, “A budai csonkatorony,” 662. 120  MOL Batthyány cs. lvt. P 1314, Missiles No. 9299 (1660). Cf. András Kubinyi, “Rabok feliratai a budai Csonkatoronyban [The Inscriptions of Captives in the Stump Tower of Buda],” Budapest régiségei: A Budapesti Történeti Múzeum évkönyve 18 (1958) 523: note 16. 121  “His Greatness, the bey called us to him, since the Turks had divans, for the reason that they would [debate] the matter of the prisoners” (Szolnok, 1656); Thaly, “Szólnoki rabság,” 215. Cf. Fekete, Türkische Schriften, 243: No. 9.

Ransom Slavery Along the Ottoman–Hungarian Frontier

947

to the captives (and to the Ottomans concerning the captives) were read out and translated into Turkish by the prisoners’ clerks or by the Hungarian scribes of the pashas and beys. The divan also provided an opportunity for the Ottoman officers in Buda, the kadı and the jurisconsult (müfti), to decide in matters relating to the ownership of captives and the payment of their ransoms.122 Thus most tasks of the stewards were related in some way to the ransom fee. Apart from their role in the bargaining process, stewards also advised fellow captives in matters such as the choice of bailsmen and messengers. Similarly, it was the stewards who managed funds bequeathed to the prisoner community. Generally, such funds were used to purchase groceries. In the course of the seventeenth century, it became the accepted custom among the soldiers of the border fortresses and the nobles of the border districts to bequeath a sum of several florins to prisoners in the neighbouring Ottoman areas, especially to those imprisoned in the Stump Tower of Buda.123 There were other benefactors, too (for example, the market towns of the Great Plain).124 Their donations of cash and food provisions were usually handed over to the leaders of the prisoner community by the Buda magistrates. Meanwhile, those Christians who were still residing in Buda or Pest gave food to the captives in return for military tax exemptions. Indeed, when in 1596 Miklós Pálffy, captaingeneral of the mining border district, attempted to force citizens of the two towns to pay tax, it was the local prisoner community that rose to the defence of the local inhabitants. The prisoners confirmed that the townspeople had been providing them with food.125 Nevertheless, ensuring the prisoners’ food supply was no small task for the steward. While prisoners would sometimes be allowed to go and have meals at the houses of local burghers, the Christian inhabitants (who were decreasing in number) were unable to meet the needs of a rapidly expanding prisoner population. Since the prison ransom paid by the captives to their keepers only covered the expenses of their guards, the prisoners could not 122  MOL Batthyány cs. lvt. P 1314, Missiles Nos. 9278, 9284, 9289, 9294 and 9296. 123  József Horváth, Győri végrendeletek a 17. századból. I. 1600–1630. [Seventeenth-Century Testaments from Győr]. Győr, 1995, 30: No. 4 (1604); ibid., II. 1631–1654. Győr, 1996, 168: No. 212 (1652) and MOL Daróczy cs. lvt. [Archives of the Daróczy family] P 961, Tétel 1: Paksy család iratai [Documents of the Paksy family] fols. 42–46 (1663). 124  Nagykőrös, 1630: Áron Szilády – Sándor Szilágyi, Okmánytár a hódoltság történetéhez Magyarországon [Archive on the History of Ottoman Rule in Hungary]. (Török–magyarkori történelmi emlékek I: Okmánytár, I.) Pest, 1863, 18: No. XVIII and 1650: ibid., 161: No. LXIX. 125  Pál Jedlicska, Adatok erdődi báró Pálffy Miklós, a győri hősnek életrajza és korához 1552–1600 [Data on the Era and Biography of the Hero of Győr, Miklós Pálffy, Baron of Erdőd]. Eger, 1897, 593: No. 928/b.

948

Pálffy

expect to be supplied with provisions, and thus many of them went hungry.126 Their requests for assistance sometimes fell on deaf ears. The Ottomans therefore permitted captives released on bail to bring food (meat and bread bought/ begged for in the Kingdom of Hungary) to their fellow prisoners. In the spirit of reciprocity, this practice was also permitted by the other side—although on a number of occasions it was forbidden after captives on release failed to return. Subsequently, the Ottomans would encourage the Hungarian captain-generals to follow their example and permit the transfer of goods once again.127 Subject to the permission of the gaolers, stewards could also travel to areas in the Kingdom of Hungary in order to manage the affairs of their fellow prisoners. Such journeys appear to have been commonplace. This would explain the fact that both in Eger and in Várad, two stewards were elected by the prisoner community; if one steward was absent, the other could continue to manage the affairs of the prisoners. We know of several interesting visits made by stewards to the royal territories in the second half of the seventeenth century. In 1661, for example, András Posgay, the steward in Buda, acting on behalf of his fellow prisoners, sold a property in the district of Pórládony (Sopron county), which had come to be owned by the prisoners. The whole story was as follows: A prisoner on release was making his way back to Buda with a part of his ransom when he was murdered and robbed in the Bakony hills by a man from Tüskevár. The murderer was soon identified and caught by his lords. Subsequently, the fellow prisoners of the dead man offered up the murderer to the Ottomans in lieu of the deceased. The Ottomans, however, rejected the offer. Given that they had not received the body of the deceased captive, the Ottomans demanded the ransom sum of 2,000 thalers from his bailsmen in accordance with customary law. The court of the lord (sedes dominalis) subsequently sentenced the guilty man from Tüskevár to death and ordered that all his property and belongings be given to the captives. Thus, a house-session in the village of Tompaháza (Sopron county), together with the neighbouring wood and a strip of agricultural land, passed into the ownership of the prisoners. In the spring of 1661, this property was sold to István Bakodi and his wife in the presence of lieutenant János Niczky, András Posgay (the prisoners’ 126  The awful consequences of this practice are vividly and accurately described in the 1663 diary of Johann Ferdinand Auer, a citizen of Pozsony who was being held close to the Stump Tower; Auer János naplója, passim. 127  See some examples from the seventeenth century: MOL Batthyány cs. lvt. P 1314, Missiles Nos. 7763, 9279, 9293 and 26.681. Request for food to Ádám Batthyány from the Stump Tower: Ibid., No. 9307. Similar request by the Ottoman prisoners of Szalónak (castle of Ádám Batthyány) for provisions for the Hungarian captives of Kanizsa: MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok Nos. 118, 123 and 133.

Ransom Slavery Along the Ottoman–Hungarian Frontier

949

steward), György Szolnai (the prisoners’ priest), and a further prisoner. The sum received was then used by the captives to pay off at least part of the ransom of their murdered companion.128 In the second half of the 1660s, the stewards of Eger and Várad, accompanied by several other older prisoners, travelled to the military court of Ferenc Csáky (captain-general of Upper Hungary) on several occasions.129 As in the case of the Fülek military court case of 1680 cited above—in which an iron captive had sued the royal fortress castellan—in these latter instances the prisoner communities of Várad or of Eger took a number of personages from Royal Hungary to court. At the military courts it was primarily the stewards who defended the interests of the prisoners. For instance, in the autumn of 1665 the prisoners’ steward in Várad, Mihály Dobai, commenced an action against a serf woman belonging to the landowner Mrs Zsigmond Homonnai Drugeth. The serf woman’s son had been serving his time as a captive of the ağa in Várad. He had then been released with his messenger on bail. The boy had then returned to Terebes to visit his parents. As the boy’s escape was made impossible by the messenger, he and his parents elaborated a most vile plan. One night “they suffocated the messenger, struck him on the head, and killed him”. When, subsequently, the two captives failed to reappear, the gaolers took disciplinary action. The prisoners then decided to send some more messengers out to search for the boy and his messenger. They managed to find the boy and brought him back to the prison. His companion, however, they found to be dead and buried. Since all this happened when his term had already expired, according to customary law the guarantors were required to undertake to pay the ransom themselves. However, referring to the same customary law, they claimed that “it is the law among Ottomans and Hungarians that the ransom (of a murdered person) be paid where the murder has taken place.”130 Through the intervention of captain-general Csáky and Palatine Ferenc Wesselényi, the prisoners demanded the messenger’s ransom from the wife of Homonnai Drugeth, the owner of the supposed killers. But she was slow to pay up and therefore the prisoners requested that the murdering “old woman” be handed over to them. However, they first had to prove that she really was one of the perpetrators of the crime.131 It was at the above military court, rather than the county civil court, that the captives of Várad brought a lawsuit against the nobleman Sándor Semsey 128  Nagy, op. cit., 661–663. 129  On the organisation of the military court (with examples), see Pálffy, op. cit., 115–132. 130  Izsépy, “Rablevelek,” 316. 131  MOL Csáky cs. lvt. P 71, Fasc. 264. Tétel 5. Sedes bellica. October 19, 1665, Kassa. Cf. also Izsépy, op. cit., 320–321.

950

Pálffy

of Abaúj county. The story ran as follows: Semsey sold a Turkish captive to János Fekete, a prisoner in Várad, who wished to hand over the Turkish captive in return for his own release. However, the Turks in Várad did not agree to this exchange. Semsey resold the Turk, even though by now he was the property of Fekete. The second purchaser was a Hungarian captive in Eger, who in possession of the Turk succeeded in negotiating his own release. Fekete then left Várad accompanied by a messenger with his fellow prisoners acting as bailsmen. However, instead of searching for Semsey, shortly after their release the two men took to their heels and fled. The gaolers then tortured the bailsmen and demanded the payment of the escaped prisoners’ ransoms. The bailsmen turned to Csáky’s military court, asking that Semsey return at least the sum received for the Turk from Fekete.132 We know that similar things happened to Turkish captives, too. In 1640 bailsmen in the dungeons of the castle of Fülek demanded 900 guruş blood money from the town of Gyöngyös after the murder of their fellow captive who had been begging at the town’s market, but this time their attempt remained unsuccessful.133 In addition to such cases at the Hungarian military courts, it would seem that the prisoner communities gave expert opinions on matters concerning their Ottoman fellow captives, and sometimes acted as a “court of justice” and passed judgements (not infrequently under the pressure of necessity). For example, when a captive on release failed to reappear, in order to ensure payment of the ransom a ruling was passed in common by the captain-general and the prisoners.134 In 1652, the Christian captives of the Stump Tower determined the ransom issue of their fellow captive on release, Ferenc Palotai, in the same manner.135

Authentication of Documents in the Prisoner Communities

The issuing of documents and petitions (letters of bail, pledges, etc) was a difficult task for the steward. In Buda and other Ottoman border fortresses it became customary in the early seventeenth century for the prisoners to employ a clerk. The clerk was required to compose the imploring letters of his fellows day after day and in a uniform style and structure.136 Authentication 132  Ibid., 321–322. 133  Fekete, “Gyöngyös város,” (1932) 316: Nos. 70–71 and Idem, op. cit., (1933) 113–114: Nos. 154–155. 134  Jedlicska, Eredeti részletek, 138–139: No. 259 (1635). 135  MOL Batthyány cs. lvt. P 1313, Török vonatkozású iratok No. 343. 136  Izsépy, op. cit., 317–318.

Ransom Slavery Along the Ottoman–Hungarian Frontier

951

of written documents was also necessary, because captives on release, who were delivering their letters in person,137 often had to prove that they were on the right road. Owing to the general disorder in the country, the number of false captives carrying false documents soared. Official documents could easily be forged. Authentication of written documents in Hungary was usually performed with the aid of a seal (sigillum). The seventeenth century therefore became the golden age of Hungarian seals.138 Before long the Hungarian prisoner communities of the occupied territories were also using seals. It would seem that the seals were guarded by the stewards, who used them to authenticate documents prepared by the clerks. Since few of these documents have survived, we should use caution when interpreting the employment of seals by prisoners. Still, a comprehensive examination of these seals would doubtless produce some interesting results. Such a study is very much needed for the use of Hungarian seals in the early modern period remains largely unexplored.139 The data introduced below will hopefully demonstrate some of the possibilities. The earliest “prisoner seal” dates from a letter written on February 15, 1651 by the prisoners of the Stump Tower in Buda to Ádám Batthyány (illustration No. 1).140 Although the seal is small and somewhat clumsy, the initials “CT” clearly demonstrate that the seal was used by the Christian captives of the Csonka Torony to authenticate their documents. Unfortunately we still do not 137  MOL Batthyány cs. lvt. P 1314, Missiles Nos. 9307 and 35.722. 138  For this process, see: József Lugossy, “Két magyar köriratú pecsét 1500. évből [Two Seals of 1500 with Hungarian Inscriptions],” Magyar Történelmi Tár 1 (1855) 172–192. Kálmán Thaly, “Régi magyar községi pecsétek [Old Hungarian Community-Seals],” Századok 3 (1869) 571–575. Gy[ula] N[agy], “Régi magyar községi pecséteink statistikája [Statistics of the Old Hungarian Community-Seals],” Századok 5 (1871) 513–517 and Bernát L. Kumorovitz, “A magyar szfragisztika múltja [The History of Hungarian Sphragistics],” in Emlékkönyv Szentpétery Imre születése hatvanadik évfordulójának ünnepére [Festschrift for Imre Szentpétery to his 60th Birthday]. Budapest, 1938, 274–278. Among the various seals of the authorities, we know of one from 1680, that of the military judge of Fülek that had merely a Latin text (with mistakes): SIGILLV[M] IVDICII BELLICIS [sic!] PRESIDII FILEKIENSIS. MOL Esterházy cs. lvt. P 125, Pál nádor No. 11.378, and Pálffy, op. cit., 144, and 228: note 2. Cf. also the interesting Peasant county seals: Ferenc Szakály, Parasztvármegyék a XVII. és XVIII. században [The Peasant Counties in the Seventeenth and Eighteenth Centuries]. (Értekezések a történeti tudományok köréből, Új sorozat, 49.) Budapest, 1969, 142–143. 139  Bernát L. Kumorovitz as early as in 1938 opposed the notion that “sphragistics should above all deal with the medieval material”; see Kumorovitz, “A magyar szfragisztika,” 278. 140  MOL Batthyány cs. lvt. P 1314, Missiles Nos. 50.743 and 35.790. Here I should like to express my thanks to Sándor Kákonyi for his drawings of the seals. (The seals of the captives appear in their original size and form in the attached illustrations.)

952

Pálffy

know when the prisoners began to use this seal. We may state, however, that there are no traces of the seal on documents issued in 1562, 1575, or even in 1605.141 On one of the prisoner-letters dating from 1636, we find a round-shaped seal that is divided into eight equal pieces. In the following decades, the seal of an imprisoned nobleman or a Turkish official was commonly attached to documents.142 Thus, we may infer that seals began to be used by Christian prisoners in the Stump Tower in the 1640s. (However, many documents were issued without seals even after this time.) We came across another seal of the Buda captives stemming from the autumn of 1651 in the valuable missilis-collection of the archives of family Batthyány (illustration No. 2).143 This seal was used until at least 1660. The seal is larger than the others. Above the “CT” inscription, we see a depiction of a foot cuff linked by three rings, which was one of the most widely-used Ottoman shackles in the seventeenth century.144 The monogram of the seal was usually pressed into red wax or sometimes on to a piece of paper. Beneath the monogram, we see a depiction of another form of shackle. To the left and right of this depiction, there are strange-looking instruments, which were perhaps used by the Ottomans to torture people (probably hand and tongue stretchers or a hangman’s chair). In the 1650s, the captives were also using a smaller seal (illustration No. 3), in actual fact a reduced version of the larger one. The smaller seal very closely resembled the larger one. Even so, in addition to the obvious difference in size, there were a number of other minor differences, too. For example, on the smaller seal the letters “C” and “T” were bolder, while above these letters there was a depiction of a cuff held together by an iron rod rather than chain links. Since undated letters (from the 1650s) are the only surviving documents demonstrating the use of the smaller seal, we cannot exclude the possibility that the smaller seal preceded the larger one; we simply do not have the evidence. It seems certain, however, that the two seals were in use at the same time, for they both appear on letters issued by the clerk Péter Rácz 141  1562: The letter of György Csarkó, border fortress soldier from Komárom to King Ferdinand I (1526–1564): ÖStA HHStA Türkei I. (Turcica) Karton 16. Konv. 2. fol. 133; 1575: The letter of Benedek Vadlövő to Péter Melith, commander of Diósgyőr: MOL Zichy cs. lvt. P 707, Missiles No. 2102; Two letters from 1605, one by Ferenc Wathay, the other by István Komornyik, to Ferenc Batthyány: MOL Batthyány cs. lvt. P 1314, Missiles No. 51.262 (Wathay’s letter has been published on several occasions, see for example Wathay Ferenc énekes könyve, II. 164–165) and Ibid., No. 27.083. 142  1636: Ibid., No. 9277; 1646: Ibid., Nos. 9279–9280; 1649: Ibid., No. 35.722, and without year: Ibid., No. 9309. 143  September 4, 1651: Ibid., No. 31.527. 144  Temesváry, Büntető eszközök, 54.

Ransom Slavery Along the Ottoman–Hungarian Frontier

953

(mentioned above). More seals were needed for two main reasons. First, there was an increase in the number of documents issued (after a rapid increase in the number of prisoners from the middle of the 1650s). Second, the stewards, who guarded the seals, continued to issue letters of bail and other documents even when they were travelling in royal Hungary. One letter of the scribe Rácz dating from the early 1650s carries a most remarkable seal depicting the Stump Tower itself, on a so-called Renaissance shield and in stylised form (illustration No. 4).145 Above the shield, we see the initials “BCT” (= Budai Csonka Torony)—a conclusive proof that the depicted building is really the Stump Tower. It stood at the intersection of the east-west and north-south defence lines of the castle (today the inner courtyard of the Royal Palace is located at this site). The tower was the most important prison in Ottoman Hungary. Tradition has it that its name reflects the fact that construction (begun during the reign of King Sigismund of Luxemburg [1387–1437]) was never finished and thus the “stump” of the building remained exposed to all weathers.146 Although captives were kept in the tower in the middle ages,147 it was during the Ottoman occupation that it became a real prison. According to one source, by 1551 Ali pasha of Buda was already keeping a large number of Christian prisoners in the tower.148 The building—the wooden beams of which fell on to prisoners at the time of the great gunpowder explosion of 1578 (killing many of them)—was surrounded by a guardhouse and the barrack-like houses of the gaolers, Janissaries, and other Ottoman

145  MOL Batthyány cs. lvt. P 1314, Missiles No. 15.615. 146  “Opposite stands the great building of Sigismund [of Luxemburg], king of Hungary, it gives an impression more for its size than for its beauty. Apart from a square tower, it has six other towers, [the square tower] is incredibly wide, and because it remained unfinished it is called the Stump Tower” (Description of Gaspar Ursinus Velius in 1527). Florio Banfi, “Buda és Pest erődítményei 1686–ban [The Fortifications of Buda and Pest in 1686],” Tanulmányok Budapest Múltjából 5 (1936) 105. “Unfinished and without a roof, the rain falls in, therefore it is called the Stump Tower, that is a half-built truncated tower” (Hans Dernschwam’s description form the middle of the sixteenth century). Hans Dernschwam, 121. Cf. the archaeological discovery of the tower: László Gerevich, “A Budai Vár feltárt maradványainak leírása [Description of the Unearthed Remains of the Castle of Buda],” in Budapest műemlékei I. [National Monuments of Budapest]. (Magyarország műemléki topográfiája, 4.) Budapest, 1955, 248–251. The inscriptions carved by captives kept in the tower were published by Kubinyi, “Rabok feliratai,” 519–522. 147  MOL Dl. 24 005 and Kubinyi, op. cit., 519–520. 148  The letter of Imre Thelekessy to Ehrenreich von Königsberg, November 8, 1551, Komárom: ÖStA KA Alte Feldakten 1551/11/14.

954

Pálffy

residents.149 The tower itself, as demonstrated by the seal, was a building of several floors. The cells varied in terms of both size and “comfort”. The worst fate awaited those captives who had been assigned to the lower and darker dungeons (for example, the so-called Alsókula, i.e. “Lower kula”) or to the “extremely cold torture cellars”. Another area of the building to which prisoners were sent by way of punishment was the so-called Büdös Kula, i.e. “Smelly kula”. During the day, however, most prisoners worked outside on the prison courtyard. It was here that they would issue documents carrying their own seal.150 The seals used by the prisoners were fully recognised; they were, in practice, the official and authentic seals of the prisoner community. This is shown by the seal employed (according to our present knowledge) by the prisoners between 1657 and 1661 (illustration No. 5). This seal probably replaced both the smaller and larger seals later on. The newer seal was very different from what had gone before, although the depictions were quite similar. The circle of words in Hungarian (B[udai] CZONKA TORONY = Stump Tower in Buda) reflected the new standard practice of Hungarian seals in this period. However, the inner part of the seal with its depictions of two forms of shackles and imagined implements of torture, symbolised continuity. In 1661, when agreeing to the sale of the property in Sopron county (mentioned above), the prisoners referred to this seal as their “usual” seal: “We give testimony to this our letter with the signature of the steward of our miserable crowd, and confirming it with our usual seal.”151

149  Endre Veress, “Musztafa budai basa álma, s a nagy lőpor-robbanás [The Dream of Mustafa Pasha of Buda and the Great Gunpowder-Explosion],” Történelmi Tár (1896) 742. Cf. Monumenta Hungariae Judaica, X. 1150–1766. Ed. by Sándor Schreiber. Budapest, 1967, 98–99: No. 68: Viel gefangne Christen seyn darinn von dem Fewer verdorb. See further Szakály, Mezőváros és reformáció, 277: note 283: Turris quoque, quem Chongga seu curtam et intectam vocant, in quo passe captivi Christiani magNo. numero detenti fuerunt, adeo discussa sit, ut tabulata, sub qua captivi agunt, collapsa, miseros oppresserit, quod tres vivi tantum eximi potuerin. A detailed description of the explosion in 1578 by Gábor Ágoston, “Ottoman Gunpowder Production in Hungary in the Sixteenth Century: the Baruthane of Buda,” in Hungarian–Ottoman Military and Diplomatic Relations in the Age of Süleyman the Magnificent. Ed. by Géza Dávid – Pál Fodor. Budapest, 1994, 158–159. 150  For the lower dungeon: MOL Batthyány cs. lvt. P 1314, Missiles Nos. 9288, 9300 and 26.680. For the cold cellar: Ibid., No. 9307. For the “Smelly kula”: Ibid., No. 9298. For the prisoncourt: Ibid., Nos. 9284, 9297, 9299 and 31.525. 151  Nagy, op. cit., 662.

Ransom Slavery Along the Ottoman–Hungarian Frontier

955

The prisoners of the Stump Tower cut a new seal for the final time in 1678 (illustration No. 6).152 The outer circle of words of this egg-shaped seal contained more words than before (BVDAI CZONKA TORONY RABOK P[ECSÉTJE] = seal of the captives in the Stump Tower in Buda). The inner part depicted the coats-of-arms on a decorated Renaissance shield in accordance with the rules of heraldry and sphragistics. The three fetters resembled the depictions of shackles on the earlier versions, while above the shield the year 1678 was carved with an inverted seven. Although the development of depictions and inscriptions on these seals differed from that of the village communities, the seals themselves constituted a significant element of the increasing prerogatives obtained by prisoner communities. In addition to the seal of Buda, we know of the seals of the prisoners in the Ottoman fortresses of Esztergom (illustration No. 7) and Vác used from the 1650s.153 In Esztergom the seals were probably prepared in the first half of the decade. Before and during this period the prisoners seem to have used the seals of nobles to authenticate their documents.154 Later, however, this practice became unknown. Instead the prisoners used their own seal which carried a long inscription in Hungarian (ESTERGAMI TOMLOCZBEN NIOMORGO / RABOK = the destitute captives in the prison of Esztergom) on the outer circle. Inside of this circular inscription, we can see a long-haired figure standing in a cloak whose hands are tied behind the back with a pair of foot straps; there is also a six-pointed star with a moon on his right, and a sun on his left. In 1683, as a prisoner at Esztergom, László Kisfaludy used the normal seal of the prisoner community to authenticate his letters: “By way of stronger testimony I gave my sealed letter using the seal of the community of the poor captives and confirming it by my handwriting.”155 Yet, as we know, Kisfaludy generally authenticated his letters with a private seal bearing the initials “KL”. The widespread use of the prisoner seals is demonstrated by the fact that we know of seals being used among the prisoner communities in Várad 152  In 1681 the captives set their seal in green wax when writing to the nobleman Péter Káldy: MOL Sibrik cs. lvt. [Archives of the Sibrik family] P 1865, Tétel 4: Káldy cs. lvt. [Archives of the Káldy family] Aláírás nélküli levelek [Unsigned letters] fol. 46. 153  The seal from Esztergom that was set in red wax or onto a small piece of paper (1654– 1662): MOL Batthyány cs. lvt. P 1314, Missiles Nos. 50.731–50.736 and 50.738. The seal of Vác with the monogram “VR” and two crosses (1659) is known only from a description made in the nineteenth century: Albert Nyáry, “Váczi rabközség pecséte, 1659 [The Seal of the Prisoner Community of Vác in 1659],” Századok 8 (1874) 585. 154  MOL Batthyány cs. lvt. P 1314, Missiles No. 11.033 (1652) and No. 50.736 (1657). 155  Lengyel – Lovas, “Kisfaludy László,” 89: No. 15. For his own seal, see Ibid., 92: No. 21.

956

Pálffy

(illustration No. 8), Eger (illustration No. 9), and Szolnok.156 Within two years of the fall of Várad, in 1662 Christian captives in the city had cut themselves a seal with the circular inscription VARADI SZEGENY RABOK PETSETI (the seal of the poor captives in Várad). The seal of the captives in Eger was relatively simple and carried the monogram “EG”. Nevertheless, this seal was just as effective, in terms of its function, as any of the more decorative ones.157 Evidence uncovered to date indicates that Ottoman prisoners in the Kingdom of Hungary did not use prisoners’ seals in letters to their keepers.158 If they wished to authenticate their letters, they would do so with the private seal of the border fortress captain or one of his senior officers. In 1683, for example, “the poor Turkish captives” in Léva wrote to captain-general Pál Esterházy and confirmed their letter with the seal of vice-captain György Farkas carrying the inscription FARKAS GIORGI.159 Turkish messengers and captives on release would send letters from Buda (or some other place in Ottoman Hungary) to their guarantors. Each letter would carry the seal of a Turkish

156  The seal with the inscription SZONAKI SZEGINY RABOK PECZETI (the seal of the poor captives in Szolnok) in the middle of the crown shield and depicting a foot chain like the seal from Buda was published by Thaly, “Régi magyar községi pecsétek,” and “Szólnoki rabság,” 215–216. For the incomplete and misunderstood descriptions of the seals from Várad and Eger, see Izsépy, “Rablevelek,” 317 and “Az egri törökök,” 160. At the same time, we have no explanation for the fact that we never find a seal on the numerous letters of the captives of Kanizsa. However, in the small prison of Kaposvár in 1648 the captives’ letters were authenticated using a Hungarian nobleman’s seal. MOL Batthyány cs. lvt. P 1314, Missiles No. 31.522. 157  It is worth noting that the prisoner communities in Istanbul also used some kind of a seal, however, we could not give meaning to the illustrations on the two dry seals known to us, despite using an ultra-violet lamp. 1593: HL Törökkori Gyűjtemény [Collection of the Turkish Period] 1593/12, and without date, but from the middle of the seventeenth century: MOL Batthyány cs. lvt. P 1314, Missiles No. 35.983. 158  See the letter of the Ottoman captives of Németújvár to Ádám Batthyány (December 20, 1653): Ibid., No. 33.635. Even recent Ottomanist research into seals has no information on these. A good summary: Claudia Römer, Osmanische Festungsbesatzungen in Ungarn zur Zeit Murāds III. Dargestellt anhand von Petitionen zur Stellenvergabe. (Österreichische Akademie der Wissenschaften, Philosophisch-Historische Klasse, Schriften der BalkanKomission, Philologische Abteilung, 35.) Wien, 1995, 103–116. 159  MOL Esterházy cs. lvt. P 125, Pál nádor No. 5928. For another example of the use of a nobleman’s seal by an Ottoman captive from Sárvár, the castle of the Nádasdy family (1627), see MOL Esterházy cs. lvt. P 123, Miklós nádor [Palatine Nicolaus Esterházy] Tétel II/i. fol. 120. For the Turks in Hatvan (1680), see MOL Esterházy cs. lvt. P 125, Pál nádor Nos. 6798–6799.

Ransom Slavery Along the Ottoman–Hungarian Frontier

957

 ungarian prisoner seals from the H seventeenth century.

official or a Hungarian nobleman.160 And while the Ottoman prisoners did not use their own seals, they doubtless sent their letters to inform their fellow countrymen about the intentions of the Christians. For their part, the Hungarian captives supplied valuable information about impending Ottoman raids161 to the border fortress captain-generals in their sealed and rarely censored letters.162 The liberation war at the end of the seventeenth century (1683–1699) marked the beginning of a new era in the life of Hungary and Central Europe. The defeat of the Ottomans and the establishment of the military border zones (Militärgrenze) concluded 150 years of Hungarian–Ottoman coexistence. The customary law of the border zone—a series of rules and regulations that determined the daily lives of tens of thousands of people—ceased to exist for ever. We have been left with little or nothing to remind us of those peculiar customs. 160  Turkish seal: MOL Batthyány cs. lvt. P 1314, Missiles No. 9309 (undated) and No. 49.233 (1655). Seal of a Hungarian nobleman: Ibid., No. 49.224 (1652). 161  MOL Batthyány cs. lvt. P 1314, Missiles No. 35.788. Cf. Ibid., No. 9307. MOL Esterházy cs. lvt. P 125, Pál nádor No. 6796 (Eger, 1678) and Izsépy, “Rablevelek,” 324–326; see further Jenei, “Iratok,” 206: No. 34/a and Jedlicska, “XVI. századi török–magyar levelek,” 699–700: No. XIX. 162  Cf. footnote 26.

The Black Sea and the Slave Trade: The Role of Crimean Maritime Towns in the Trade in Slaves and Captives in the Fifteenth to Eighteenth Centuries1 Mikhail B. Kizilov The son went as a hostage and the daughter went off to the Crimea [i.e., to slavery] modern Kazakh proverb2

Situated at the juncture of trade routes leading from Italy and Byzantium (later Turkey) to Poland, Russia and the East, the Crimea had always been attractive to international commerce, of which the slave trade was one of its most important components. The earliest data on the slave trade in the region dates back to antiquity. By the late Middle Ages this business was carried out mainly by Italians, with the Tatars as the most important purveyors of “live merchandise.” Numerous published and archival sources (accounts of European and Ottoman travellers, diplomatic correspondence, letters and memoirs of captives, Turkish defters (registers), Russian and Ottoman chronicles, etc.) show that the most important role in the trade in slaves and prisoners of war was undertaken by maritime Crimean towns, such as Caffa (with the dependent ports of Taman, Azak and Kerş) and Gözleve.3 Another important centre of the slave S ource: Kizilov, Mikhail B., “The Black Sea and the Slave Trade: The Role of Crimean Maritime Towns in the Trade in Slaves and Captives in the Fifteenth to Eighteenth Centuries,” International Journal of Maritime History, 17(1) (2005): 211–235. © 2005 by SAGE Publications, Ltd. Reprinted by Permission of SAGE Publications, Ltd. 1  An earlier version of this article was read at the Fourth International Congress of Maritime History in Corfu in June 2004. I would like to thank my co-panellists and the audience for their insightful comments and suggestions. In this essay I use a slightly simplified system of transliteration of Turkish and Arabic names and terms. 2  “Ulı ırımğa, qızı Qırımğa ketti,” cited and translated by Peter Golden, “The Codex Cumanicus,” in H. Paksoy (ed.), Central Asian Monuments (Istanbul, 1992), 40. See also D. Kshibekov, Kochevoe obshchestvo (Alma-Ata, 1984), 71. I am indebted to Professor Peter Golden for this valuable reference. 3  This is its Turkish name. Gözleve, renamed Eupatoria by the Russians at the end of the eighteenth century, was the most important port of the Crimean Khanate. Unfortunately, fiscal data from Gözleve, together with hundreds of other precious manuscript sources, disappeared during the numerous eighteenth-century Russo-Turkish wars. This is why this paper © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_032

The Black Sea and the Slave Trade

Figure 1

959

Sixteenth-century map of the Black Sea. Source: Martinus Broniovius, Tartariae Descriptio (Cologne, 1595).

trade was situated farther inland in Karasubazar. This article discusses the history of the slave trade in the Crimea between the fifteenth and eighteenth centuries, with a special emphasis on the role of the maritime towns (see figure 1).

Slave Trade in Genoese Caffa

The importance of the Black Sea slave trade began to increase in the thirteenth and fourteenth centuries, soon after the establishment of the Genoese colony of Caffa (also known as Kaffa or Capha; the Turkish name was Kefe while in modern Russian it is Theodosia) on the eastern shore of the Crimean peninsula in the 1260s.4 Caffa became the largest and the most important

will discuss mostly Ottoman maritime possessions in the Crimea, since unlike Tatar manuscripts, most Turkish documents were preserved. 4  The exact date of the arrival of the first Genoese settlers in Caffa is still a matter of debate. One of the earliest students of the problem, the Genoese abbot Gasparo Oderico, suggested

960

Kizilov

Crimean port in the late Middle Ages. The Spanish traveller Pero Tafur, who visited Caffa in 1435, wrote that “[t]his city is very large, as large as Seville, or larger, with twice as many inhabitants, Christians and Catholics as well as Greeks, and all the nations of the world.”5 Indeed, Caffa was inhabited by a mosaic of different nationalities, including Genoese, Tatars, Armenians, Greeks, Slavs (Russians and Bulgarians), Syrians, Goths, Alans, Cumans, Caucasians (Georgians, Mingrelians, Circassians, Abkhazians and Lezghins), Hungarians, Persians, Jews, and some other ethnic groups. Accordingly, its urban landscape was also highly varied, with Christian (Greek, Roman, Syrian and Armenian) churches located near Tatar mosques and Jewish synagogues. There is no unanimity of opinion concerning Caffa’s population: medieval sources range from the highly exaggerated number of 21,000 households (perhaps equal to 100,000 souls) at the beginning of the fifteenth century to only 4000 inhabitants in 1467.6 According to rather convincing estimates by modern scholars, there were about 20,000 inhabitants before the fall of Constantinople in 1453 and about 10,000 in the last years before the Ottoman conquest of the city in 1475 (see figure 2).7 The slave trade was an integral part of the city’s economy from the beginning of Genoese settlement. According to Pero Tafur, the slave market of Caffa was frequented by Tatars and residents of surrounding countries who sold not only slaves but also their own children and brothers. The traveller described the process of selling slaves in Genoese Caffa in the following way:

that it was built some time after 1266; see Gasparo Luigi Oderico, Lettere ligustiche, ossia, Osservazioni critiche sullo stato geografico della Liguria fino ai tempi di Ottone il grande, con le memorie storiche di Caffa, Ed altri luoghi della Crimea posseduti un tempo da’ Genovesi, ed. Franseco Massola (Bassano, 1792), 120–122. Michel Balard and other authoritative scholars believe that Caffa was established as a small Genoese trading station in the 1260s and definitely existed in 1270–1275. In 1281 there was a Genoese consul in Caffa; see Michel Balard, La Romanie génoise (XIIe-debut du XVe siècle) (2 vols., Rome, 1978), I, 114–118; and Balard, and Gilles Veinstein, “Continuité ou chagement d’un paysage urbain? Caffa génoise et ottomane,” Le paysage urbain au Moyen Age (Lyon, 1981), 82. 5  Pero Tafur, Travels and Adventures (1435–1439), trans. and ed. Malkolm Letts (London, 1926), 132–133. This work, written in Spanish, was first published as Andances e Vijaes por Diversos Partes del Mondo Audios (1435–1439) (Madrid, 1874). 6  This is according to the Bavarian soldier and Tatar captive Johann Schiltberger, Hans Schiltberger Reisebuch (Tübingen, 1885), 63. Cf. the exaggerated data of the Florentine provided by the merchant Benedetto Dei in 1475, who claimed there were 8000 households and 70,000 inhabitants (Balard and Veinstein, “Continuité,” 83). 7  Balard and Veinstein, “Continuité,” 83.

The Black Sea and the Slave Trade

Figure 2

961

Seventeenth-century Caffa. Source: Guillaume Levasseur de Beauplan, Opys Ukrainy/ Description d’Ukranie (2 vols., Kiev, 1990).

The sellers make the slaves strip to the skin, males as well as females, and they put on them a cloak of felt, and the price is named. Afterwards they throw off their coverings, and make them walk up and down to show whether they have any bodily defect. The seller has to oblige himself, that if a slave dies of the pestilence within sixty days, he will return the price paid.8 The price of slaves rose rapidly after the Black Plague of 1348 which started, according to some accounts, during the Tatar siege of Caffa (see figure 3).9 In 1385–1386 about 530 slaves lived in Caffa; according to their names, most were of Tatar and Circassian origin.10 In these years, c. 1500 slaves were 8  Tafur, Travels and Adventures, 132–133 and 136. The traveller purchased two female and one male slave in Caffa. 9  See the chronicles of Gabrielle de Mussis and Gilles li Muisis in Rosemary Horrox (ed. and trans.), The Black Death (Manchester, 1994), 14–19 and 46–47. When describing the siege of Caffa, the chroniclers depicted the Crimean Tatars as some of the worst infidels, who were punished by God with a terrible decease (the plague) which they later spread to the Christians. 10  For example, Achoga, Cotolboga, Jalavichi, Mogalboga, Jharcasius, Chexica and Torontai (as provided in Latin sources). For more details on the slave trade in Genoese Caffa, see Balard, La Romande génoise, I, 290–302; and Charles Verlinden, “Esclavage et ethnographie sur les bords de la mer Noire,” in Miscellanea Historica in honorem Leonis van der

962

Kizilov

Figure 3

Siege of Caffa by Zaporozhian Cossacks in 1622. Source: Adrian Kashchenko, Opovidannia pro slavne Viisko Zaporoz’ke nizove (Kiev, 1992).

sold annually in Caffa at the prices ranging from 130 aspres for an aged slave to 1000 aspres for a young and strong male. Most were Abkhazian, Tatar, Circassian and Mingrelian males. They were sold in Caffa and usually taken to destinations such as Genoa, Mamluk Egypt (Cairo and Alexandria)

Essen (2 vols., Brussels, 1947), I, 287–298. For general surveys of the Genoese slave trade, see Balard, La Romanie génoise, II, 785–832; S. P. Karpov, “Rabotorgovlya v Yuzhnom Prichernomor’e v pervoi polovine XV veka,” Vizantiiskii Vremennik, XLVI (1986): 139–145; and Paul Meinrad Strassle, Der internazionale Schwarzmeerhandel und Konstantinopel 1261–1484 im Spiegel der sowjetischen Forschung (Bern, 1990).

The Black Sea and the Slave Trade

963

and the East.11 Christians were authorized to purchase and keep slaves by a special papal bull “to prevent their falling into the hands of the Moors and renouncing the faith.”12

Slave Trade in the Ottoman Crimea

After the Ottoman conquest in 1475, the political and administrative structure of the Crimea changed completely. Those regions, ports and towns which were most useful from a mercantile and administrative perspective came under Ottoman jurisdiction, while the rest of the Crimea, along with the southern part of the contemporary Ukraine, the northern Caucasus and Bucak (Budjak), was ruled by the Crimean Khan. In spite of the massacre and the plight of a considerable part of the local Christian populace Caffa’s military and commercial significance did not wane after the Ottoman conquest. It became the centre of Caffa eyalet (Ottoman province) and remained the largest Crimean maritime town, often called by contemporaries “Küçük Istanbul” (“Small Istanbul”).13 Indeed, in the sixteenth century the population of Caffa was as heterogeneous and numerous as during Genoese times: according to Ottoman data, there were more than 3000 households (c. 20,000 souls) in over sixty quarters of the town.14 Despite a lack of 11  Among them also were less numerous Russian, Hungarian and Bulgarian slaves. On Hungarian slaves in Tatar countries, see Lajos Tardy, Sklavenhandel in der Tartarei: die Frage der Mandscharen (Szeged, 1983); and Lajos Tardy, A tatàrorszàgi rabszolgakereskedelem és a magyarok a XIII–XV. szàzadban (Budapest, 1980). In 1420 Mamluk sultan bought 2000 slaves in Caffa; see Balard, La Romanie génoise, I, 302; and Balard, “Genes et la mer Noire (XIII e–XV e siècles),” Revue Historique, CCLXX (1983), 45. At the end of the thirteenth century a thirty-year-old male slave in Caffa was sold for 130 aspres, and Abkhazian girl for 500 aspres and a fourteen-year-old Circassien for 750 aspres (Michel Balard, “Notes sur l’activité maritime des Génois de Caffa à la fin du XIII e siècles,” Société et compaignies de commerce en Orient et dans l’Océan Indien [Paris, 1970], 384). By comparison, wheat was sold in Caffa in 1386 for 175 aspres per muid (hogshead); Balard, La Romanie génoise, I, 765. 12  Tafur, Travels and Adventures, 133. 13  Balard and Veinstein, “Continuité,” 115. 14  There were sixty-five quarters and 3007 (or 3017) households in Caffa about 1520; and sixty-eight quarters and 3067 (or 3043) households in 1542. At this time Muslims comprised less than half the city’s population; see Gilles Veinstein, “La population du sud de la Crimée au debut de la domination ottomane,” in Mémorial Ömer Lutfi Barkan (Paris, 1980), 227–249. Cf. Halil Inalcik, Sources and Studies on the Ottoman Black Sea. Vol. 1: The Customs Register of Caffa, 1487–1490 (Cambridge, MA, 1996), 150. Caffa registers as a source

964

Kizilov

precise data for the seventeenth and eighteenth centuries, travellers’ reports that there were 4000–6000 households suggest a growing population.15 Caffa’s role as the centre of the slave trade expanded considerably after the Ottoman conquest. A detailed description of the slave market in Caffa, and of the tortures and humiliation of the Christian population there, can be found in the work of Michalon Lituanus (Litvin) from the mid-sixteenth century. The author reproduced a long and mournful monologue of one of his PolishLithuanian compatriots who had become a slave.16 Ebülgazi Bahadir Han mentioned about 1664 that Caffa “has no Trade at present, except that of Slaves.”17 Evliya Çelebi mentioned admiringly that a multitude of beautiful male slaves were sold in Caffa in 1666. Travellers also remarked that the slave market was situated near the shoe market.18 According to European sources, in this period the number of slaves sold in Caffa exceeded 30,000.19 Indeed, the number was so large that a mid-sixteenth century Christian visitor wrote with disdain that of information on the slave trade are also analysed in Olexander Halenko, Skhidna Europa u mizhnarodnii rabotogrovli (forthcoming). 15  Reports from travellers, such as Jean Chardin, Julien Bordier, Evliya Çelebi and Guillaume de Beauplan, should be always treated with some scepticism. For a detailed analysis, see Balard and Veinstein, “Continuité,” 95–121. 16  Michalon Lituanus, “De Moribus Tartarorum, Lituanorum et Moschorum, Fragmina X,” in Russia, seu Moscovia, itemque Tartaria (Leiden, 1630), 191; and Michalon Lituanus, Traktat o nravakh tatar, litovtsev i moskovitian, trans. V. I. Matuzova (Moscow, 1994), 73–74. First published in Basel in 1615, this latter work had appeared at the court of the Lithuanian and Polish king Sigismund II Augustus by 1550. Its author concealed his identity under the penname of Michalon Lituanus (Mikhail/Mikhailo Litvin). This valuable work, which was a sort of didactic treatise for the author’s co-inhabitants, needs to be analysed carefully and cautiously. 17  Ebülgazi Bahadir Han, A General History of the Turks, Moguls, and Tatars, Vulgarly Called Tartars. Vol. 2: An Account of the Present State of the Northern Asia (London, 1729), 599. Ebülgazi Bahadir Han (1603–1663) was a Khan of Khorezm; this book, originally written in Jagatay, was completed by his son Enuöse Han in 1644. 18  Evliya Çelebi, Kniga puteshestviya, transl. E. Bakhrevskii (Simferopol’, 1999), 96. The Ottoman traveler Evliya Çelebi (1611–c. 1679), who visited the Crimea in 1665–1666, left precious data on the history, ethnography, linguistics and historical geography of the countries he had seen in his ten-volume travel account entitled Seyahatname (The Book of Travels). For more on his Crimean travels, see Mikhail Kizilov, “Kniga i ee avtor,” afterward to Çelebi, Kniga Puteshestviy, 184–190; and Kizilov, “Kniga Puteshestvii. Evlii Çelebi kak istochnik svedenii o byte Krymskikh tatar,” Materialy po Arkheologii, Istorii i Etnografii Tavriki, III (1993): 273–275. The Crimean portion of Çelebi’s travelogue has been published several times in Turkey and has also been translated into Russian and Polish. 19  Guillaume Levasseur de Beauplan, Description d’Ukranie qui sont plusieurs Provinces du Royaume de Pologne (Rouen, 1660), 32.

The Black Sea and the Slave Trade

965

“because of this practice the city of Caffa may well be called a heathen giant who feeds on our blood.”20 The Ottoman defters (registers) provide us with detailed information on the slave trade in the Caffa eyalet. While the customs dues (4.2 percent) were the same for Muslims and non-Muslims, slaves represented a special case. The earliest register of 1487–1490 informs that the full customs duty on a slave “coming from outside” (der-amadan-i birun) was 210 Ottoman akça.21 If a slave were imported from Azak and Taman, only half of the duty was charged (i.e., 105 akça). If such a slave were later sold in Caffa, the other half had to be paid.22 In the sixteenth century customs dues on slaves in Caffa remained practically unchanged: 210 akça per head (so called “regular tax,” which was collected for the sultan’s treasury); when resold, twenty-seven akça per head (paid by the buyer); and for a slave taken out of the city by his or her owner (möhür verme), 8.25 akça.23 In 1520 the total customs dues on slaves in Caffa (together with two-thirds of the brokerage due) amounted to 690,000 akça; in 1542, they were 157,790 akça.24 In 1529 slave taxes in Caffa amounted to 650,000 akça (c. 10,000 gold pieces), or about one-fourth of total customs revenue.25 A merchant who shipped slaves by sea from Caffa also had to pay an export duty upon leaving the Crimea and another tax in Istanbul if he sold his merchandise there. Kerş, another important port on the Crimea’s eastern shore, also actively traded in slaves, which were delivered from Taman, Circassia, Azak and Caffa.26 20  Cited in Alan Fisher, “Muscovy and the Black Sea Slave Trade,” Canadian-American Slavic Studies, VI, No. 4 (1972), 584. 21   Akça (aqçe) was a silver coin in Turkey and the Crimea. One Ottoman akça was equal (depending on the period) to between two and five Caffa akça. There were old and new Kefevi akça; Inalcik, Customs Register of Caffa, 185–188. 22  Ibid., 57 and 93. 23  Ibid., 145–146. According to Inalcik, this tax was quite high. In other Ottoman regions it was even higher: in Kerpe-Trebizond the duty on a slave (called the pencik, i.e., one-fifth) was 250 akça, whereas the price of a slave fluctuated between twenty-five and fifty golden pieces (fifty-sixty akça = one golden piece). For a comparison, the custom duty on a small boat of Crimean fish or fruit was only twenty-five akça in Caffa in 1542; the value of 100 pieces of exported leather in Caffa in 1487–1490 was 1500 akça, while one sword was 400 akça and fifty batman (1 batman=7.698 kg.) of rice was 1000 akça (ibid., 53–77 and 147). 24  Ibid., 143. 25  Alan Fisher, “The Ottoman Crimea in the Sixteenth Century,” Harvard Ukrainian Studies, V (1981), 139. 26  Kerş (Kerç, Kerc) was the Turkish name; in Italian it was Vosporo or Cerchio; in Greek, Bospor; and Korchev in medieval Russian documents. At present it is Kerch, a port in the Crimean Autonomy (Ukraine). The ports from which the slaves were shipped can be seen in Inalcik, Customs Register of Caffa, 139.

966

Kizilov

Azak (Russian Azov; Venetian Tana), situated advantageously on the Sea of Azov at the mouth of the Don, was another important centre of the slave trade, and many of the captives sold there were later transported and resold in Caffa. In 1529 a full tax for a slave sold in Azak was 256 akça; if a slave was to be taken to Caffa, the tax was only 128 akça.27 Merchants came to Caffa from all over, including Sinop (Anatolia), Sokhum (Georgia), Azak, Mangup (southwestern Ottoman Crimea) and Persia. Slaves purchased in Caffa were later mentioned in the registers of Kili (Kilia), Ak Kerman and Istanbul; some were taken as far as North Africa.28 Although many slaves were delivered to Caffa and other ports by sea, most were sent overland for sale in Tatar and Ottoman ports and towns in the Crimea (see figure 4). From the seventeenth century onwards, due to the blockade of the Dardanelles Istanbul became practically the only destination for the Crimean slave trade. The journey from Caffa to Istanbul normally took about ten days. Once in Istanbul the slaves were sold to merchants from Mamluk Egypt, Anatolia, Iran, Abyssinia and even Western Europe.29 In the eighteenth century the price of a slave on the Crimean market could be from sixty to five or six thousand piastre.30 In both the Ottoman and Tatar parts of the Crimea there were a few minor legal limitations for non-Muslims concerning the purchase of slaves. Normally Jews and Christians were not allowed to purchase Muslim slaves.31 According to de Peyssonel, Christians and Jews were also forbidden to purchase the best slaves—Circassians and Abhazians, who were supposed to be bought only by Muslims.32 Nevertheless, as the Crimean Khan Kırım Giray confessed to the eighteenth-century French diplomat, Baron de Tott, Islam permitted him to give male slaves to Christians but commanded him to keep females in order to convert them.33 Thus, the 27  Ibid., 146. On Tana as an important center of the Italian slave trade see Charles Verlinden, “La colonie vénitienne de Tana, centre de la traite des esclaves au XIVe et au début du XVe siècle,” Studii in onore di Gino Luzzatto, II (1950), 1–25. 28  Ak Kerman (also known as Cetatea Alba or Maurocastron) is the modern Belgorod Dnestrovskii. 29  In 1678 there were Russian and Polish slaves on the galleys of Marseilles; see Fisher, “Muscovy,” 584. 30  Charles de Peyssonel, Traité sur le Commerce de la Mer Noire (2 vols., Paris, 1787), I, 180. 31  Fisher, “Muscovy,” 585. 32  De Peyssonel, Traité sur le Commerce, I, 180. 33  François de Tott, Memoirs of Baron de Tott, Including the State of the Turkish Empire and the Crimea, during the Late War with Russia (2 vols., London, 1785), I, 199. François Baron de Tott (1733–1793) was born in France into the family of an Hungarian noble. In 1755 he was sent to Istanbul and in 1767 to the Crimea. His account, describing his most incredible

Figure 4

Crimean Peninsula. Source: Alexander Vasiliev, The Goths in the Crimea (Cambridge, MA, 1936).

The Black Sea and the Slave Trade

967

968

Kizilov

captives were purchased and used not only by the Turks and Tatars but also by local Christians and Jews (Karaite and Rabbanite alike).34

Slave Trade in the Crimean Khanate

The first major Tatar raid for captives occurred in 1468 and was directed at Galicia.35 The number of raids increased considerably after the Ottoman conquest. According to almost all written sources, the main income of the Crimean Khanate came from raids into adjacent countries and the trade in slaves captured during these campaigns. Since the economy of the Crimean Khanate was not particularly prosperous, plundering neighbouring countries was a way of boosting it; slaves had always been in great demand both within and without the country. As the Dominican friar Giovanni da Lucca remarked around 1630, “there are always slaves to be sold [in the Crimea]. The Turks,

(and perhaps slightly exaggerated) adventures in Turkey and the Crimea was later used extensively by the German writer Rudolf Erich Raspe to create the image of the famous Baron Munchausen (see Baron Munchausen [Rudolf Erich Raspe], Gulliver Revived: Or, The Vice of Lying Properly Exposed (7th ed., London, 1793), 157–168. 34  According to Fisher, “Muscovy,” 584, this business in Istanbul was run exclusively by a guild of 2000 Jewish slave traders. The first reference to the Jewish slave traders of Cherson (Chersonese) dates from 1096; see “Zhitie prepodobnogo ottsa nashego Evstratiya,” in Kievo-pecherskii Paterik (Kiev, 1903), 147–150. Jewish slave owners were also mentioned in the documents of the Massaria of Caffa in 1385–1386; see Balard, La Romanie génoise, I, 302. The Caffa register of 1542 mentions six female slaves (esir) belonging to the members of the Karaite Jewish community of Caffa; see Oleksander Halenko, “Iudeiski hromady Osmanskoyi Kefy seredyny XVI st.,” Skhodoznavstvo, III–IV (1998), 59. In 1614 a Crimean Jew, most likely a Karaite from Çufut-Kale, named Abraham ben Berakhah ransomed the Nogay prisoner Mamay bin Mohammed for 120 florins, which the latter was supposed to re-pay; Library of the Lithuanian Academy of Sciences, Manuscript Department (MS, LMAB), F. 143–1177, folios 3v–4r. A document from 1613 mentions the Russian female slave Servinaz (a corruption of the Russian Serafima?), who was kept in slavery by a Karaite Jewess of Çufut-Kale Malkah, the daughter of Elijah (MS, LMAB, F. 143–1177, folio 1). See more in Mikhail Kizilov, “Jewish Population and the Trade in Slaves and Captives in the Crimean Khanate in the Seventeenth Century,” Proceedings of the Thirteenth World Congress of Jewish Studies (Jerusalem, 2005, forthcoming). 35  Fisher, “Muscovy,” 579.

The Black Sea and the Slave Trade

969

Arabs, Jews, Armenians, and Greeks often buy them … [The Tatars] sell mainly the slaves captured from people against whom they make war.”36 Marcin Broniewski (Broniovius) described how the Crimean Tatars conducted their military raids and captured and tortured prisoners, cunningly forcing their relatives to pay as much for their release as they could afford.37 According to his account, the Crimean Tatars began to prepare for a raid three or four weeks before the proposed date of a campaign. Winter, when all marshes, rivers and muddy roads were frozen, was the usual time for raids. Several skilled and experienced soldiers were sent to capture informers from the country they were going to plunder to gather information. After entering the territory of a rival, they avoided direct confrontations, moving quickly from place to place and trying to seize as many captives and as much booty as possible while plundering and burning everything they encountered.38 The Tatar army consisted mostly of cavalry. Each Tatar had between two and four horses to increase the speed of the army’s movement. Indeed, many travellers remarked on their astonishing speed and mobility. According to Evliya Çelebi (1665–1666), when the Tatar army was headed by the Khan himself, its strength was 80,000 soldiers. But it was only 50,000 when headed by a kalğa sultan (the second in command after the Khan) and 40,000 when headed by a nureddin sultan (the third in command).39 Each Tatar soldier could take a few captives to the Crimea after a successful raid. After the completion of a campaign, the Khan received a tenth of all captives. Thus, in 1769 the Crimean Khan Kırım Giray received about 2000 slaves of the 20,000 captured during raids on southern Russia and the Ukraine.40 36  Giovanni da Lucca, “Fatta da me Fra Giovanni da Lucca Dominicano circa il modo di vivere colle particolarita de costumi delli Tartari Percopiti, Nogai, Circassi, Abbaza etc. Mangrilli e Giorgiani,” in Sebastiano Ciampi (ed.), Bibliogaphia Critica delle Antiche Reciproche Corrispondenze (Firenze, 1834), 55; and Giovanni da Lucca, “Opisanie Perekopskikh i Nogaıskikh Tatar, Cherkesov, Mingrelov i Gruzin,” Zapiski Odesskogo Obshchestva Istorii i Drevnostei, XI (1879), 482. 37   Martinus Broniovius (Marcin Broniewski), Martini Bronovii de Biezdzfedea bis in Tartariam nomine Stephani Primi Poloniae Regis legati Tartariae Descriptio (Cologne, 1595), 17. Martinus Broniovius, an ambassador of Stephan Bathory to the Crimean Khan Mehmed Giray in 1578–1579, described in his travel diary the majority of the towns and castles of the Crimea and the way of life of the Crimean Tatars. 38  Ibid., 19–22. 39   Ibid., 22–23. Cf. Evliya Çelebi, Księga podróży Ewliji Czelebiego, trans. and ed. Z. Abrahamowicz, et al. (Warsaw, 1969). 40  De Tott, Memoirs, I, 196.

970

Kizilov

After the Khan received his share, well-to-do captives were often tortured to induce their relatives to pay ransom. According to some estimates, in the first half of the seventeenth century the number of the captives taken to the Crimea was between 150,000 and 200,000, about 100,000 of whom were captured in the period 1607–1617.41 The Crimean Tatars invaded Slavic lands thirty-eight times between 1654 and 1657; 52,000 people were seized in the spring of 1655 during the incursion into the Ukraine and southern Russia.42 The “overproduction of slaves” in the 1650s, during the years of Chmielnicki’s uprising and afterwards, caused a decline in prices and a dangerous accumulation of rebellious slaves in the Crimea. The problem was solved in Istanbul by the construction of more galleys and by selling excess slaves there.43 The number of Tatar raids seems to have diminished in the eighteenth century due to the growth of Russian strength and a few Russo-Turkish wars, some of which took place in the Crimea. Nevertheless, in 1758 there were around 40,000 slaves captured during the raid to Moldavia.44 And in 1769, during one of the very last Tatar incursions into Russian territory, the amount of “live booty” was about 20,000 souls.45 Most Tatar raids inland were made into adjacent regions and countries, such as Russia, Poland, Moldavia, Georgia, Mingrelia and the Caucasus. On their return they had to traverse the narrow isthmus defended by the fortress of Or (Perekop), where the customs office was situated.46 In the sixteenth century this office was headed by a Jew who asked Michalon Litvin about the number of slaves in Slavic countries.47 Customs had to be paid there first, and later in

41  A. L. Yakobson, Srednevekovyi Krym (Moscow, 1973), 141; A. A. Novoselskii, Bor’ba moskovskogo gosudarstva s tatarami v pervoi polovine XVII veka (Moscow, 1948), 436. Evliya Çelebi, Księga podróży, 355, 362 and 367, provided the impossibly exaggerated numbers of captives in the Crimea in 1666 as from 400,000 to 600,000 and even 920,000. These exaggerated numbers certainly should not be taken literally but as an indication that the number of captives in the Khanate was quite large. 42  G. A. Sanin, Otnosheniia Rossii i Ukrainy s Krymskim Khanstvom v seredine XVII veka (Moscow, 1987), 243. 43  Ibid., 194. 44  According to de Peyssonel, most were later sent back; Charles de Peyssonel, An Appendix to the Memoires of Baron de Tott (London, 1786), 25–26. 45  This is according to an eyewitness, the French diplomat Baron de Tott, who even took part in this incursion; de Tott, Memoirs, I, 189. 46  The Turkish and Tatar names were Ferahkerman (official), as well as Orkapı/Orkapısı or Oragzı; the Slavic names were Perekop or Przekop. 47  Lituanus, “De Moribus Tartarorum,” 189–214; and Lituanus, Traktat, 72–74.

The Black Sea and the Slave Trade

971

Kırk Yer (Chufut-Kale) and Kırk Yeti Yer; dues were collected by special officers called tamğacı and tartnaqcı.48 The fate of the captives taken to the Crimea varied. Some remained in the hands of the Tatars, who used them mostly as domestic and agricultural workers or artisans. According to Marcin Broniewski (1578), the Tatars seldom cultivated the soil themselves, leaving this task to Russian, Hungarian, Walachian and Moldavian slaves.49 The Ottoman traveller, Evliya Çelebi, mentioned that during the absence of the men, Tatar women took their men’s sabres to control the captives’ work in the fields.50 The most important and valuable captives were usually imprisoned in the mountainous fortress of Çufut-Kale, not far from the capital of the Khanate, Bahçeseray. Among the “reluctant travellers” imprisoned in ÇufutKale were such famous persons as the Prince of Transylvania, Janos Kemeny (1657); the Poles Potocki and Kalinowski (1648); the Russian Vasili Sheremetev (1660–1681); and Russian ambassadors V. Aytemirov (1692–1695) and A. Romodanovski (1681).51 Some captives were delivered to the Crimean Khanate’s western port of Gözleve (Eupatoria). According to Evliya Çelebi, the slave market of Gözleve was situated close to the main gates of the port; hundreds of beautiful young men and women were sold there each day. Some slaves were taken eastward to the towns of Eski Kırım (Italian: Solkhat/Surhat; Russian: Staryi Krym) and Karasubazar (modern Belogorsk), situated close to the border with the Ottoman Crimea. Çelebi sorrowfully wrote about what he considered to be the “shameful” slave market in Karasubazar: “A man who had not seen this market, had not seen anything in this world. A mother is severed from her son and daughter there, a son—from his father and brother, and they are sold amongst lamentations, cries of help, weeping and sorrow.”52 He also added a few interesting details about the skills of the local slave traders who could deceive the

48  This is according to the soyurğal (grant) of Hacı Giray of 1453; see Alexander Bennigsen, Pertev Naili Boratav, Dilek Desaive and Chantal Lemercier-Quelquejay, Le Khanat de Crimée dans les Archives du Musée du Palais de Topkapı (Paris, 1978), 33–38. In 1500 three merchants from Moscow entered the Crimea via Orkapı (Perekop) and paid a tamğa (tax) of seven percent ad valorem and three gold pieces to the head of the clan of Shirins. Had they done this via Caffa, no tamğa would have been paid; Inalcik, Customs Register of Caffa, 98–99 and 102–103. 49  Broniovius, Tartariae Descriptio, 17. Cf. Inalcik, Customs Register of Caffa, 134. 50  Çelebi, Księga podróży, 362. 51  Mikhail Kizilov, Karaites through the Travelers’ Eyes (New York, 2003), 190–191. 52  Çelebi, Księga podróży, 308.

972

Kizilov

buyers and even sell old and exhausted slaves at high prices.53 In Caffa, if a female slave was adorned with make-up (allamak pullamak), an additional embellishment tax (’adet-i tezyin) of twelve Ottoman akça was levied.54 To avoid too many cases of cheating customers, however, a special edict was issued in 1680 forbidding merchants from “colouring the faces of slave women or embellishing them in any way.”55

Selling and Ransoming the Captives

The captives were not always sold; often they were exchanged for other goods. Pierre Chevalier in 1663 wrote that the Tatars often accepted Turkish horses, weapons, clothes and other merchandise in payment.56 The price asked for the captives varied over time, depending on the number of captives caught and the quality, age, strength, beauty, skills and gender of the slaves. More respected and important captives (wealthy merchants, military commanders, nobles and the like) could cost a fortune: for the Russian boyar Vasilii Sheremetev, for example, the Crimean Khan demanded from the Tsar not less than Kazan and Astrakhan, demands which were certainly too high even for such an important person.57 As a consequence, he spent twenty-one years in captivity in ÇufutKale and died in Russia a few months after his release in 1681. Eunuchs were also expensive, costing at least 1000 Reichstaler in the seventeenth century.58 Both female and male slaves were often used for sexual purposes.59 After a successful military campaign, when each Tatar soldier might bring back between ten and twenty captives (some sources say up to thirty or forty), prices decreased rapidly. The Crimean Tatar chronicler Hacı Mehmed Senai 53  He added, however, that these traders were not Tatars but Turkish merchants from Anatolia; ibid., 306. 54  Inalcik, Customs Register of Caffa, 147 and 149. 55  Fisher, “Muscovy,” 584. 56  Pierre Chevalier, Histoire de la Guerre des Cosaques contre la Pologne (Paris, 1663), 49–83; and Chevalier, Istoriya viiny kozakiv proty Polshchi, trans. Y. I. Nazarenko (Kiev, 1993), 60–61. 57  L. P. Kruzhko, “Tsennyı plennik krymskikh khanov,” Krymskii pushkinskii nauchnyi sbornik, I (2001), 286. 58  J. Matuz, “Eine Beschreibung des Khanats der Krim aus dem Jahre 1669,” Acta Orientalia Societatis Orientalis Danica Norvegica Svecica, XXVIII, No. 2 (1964), 140–142. 59  Numerous cases of the use of male slaves-gulams for homosexual purposes were described by Çelebi, Kniga puteshestviya, 34 and 95–96.

The Black Sea and the Slave Trade

973

mentioned that after the successful raid into Poland and Ukraine during Chmelnicki’s Rebellion (1648), each Tatar soldier killed ten or fifteen captives for his amusement.60 According to Evliya Çelebi, in the 1640s and 1650s the Crimean Khan Islam Giray III invaded Polish lands seventy-one times and captured 200,000 Jews, later selling each for the price of a full tobacco pipe.61 Despite the fact that such information was probably an exaggeration, it gives some idea of the number of captives taken during such raids and the attitude towards them. It was often the case that the captives were sold by the Tatars not to their fellow Muslims from the Crimean Khanate but to wealthy Christian or Jewish merchants, who could retain them for their own purposes or try to sell them back to the countries of origin. Especially interesting is the information of Marcin Broniewski in the chapter “Captivorum apud Tartaros ratio” of his Tartariae Descriptio (1578). He mentioned that ambassadors from Christian countries often tried to bribe the Jews or Tatars in order to ransom Christian captives for a price lower than would be demanded by Tatar officials. In his opinion, these bribed Jews and Tatars were extremely important for the successful ransoming of the captives.62 Travellers’ information about the Jewish mediators in ransoming slaves is confirmed by other sources. A merchant from Caffa, Hoca Bikeş Gökgöz (in Russian sources Hozia Kokos), perhaps the most famous medieval Crimean Jew, had special dealings with the Russian Tsar Ivan III over the redemption of Russian prisoners in the 1470s. In spite of the fact that the Russian merchants who had been captured by the Şirin bey Mamaq were very grateful to Hoca and even gave him some money, the Jewish merchant tried to get some additional money by cheating the Tsar. Regardless, he certainly played a crucial role in the release of these captives.63 60  He was also known as Kırımlı (“of the Crimea”). Hacı Mehmed Senai, Historia chana Islam Gereja III, trans. and ed. Z. Abrahamowicz (Warsaw, 1971), 64. Cf. also the recent Russian translation: Hacı Mehmed Senai Kırımlı, Kniga pokhodov. Istoriia khana Islam Giraya III (Simferopol, 1998). 61  Çelebi, Seyahatnamesi, 527, as quoted in Senai, Historia, 64. In 1666, to avoid additional taxation on the slaves introduced by Adil Çoban Giray, local slave-owners were selling three healthy slaves for one golden piece (Çelebi, Księga podróży, 361). 62  Broniovius, Tartariae Descriptio, 21–22. 63  Hoca Bikiş/Bikeş, a Jew, son of Gökgöz, an influential merchant in Kefe (also called “Bikiş son of Gökgöz”), was mentioned in the Caffa register of 1497 (Inalcik, Customs Register of Caffa, 74). Previous scholars could not identify the proper name of this interesting person and called him according to a corrupt Russian spelling “Hozia Kokos.” Cf. VI. Ogorodnikov,

974

Kizilov

Highly interesting is the destiny of the Jewish merchant Meir Ashkenazi (d. in the second half of the sixteenth century), who was born in Poland, lived for some time in Ottoman Caffa, and was appointed Tatar ambassador to Poland about 1567. He often travelled to such remote countries as Egypt and India, and was killed by corsairs near Genoa while transporting prisoners from Egypt.64 Polish archival documents mention also the active role of the Armenian merchants as mediators in the ransoming process.65 Slavic captives in Turkey or the Crimea occasionally were exchanged for Tatar captives—as was Bohdan Chmielnicki (Khmel’nyts’kyi), who was released from captivity in Istanbul (1620–1622/1623) in exchange for Tatar captives taken in the Battle of Chocim.66 The problem of ransoming Russian slaves from Tatar captivity became an important issue in sixteenth- and seventeenth-century Russia. In 1551 the Zemskii sobor (parliament) regularized the ransom of Russian captives from the Crimea, while Tsar Ivan IV encouraged Crimean merchants to come to Moscow to sell them. In 1649 Alexei Mikhailovich established a special annual tax for this purpose, and between 1667 and 1672 a special Polonianichnyi prikaz (department for captives) was organized.67 The Polish government, in its turn, paid a high yearly tribute to the Tatars to prevent raids and the taking of captives.68 In Moscow the Crimean Tatar merchants could not only sell their captives but also buy new ones, mostly of Lithuanian, Polish and German origin, from the Russian army.69 “Ivan III i zarubezhnye evrei,” Sbornik statei v chest’ Dmitriia Alexandrovicha Korsakova (Kazan’, 1913), 57–63; Regesty i nadpisi. Svod materialov dlya istorii evreev v Rossii (2 vols., St. Petersburg, 1899), I, 77–79; and Yulii Gessen, Istoriya evreiskogo naroda v Rosii (2 vols., Petrograd, 1916), I, 23–24. 64  Maurycy Horn, “Udział Żydćw w kontaktach dyplomatycznych i handlowych Polski i Litwy z zagranicą w XV–XVII w.,” Biuletyn Żydowskiego Instytutu Historycznego, III–IV (1990), 7. Cf. I. V. Achkinazi, Krymchaki (Simferopol’, 2000), 66. 65  Archiwum Glówne Akt Dawnych: Archiwum Koronne Warszawskie, Dział Tatarskie (AGAD, AKW, Dz. Tatarskie), k. 61, t. 69, no. 211; and k. 61, t. 62, no. 204. 66  There is another version that he might have been ransomed by his mother; see commentaries to Senai, Historia, 75. 67  Alan Fisher, The Crimean Tatars (Stanford, 1978), 28. 68  The whole sum paid as “Tatar gifts” for the period from 1663–1667 was as large as 176,310 złotys; it was financed mostly by the Jewish poll-tax; Zbigniew Wójcik, “Some Problems of Polish-Tatar Relations in the Seventeenth Century,” Acta Poloniae Historica, XIII (1966), 100–101. 69  S. M. Solov’ev, Istoriya Rossii s drevneishikh vremen, IV, parts 7–8 (Moscow, 1960), 139.

The Black Sea and the Slave Trade



975

Position and Everyday Conditions of Slaves

In the Islamic legal system slaves were called mal-i natik (“speaking property”), a category somewhat similar to the Latin instrumenti vocali. In the Turkic languages slaves were usually called esir (hence the Russian yasyr’ and the Polish jasyr). Despite the fact that the Koran dictated the humane treatment of slaves, in the Crimea and Turkey this principle was not always followed. Indeed, sixteenth-century Christian sources are full of descriptions of the terrible sufferings of European slaves in Crimean captivity.70 Some of them could spend the rest of their days doing some exhausting work. As the Crimean vizir Sefer Gazi Ağa mentioned in one of his letters, the slaves were often “a plough and a scythe” for their owners.71 Perhaps the most terrible fate was to become a galley-slave.72 Each galley needed approximately 150–200 oarsmen, and there were about 45,000–60,000 galley-slaves in the Ottoman navy in the second half of the seventeenth century.73 Subsequently, the Turkish word kadırga (galley) in the form katorga became a synonym for “prison” in Russian. Many of the slaves were allowed to marry, but their offspring usually became slaves themselves.74 The position and day-to-day conditions of a slave depended largely on his/her owner. Some masters took care not to exhaust their slaves and to prolong their productive lives, while others were entirely indifferent to the fate of their chattels. Driven to desperation slaves could easily turn against their masters, as was the case with the eighteenth-century

70   Blaise de Vigénére, “Izvlechenie iz zapisok Blaise de Vigénére,” in Memuary otnosiashchiesia k istorii Yuzhnoi Rusi (Kiev, 1890), 81–82, mentioned that small and old slaves were given to Tatar boys to be tortured to death. Michalon Lituanus, Traktat, 72, mentioned that the strongest of the Christian slaves were castrated, tortured and fed rotten meat. It seems, however, that many of these sources were heavily distorted by the political and didactic agenda of the time. 71  AGAD, AKW, dz. Tatarskie, k. 61, t. 135, no. 277 (1661). The Tatar collection of AGAD contains hundreds of other interesting documents related to the fate of Polish captives in Tatar and Turkish captivity. 72  Their sufferings were poeticized in many Ukrainian dumas (songs); Adrian Kashchenko, Opovidannia pro slavne Vijsko Zaporoz’ke nizove (Kiev, 1992), 29–33. 73  Sanin, Otnosheniia, 194. 74  An English captive known to us as “Astrakan” received a Wallachian woman to be his wife. Astrakan later took her with him after their release from Crimean captivity; Astrakan, “A Short Description of All the Kingdoms which Encompass the Euxine and Caspian Seas,” in J. B. Tavernier, A New Relation of the Inner Part of the Grand Seignors (London, 1677), 119.

976

Kizilov

Crimean Jewish slave-owner who was killed by his slave in his own vineyard. By order of the Khan Maksud Giray the murderer was delivered into the hands of the Jewish community which, according to Old Testament prescription, stoned the culprit to death.75 As with other forms of property, slaves were transferred from a deceased owner to his or her heirs. Some slaves could also earn their freedom, and many preferred to stay in the Crimea instead of going back to the land of their birth.76 Especially significant is the story of the Cossack Ivan Sirko, who freed about seven thousand imprisoned Slavic slaves during the Crimean raid of 1675. On their way from the Crimea he suggested that some might choose to return—and many decided to do this, saying that “in the Crimea they had household and possessions.” Enraged by their wish to go back to their Muslim lords, Sirko massacred all three thousand of those who decided to return.77 It seems that the captives were usually allowed to retain, and even to practise their non-Muslim religions.78 According to Muslim doctrine, it was considered more desirable to convert female than male slaves.79 Those who converted to Islam usually received new Muslim names, were treated as Muslims and could

75  It is very likely that this event happened in the vicinity of Bahçesaray (de Tott, Memoirs, I, 95–96). Draco, a rich Greek merchant from Istanbul, was hanged after one of his female slaves declared that she was a Turk and that Draco had “ill treated her, to oblige her to turn Christian” (ibid., I, 196–198). 76  Sigismund Herberstein (1549), in his Rerum Moscoviticarum commentarii, mentioned that there was a custom in the Crimea to release slaves every six years; Siegmund von Herberstein, Rerum Moscoviticarum Commentarii (Vienna, 1549). Cf. Senai, Historia, 64. This paradoxically reminds us of the Jewish tradition to release slaves every seventh year. Michalon Lituanus also mentioned that slaves were kept in the Crimea for only seven years and explained this tradition by reference to the Biblical prescription. These data seem to contradict numerous testimonies about life-long captivity. Were the testimonies of both travellers just Biblical allusions? 77  D. I. Yavornitskii, Istoriya Zaporozhskikh kozakov (2 vols., Kiev, 1990), II, 387–388. For a few other such cases, see Fisher, “Muscovy,” 592. 78  See the letter of the Polish captives from Bahçeseray containing complaints over the improper behaviour of a new priest sent from Poland to the Crimea to meet the religious needs of Catholic captives; AGAD AKW, dz. Tatarskie, t. 13, no. 13 (1660). 79  De Tott, Memoirs, I, 199.

The Black Sea and the Slave Trade

977

not be ransomed.80 The Tatar slaves in Europe, on the contrary, were usually forced to convert to Christianity.81

Escape from Bondage

Many “reluctant travellers” who were captured during the Tatar raids tried to escape from captivity. This was a difficult undertaking, since they had to undertake a journey to the nearest Christian country without food, money or proper dress, usually on foot. Because of the excessive number of slaves taken during the 1650s there was an attempt to organize a large uprising against the slaveowners, but the plot was uncovered.82 The most typical period to escape was during the harvest, when the slaves could hide in the crops, which they also consumed. They also often received support in the form of money, food and clothes from the local Christian population, especially the Greeks and Armenians.83 From the Crimea the captives often travelled via the ford near Sıvaş (Russian: Gniloe more or “putrid sea”), called in the sources Çingişke or “Tonka Voda” (“narrow water”).84 Few were successful, although we do possess data about a few travellers who managed to escape. The English captive “Astrakan” managed to escape to the Zaporozhian Cossacks via the shallow waters of Çingişke in the first half of the seventeenth century; later, however, he was caught and enslaved anew in the Crimea.85 Those slaves who were caught 80  See the Polish-Turkish peace treaties of 1502 and 1514 in the commentaries to Senai, Historia, 76. Sources describing the alleged conversion of Bohdan Chmelnicki to Islam (who was nevertheless ransomed in 1622/1623) should be dismissed as unreliable (ibid., 74–76). In light of these data we should also doubt the veracity of hypotheses suggesting that Vassili Sheremetev converted to Islam to be eligible for the post of the Khan’s translator; M. N. Berezhkov, “Russkie plenniki i nevol’niki v Krymu,” Arkheologicheskii s’ezd. Trudy, VI, No. 2 (1888), 361. Cf. Alan Fisher, The Russian Annexation of the Crimea 1772–1783 (Cambridge, 1970), 20–21. 81  Marian Małowist, Caffa-kolonia genueńska na Krymie i problem wschodni w latach 1453– 1475 (Warsaw, 1947), 84. 82  Sanin, Otnosheniya, 195. 83  See the chapter “On the escape of the captives from Turkey” in Alexander Gwagnini [Alessandro Guagnini], Kronika Sarmacyey Europskiey [sic], books 9–10 (Krakow, 1611), 79–80. 84  Also known as Yeniçi and Cencige; this is the modern town of Genichesk in the Ukraine. 85  Astrakan, “Short Description,” 118.

978

Kizilov

trying to escape were severely punished and often blinded.86 To prevent their flight, the captives were usually chained together and had to spend the night in a deep pit covered with a lid on which their guards slept.87 Even those who managed to return to their native country found that it was often not the end of the ordeal: many found their wives married to others or their houses and property lost or confiscated. Many who returned still had to pay back their ransom to the Tatars. But those who acquired a command of Turkish or Tatar could often find good positions as interpreters at the court.88

Origins of Slaves

In addition to the usual terms to denote slaves, there were a few others which can help us to establish their ethnic origins. According to these terms, the slaves were mostly Slavs, most likely from Poland, Russia and the Ukraine. Evliya Çelebi mentioned that in the Crimean Khanate female slaves were usually called difka (most likely a corruption of the Russian devka or the Ukrainian divka, meaning a girl) and maria.89 The former term was usually used to denote young women and virgins, while the latter was applied to adult women. For the male captives the terms kopna or konia were used (kopna is likely a corruption of the Polish word chłop, meaning a peasant or a young man).90 The register of the Jewish community of Caffa of 1542 mentions six female slaves whose names were indicated as “marya rus” or “marya;” one of them had a typical Russian name Ulyana.91 Especially worth mentioning are the prisoners of war: the Zaporozhian Cossacks, undaunted pirate-brigands from the southern Ukraine whose frequent maritime and inland raids devastated many ports and towns of the Crimea and Turkey in the sixteenth and seventeenth centuries.92 Small Cossack 86  In the Ukraine many such blinded captives became the so-called kobzars, or blind travelling singers; see Kashchenko, Opovidannya, 33. 87  Sanin, Otnosheniya, 195–196. 88  See a few cases of this kind in Fisher, “Muscovy,” 590. 89  This most probably was because “Maria” was one of the most widespread names of these captives; Çelebi, Księga podróży, 355. 90  Ibid., 355, 362 and 367. 91  Halenko, “Iudejski gromady,” 59. 92  Their raids in the Crimea started as early as in the second half of the sixteenth century; Minhea Berindei, “Le problème des ‘Cosaques’ dans la seconde moitié du XVIe siècle,” Cahiers du Monde russe et soviétique, XIII, No. 3 (1972): 338–367. Alan Fisher, “The Ottoman

The Black Sea and the Slave Trade

979

vessels called chaikas (hence the Russian shaika, meaning a group of bandits), suddenly appeared in the most unexpected places and were a constant threat to the vessels and ports of the Black Sea. Tatar and Turkish captivity (nevolia) was a recurrent topic in many dumas—Cossack songs of the time. Even Bohdan Chmielnicki, a key figure in the separation of the Ukraine from Poland during the Cossack rebellion of 1648, was imprisoned in Istanbul for about two years.93 Archival sources mention the curious case of a Karaite Jew, Samuel ben Daniel, who received a certain kazak (perhaps a Zaporozhian Cossack captured by the Tatars) to be kept in his place. During the night, however, the Cossack managed to escape.94 “Kazak Istefan” (a corruption of Stepan) is mentioned as a slave in the Crimean maritime settlement of Sikita (Nikita) about 1520; his owner was his religious brethren, a Greek named Dimitrios.95 Not only Zaporozhian but also Don Cossacks attacked the Muslim Crimea and were often caught and enslaved. Especially interesting is the story of the Don Cossack Ivan (Ivashko) Vergunenok who, when caught by the Tatars about 1460 claimed to be a son of Prince Dimitri, the heir to the Russian throne. He had been sold to a Jewish merchant in Caffa and was later bought by the Crimean Khan, who apparently wanted to use him in the planned war with Russia. The Khan kept him in chains in Zhidovskii Gorodok (Russian for CufutKale) for three years and in 1646 sold him to the Turkish sultan.96 Quite numerous among the slaves as well were Caucasian (Circassian, Abkhazian, Georgian and Lezghin) and Kalmykian captives. According to de Peyssonel, Circassian women, who were renowned for their beauty, were often taken by the Crimean khans as concubines; as a consequence, many of the Crimean princes were of Circassian origin. De Peyssonel also remarked that Crimea in the Mid-Seventeenth Century,” Harvard Ukrainian Studies, IV (1979–1980), 216 and 221, supposed that one of the main reasons for the worsening demographic and economic situation in the seventeenth-century Crimea was the frequent devastating raids of the Don and Zaporozhian Cossacks. For an analysis of Cossack naval raids on the Crimea and Turkey, see E. V. Venikeev and L. T. Artemenko, Peniteli Ponta. Piratstvo v Chernom more (Simferopol’, 1992), 72–129; and Victor Ostapchuk, “The Human Landscape of the Ottoman Black Sea in the Face of the Cossack Naval Raids,” Oriente Moderno, XX (2001). 93  On his captivity, see the commentaries to Senai, Historia, 72–77. 94  MS, LMAB, F. 143–1177, folio 4. 95  Olexander Halenko, “Sikita/Nikita osman’skoho chasu,” Istoricheskoe nasledie Kryma, VIII (2004), 84 and 91. 96  See the Russian archival documents published in Solov’ev, Istoriya, V, parts 9–10, 462–463 and 466. Alan Fisher’s reading of these documents is erroneous (cf. Fisher, “Muscovy,” 587).

980

Kizilov

the beautiful Circassians were also favourite concubines of Turkish sultans.97 A Ruthenian captive, Alexandra-Roxolana, the favourite wife of Süleyman the Magnificent, was perhaps one of the most influential women in the history of the Ottoman Empire.98 Ebülgazi Bahadir Han mentioned that the Crimean Tatars in general preferred female slaves for wives because of the rather plain appearance of Tatar women.99 In addition to Slavic and Caucasian captives, there were thousands of slaves from other European countries, including Hungary, Moldavia (Wallachia), Germany and even England and Scotland. Among those captured during raids into Slavic lands were also other ethnic inhabitants, including Jews, Greeks and Armenians.100 There certainly were numerous examples of slavery working in the opposite direction, i.e., when the Tatars were seized by Russians or Poles during military campaigns.101 In 1453, for example, the price of a young Tatar on the slave market in Genoa was about sixty to eighty liras.102 According to Pero Tafur, in the fifteenth century slaves of Tatar origin were three times more expensive than others because they had a reputation of being absolutely faithful to their master.103 The marvellous fidelity of captive Tatars in Poland was also mentioned by Pierre Chevalier, who remarked that the Polish nobility trusted their Tatar servants with the keys to their money and jewels.104 Considerable numbers of Tatar captives were seized by the Zaporozhian Cossacks. For example, the Cossack ataman Ivan Sirko captured 6000 Tatars during his raid on the Crimea in 1675.105 97  De Peyssonel, Traité sur le Commerce, I, 177; and de Peyssonel, An Appendix, 22. 98  Fisher, “Muscovy,” 586. 99  Han, General History, 602. 100  In 1650, during Chmielnicki’s uprising, the Tatars captured 300 Polish Armenians and carried them away to the Crimea, where most were redeemed by local Christians. See the chronicle by the seventeenth-century Armenian historian Xačatur Kafajeci (of Caffa) in E. Schütz, “Eine armenische Chronik von Kaffa aus der ersten Hälfte des 17. Jahrhunderts,” Acta Orientalia Academiae Scientiarum Hungaricae, XXIX (1975), 159. 101  See a few fifteenth-century cases of this kind in Bertold Spuler, Die Goldene Horde (Wiesbaden, 1952), 386. 102  Malowist, Caffa, 84. 103  Tafur, Travels and Adventures, 133. 104  Pierre Chevalier, A Discourse of the Original, Country, Manners, Government and Religion of the Cossacks (London, 1672), 45–46 and 54. The traveller saw a few hundred captive Tatars in Warsaw in the 1660s. 105  Yavornitskii, Istoriya, 387–388.

The Black Sea and the Slave Trade

981

Despite the fact that the most common way of obtaining prisoners was through military raids, there were frequent cases when family members sold their own relatives (especially children) during times of famine or in order to avoid payment of the tribute.106 Sometimes the captives were donated to important persons as a symbolic embodiment of the humiliation of a defeated enemy.107 Some were violently executed as an example of the fate of the enemies of the country.108 In general, however, slaves were seldom killed; the usual punishment for misbehaviour was to sell them to the owner of a galley. Conclusion Maritime towns in the Crimea were places where merchants from all over the Black Sea region, the eastern and western Mediterranean, Anatolia, Turkey, Russia and even Western European countries came to buy, sell, and exchange their goods. In this trade “live merchandise,” or “reluctant travellers,” caught by the Tatars during raids on adjacent countries were among the most important trade goods. Indeed, the Crimea was one of the main centres of the slave trade in the eastern Mediterranean and the Black Sea, with the slaves sold in a few important ports (especially Caffa) and in the hinterland (especially Karasubazar, or modern-day Belogorsk). Before 1475 the role of slave traders was fulfilled by Tatar and Genoese merchants, but after the Ottoman conquest it was carried out mostly by the Muslim Tatars and Turks, with a significant role also played by local Jewish and Armenian traders. Selling slaves and captives back to their relatives in exchange for a ransom became an important offshoot of the slave trade. Most of the slaves captured by the Tatars during raids on Poland, Russia and the Caucasus were delivered to the Crimea by land. Nevertheless, it was mostly maritime towns (Caffa, Azak, Taman, Kerş and Gözleve) where the slaves were later delivered. The slave trade was a cornerstone of the Crimean economy in the early modern period, so much so that a decrease in slave prices signified a worsening economic condition. Thousands of slaves—a number too great for the needs of the Crimea’s inhabitants—were sold annually in the slave markets of the Ottoman Crimea and the Crimean 106  Spuler, Die Goldene Horde, 385–386; cf. Tafur, Travels and Adventures, 132–134. 107  Spuler, Die Goldene Horde, 386. 108  During the reign of Mohammed Giray IV (1641–1644 and 1654–1666) a few captured Cossacks were taken to the Crimea and killed “with the most violent punishment;” Ostapchuk, “Human Landscape,” 88.

982

Kizilov

Khanate. It is due to the existence of an international slave trade carried out through the maritime towns of the region that there was a constant demand for new lots of slaves. Consequently, it was a constant stimulus for the Tatars to continue their predatory raids. The eighteenth century was marked by a growing concern in Russia to stop the Tatar raids, which annually took tens of thousands of its inhabitants. Moreover, Russian interests in this area were also stimulated by the imperial desire to expand Russian territory further south and to gain control of a long-cherished outlet to the Black Sea. In 1774–1783 the Russian annexation of the peninsula stopped forever this inhumane trade in people. Caffa (renamed by the Russian victors as Theodosia), the largest port and slave market of the Crimea, lost its importance immediately after the Russian annexation of the peninsula. This was followed by the massive emigration of the Muslim inhabitants and the abolition of the slave trade. The economic and commercial centre of the region shifted to the western part of the peninsula where construction of the port of Sevastopol began on the site of the insignificant Tatar village of Aq Yar. Caffa (Theodosia), Azak (Azov), Gözleve (Eupatoria) and Kerş (Kerch), former centres of the slave trade, were converted into Russia’s largest commercial and military ports in the south of the country. The slave trade in the Crimea was never resurrected.

The Gypsies in the Romanian Lands during the Middle Ages: Slavery Viorel Achim 1

The Age and Origin of Slavery in the Romanian Lands

From the first attestations of their presence in Wallachia and Moldavia, the Gypsies were slaves. They were to remain in this social condition for many centuries until the laws abolishing slavery in the middle of the nineteenth century. The Gypsies were also enslaved in Transylvania, more particularly in the regions that were for a time under the control of the Wallachian and Moldavian princes. Even after the end of the dominion of the two Romanian states there, the Gypsies remained for a time as slaves, a vestige of that previous era.1 The origins of slavery in the principalities have not formed the basis of an independent study in Romanian historiography. Generally speaking, works of social history, which inevitably make reference to the slavery of the Gypsies, content themselves with the statement that slavery dates from before the creation of the principalities and that its origin is unknown. Since the appearance of the Gypsies in the Romanian territories was linked with the Tatars, who ruled there after 1241, the slavery of the Gypsies has been regarded as a vestige of that era. In the view of Nicolae Iorga, the Gypsies who were the slaves of the Tatars were taken over by the Romanians preserving their state of bondage.2 Other writers explain the appearance of slavery in the Romanian territories with reference to the Romanians’ battles with the Tatars. As a result of these battles, Gypsies captured from the Tatars were transformed into slaves.3 The problem of slavery in the medieval history of Romania is not restricted to the slavery of the Gypsies. Alongside the Gypsy slaves, in the Romanian states there were also Tatar slaves. Moldavian documents of the fifteenth century tell us about this second category of slaves. These Tatars were mistakenly S ource: Achim, Viorel, “The Gypsies in the Romanian Lands during the Middle Ages: Slavery,” in Viorel Achim, The Roma in Romanian History, Budapest: Central European University Press, 1998, 27–85. 1  See below pp. 42–43. 2  N. Iorga, op. cit., p. 22. 3  Al. Gonţa, op. cit., p. 313ff. © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_033

984

Achim

identified as Gypsies and so the theory of the arrival of the Gypsies on Romanian territory in the thirteenth century during the Tatar domination was constructed on the basis of this identification. It is certain that the Tatar slaves mentioned in the Moldavian documents were a legacy of the Tatars. They were Tatars (or rather slaves of the Tatars) who had ended up in the possession of the Romanians. We find it highly plausible that at the time of the conflicts between the Romanians and the Tatars during the fourteenth century, the Romanians would have transformed Tatar prisoners of war into slaves. Having spent a long period of time under Mongolian domination, we can hypothesise that the Romanians could have adopted from the latter the practice of enslaving captured enemies. It is not, however, obligatory to regard these prisoners of war as Tatars from an ethnic point of view. As has been supposed, it is highly probable that the Tatar slaves in Moldavia were in fact a population of Cumans, established in the region before the arrival of the Tatars.4 The Romanians took them over as slaves and kept them on in this state. The question that needs to be asked is: when did they come into the possession of the Romanians? For how long had there been Tatar slaves? Since the founding of the principality in the middle of the fourteenth century? Or from a later time, when the Moldavian state had incorporated the territory to the south where Tatar domination had continued until the last decade of the fourteenth century? It is difficult to give an answer to these questions in the absence of any definite historical evidence. To historians, however, it is certain that there were also Tatar slaves in Wallachia, although they are not mentioned in official documents. The toponyms “Tătărăi” in Muntenia (identical to the toponyms “Tătăraşji” in Moldavia) appear to indicate this very population of slaves acquired from the Tatars. We believe that we may assume that “Tatar” slaves existed in Wallachia and Moldavia from the foundation of these principalities, which took place at the beginning and the middle of the fourteenth century respectively. Once passed under this new dominion, Tatar prisoners of war or Tatars’ slaves entered the service of the prince as slaves. This is not, however, an instance of a phenomenon peculiar to Romanian history. In Eastern Europe, the turning of pagan enemies into slaves was practised in the first centuries of the second millennium. It is known that prior to their Christianisation, the Hungarians would turn prisoners of war into slaves. In the Hungarian Kingdom, Muslim Saracens and Mozaic Khazars were used as slaves until the thirteenth century, when they were forced to convert to Christianity.5 In fourteenth century Hungary, however, the institution of 4  N. Beldiceanu, Irène Beldiceanu-Steinherr, op. cit., p. 13. 5  See Al. Gonţa, op. cit., pp. 307–311, with the references in the notes.

The Gypsies in the Romanian Lands During the Middle Ages

985

slavery no longer existed. The Eastern Slavs established slavery for those that they captured in battle.6 Prisoners taken by the Russian dukes from the Tatars were considered the duke’s slaves and as a rule used to populate certain border areas. The institution of holop continued for a long time, albeit in an increasingly watered down form, until the distinction between slave and subjugated peasant disappeared altogether in the seventeenth and eighteenth centuries.7 In any case, the practice of keeping domestic slaves existed almost everywhere in Europe in the early Middle Ages. The Romanians adopted slavery from the social system of the Tatars, from whom they also adopted certain institutions and elements of military, administrative and fiscal organisation. Slavery existed on the Romanian territories even before the creation of the Romanian states, since the Tatars themselves had slaves. It is evident that the origins of slavery in the Romanian lands have nothing to do with the appearance of the Gypsies there. They are two distinct questions. When the Gypsies reached the area north of the Danube at the end of the fourteenth century, slavery had already been in existence there for some time. The newcomers, foreign to the local society in every respect and with a nomadic way of life, were assured the same regime as that of the Tatars. The role of the Gypsies in the history of slavery in Romania lies in the fact that due to the relatively large number of Gypsies settled in the Romanian lands in the fourteenth and fifteenth centuries, slavery became a widespread phenomenon. The Gypsies managed to acquire the monopoly in this social institution, as by the second half of the fifteenth century they were the only slaves in the country. The Tatars, who had been present in small numbers, had disappeared, merging into the mass of Gypsy slaves. In this way the term “Gypsy” became synonymous with that of “slave”. Another question is whether the Gypsies arrived in the Romanian lands as slaves or as freemen. In order to provide a response, we need to bear in mind the social conditions of the time in the countries of South-Eastern Europe. Slavery in its medieval form was a conspicuous reality in the Byzantine Empire until later on.8 In these conditions, the Gypsies paid a special tax and were recorded in a special register. This system of taxing the Gypsies would later be adopted by the Ottoman Empire in the fifteenth century.9 In fact, in the Byzantine Empire, the Gypsies were in effect slaves of the state. It is certain 6  B. D. Grekov, Ţăranii în Rusia (translation), Bucharest, 1952, pp. 151–153. 7  See Al. Eck, “Les non-libres dans la Russie de Moyen-Age”, Revue historique de droit français et étranger, 1930, pp. 26–59. 8  See Helga Köpstein, Zur Sklaverei im ausgehenden Byzanz, Berlin, 1966 (with bibliography). 9  Ilse Rochow, K.-P. Matschke, op. cit., pp. 243–246.

986

Achim

that the same situation existed in the medieval states of Bulgaria and Serbia, however, the very few documents to have survived from these two states do not make explicit reference to the slavery of the Gypsies. This state of affairs is natural when we bear in mind the fate of Bulgaria and Serbia, conquered by the Ottoman Empire in the final decade of the fourteenth century and in the middle of the following century respectively, resulting in the destruction of their social organisation, and the fact that the period of time between the arrival of the Gypsies there and the liquidation of the two Balkan states was short in duration. Nevertheless, given the position of the Gypsies during Ottoman domination, we believe that they would also have been slaves in medieval Bulgaria and Serbia. This would mean that when in the second half of the fourteenth century the Gypsies travelling from the Balkans crossed to the north of the Danube, they were already slaves. It is worth recalling that in the Romanian lands, until the abolition of the institution (although we can presume that this was also the case in the lands south of the Danube in the fourteenth century), the status of slave did not necessarily imply being tied to a particular estate, but rather to a particular owner. The vast majority of the (enslaved) Gypsies were nomadic, tied to their master by certain obligations. The Gypsies that crossed the Danube into Wallachia, where they were “enslaved” (i.e., they entered into the possession of the Crown as slaves), did not lose their freedom, since they had never been free. By changing country, they in fact exchanged one master for another. Their social conditions did not, as a result, undergo any essential changes. The slavery of the Gypsies in the Romanian principalities can, therefore, be also explained by the latters’ location in the historical south-east European space, where Gypsies were held as slaves even before their arrival north of the Danube. The Romanian principalities acquired the Gypsies as slaves, although the institution of slavery dated from an earlier time there, from the time of the battles with the Tatars. Another explanation for the slavery of the Gypsies in the Romanian principalities has also been offered. It has been stated that the Gypsies were not slaves from the beginning of their presence in the Romanian principalities, rather that their enslavement occurred at a later stage. P. N. Panaitescu, the author of this hypothesis, considers that their enslavement had a strictly economic motive, namely the need for labour force in the Romanian principalities in the Middle Ages. After the Crusades, when the Romanian states, thanks to their geographical position, took part in the East-West trade, the reduced number of peasants and the fact that those were not good craftsmen, especially blacksmiths, of which there was great need, determined their feudal masters

The Gypsies in the Romanian Lands During the Middle Ages

987

to force the Gypsies to settle on their estates, thereby forfeiting their freedom.10 The importance of the trade route passing through the Romanian states, especially in the second half of the fourteenth century, and the relative prosperity generated by trade for more than a century are historical realities that are beyond question. It is, therefore, natural that there would have been need for labour force on the great estates. The donations and purchases of slaves are proof of this state of affairs. However, the economic role of the slaves on these estates and in the country’s economy as a whole should not be exaggerated. When the Romanian states began to have a role in European trade, the Gypsies were already in a state of slavery. The economic interest of the leading estate owners cannot explain a social and institutional status. As we have seen, the slavery of the Gypsies is an older phenomenon. 2

Categories of Slaves

The classification of the Gypsy slave population in the Romanian lands11 should be based on precise criteria in order to avoid the possible confusions and overlapping that often appear in literature on this subject. Romanians and foreigners living at the end of the era of slavery observed numerous distinctions within the Gypsy population and produced written descriptions of the different groups, while some even attempted a classification of these groups. In most cases, however, there is confusion with regard to the understanding of the different criteria that made the Gypsy slave population present a highly varied tableau. The first, essential criterion is that of belonging to a master. From this point of view, Gypsy slaves can be divided into three categories: princely slaves, slaves belonging to a monastery and slaves belonging to a boyar. The first category of slaves is referred to in documents as “princes’ Gypsies”, “princely Gypsy slaves”, “Gypsies of the Crown”, later on as “Gospodar’s Gypsies” and in the nineteenth 10  P. N. Panaitescu, Le Rôle économique et social des Tziganes au Moyen Age en Valachie et en Moldavie, in XVII-e Congrès International d’Antropologie et d’Archéologie Préhistorique. VII-e Session de l’Institut International d’Antropologie. Bucarest, 1–8 Septembre 1937, Bucharest, 1939, pp. 933–942; idem, “The Gypsies in Wallachia and Moldavia. A Chapter of Economic History”, JGLS (3), 15 (1941), pp. 58–72. 11  This problem is dealt with by G. Potra, “Despre ţiganii domneşti, mănăstireşti şi boiereşti”, Revista Istorică Română, V–VI (1935–1936), pp. 295–320; idem, Contribuţiuni, pp. 26–65; N. Grigoraş, op. cit., (I), pp. 40–70.

988

Achim

century as “Gypsies of the State”. In certain periods, these constituted the largest category of Gypsies. Gifts of Gypsies by the Wallachian and Moldavian princes to the monasteries and the boyars were made from this fund of princely slaves. It appears that initially the prince was in theory the sole owner of slaves, and he was responsible for the granting of official approval for any transfer of slaves just as in the case of estates. The number of princely slaves also grew via the acquisition of any Gypsy without a master. There are numerous cases of Gypsies passing from one country to another who thus join the ranks of princely slaves. In cases of treason, the boyar’s slaves, like the estates themselves, entered into the possession of the prince. The confiscation of slaves from disloyal boyars is recorded in official documents.12 Most princely slaves carried out their work in the countryside, apart from those who were actually forced to work at the princely court. Princely slaves were required to carry out any work demanded of them. They brought a substantial income to the Crown through the taxes that they were obliged to pay. Generally speaking, their situation was better than that of slaves belonging either to monasteries or boyars’ estates. Princely Gypsy slaves given to monasteries or estates were unwilling to accept their new situation, and for this reason some of them fled from their new masters and returned among the “bands” of princely slaves.13 There were, however, plenty of cases where princely Gypsy slaves fled to other countries. Princely Gypsy slaves often mixed with “private” slaves, especially as a result of marriages contracted without the permission of their masters. Such cases led to many disputes between princes, boyars and the monasteries, which usually resulted in the annulment of the marriages or in exchanges of slaves as compensation. Children born into such marriages were as a rule divided between the private owner and the Crown, represented by princely magistrates. If princely slaves became confused with those belonging to monasteries or estates, the magistrates were ordered to return them to the “bands of princely Gypsies”.14 A separate category of slaves that existed in Moldavia were the “princes’ slaves”. These were Gypsies who were solely the possession of the wife of the prince. They are attested for the first time in 1429, when amongst other things Alexander the Good granted his wife, Princess Marena, a number of Gypsy

12  Examples in N. Grigoraş, op. cit., p. 45. 13  See for example the complaint made in 1753 by a number of princely slaves in Moldavia awarded to the Metropolitanate: “Ioan Neculce”. Buletinul Muzeului Municipal Iaşi, fasc. 2 (1922), p. 308. 14  N, Grigoraş, op. cit., pp. 45–16.

The Gypsies in the Romanian Lands During the Middle Ages

989

slaves.15 The slaves were the possession of the princess, in the sense that she could make a gift of them or sell them. Such slaves had their own organisation, even if they were sometimes included in the bands of princely Gypsy slaves.16 Slaves belonging to monasteries mostly originated from gifts made by the princes and the boyars. Gifts of slaves made by the boyars were more numerous than those made by the princes. The monasteries managed to possess a very large number of slaves, acquired via a number of paths. Cozia monastery, for example, received a gift of 300 families of Gypsies from Mircea the Old.17 The number of slaves belonging to the monasteries also increased as a result of marriages between freemen and Gypsy men and women belonging to the monastery. The rule was that these people, as well as their descendants, had to become slaves. In the seventeenth and eighteenth centuries, enslavement as a result of marriage was a relatively common phenomenon.18 There are in existence a large number of documents concerning the monasteries’ ownership of slaves. The archives of the monasteries have preserved the deeds of donation and confirmation made by the princes, deeds of donations made by the boyars, statistics and registers (lists), and other documents; some documents record in detail the slaves’ origins, names, children, professions, and even any possible legal disputes held with regard to them. The monasteries’ registers also contain information about Gypsy slaves.19 The slaves were used either for agricultural labour or as craftsmen. When they had nothing to occupy them in the fields, they were used for the cutting and transportation of timber. Slave women were used to spin linen. When the slaves were sent to work somewhere else, payments had to be made to the monastery in exchange for their labour. Monastery slaves living around or even within the grounds of the monastery were required to carry out various special tasks. Among their number were included craftsmen and servants. It is supposed that these slaves enjoyed a better situation than those living in the villages (estate Gypsies), who were required to carry out more onerous physical labour.20 The slaves of the boyars’ estates could be procured as a result of princely gifts, inheritances, dowries and from the spoils of war. Deeds of donation show 15  DRH, A, I, pp. 124–126. 16  N. Grigoraş, op. cit., pp. 48–49. 17  DRH, B, I, pp. 25–28 (from the year 1388). 18  See below, pp. 57–58. 19  N. Grigoraş, op. cit., pp. 50–51. 20  G. Potra, “Despre ţiganii domneşti”, pp. 317–319; Olga Cicanci, “Aspecte din viaţa robilor de la mănăstirea Secul în veacurile XVII–XVIII”, Studii şi articole de istorie, X (1967), pp. 158–160.

990

Achim

that the prince granted the slaves to the boyars or that the prince confirmed their possession by official deed as “official proprietor with full economic rights” or “hereditary and inalienable proprietor”. The gifts of slaves made by the prince were usually tied to villages. This demonstrates that the slaves dwelt on estates were granted to the boyars by the prince. The boyars were in full possession of their slaves, i.e. as with any personal property or real estate, they could sell, donate, exchange, mortgage, bequeath them etc. The prince funded the buying and selling, as well as the allocation of slaves. For the boyars, slaves were a cheap source of labour. In the running of a boyar’s estate, slaves played an important role, chiefly as servants and craftsmen, but also to a lesser degree as agricultural labourers.21 Clearly, classifying the Gypsies according to which of the three categories of feudal masters they served tells us little about the occupational and cultural diversity of this population. The Gypsies were far from constituting a homogeneous group. The tableau presented by the Gypsy population during the Middle Ages was particularly varied. Spread throughout the country in relatively large numbers, the Gypsies formed distinct groups that were specialised in certain occupations, with their own cultural and ethnographical characteristics and sometimes even speaking their own separate dialects. Documents produced inside Wallachia and Moldavia attest to these characteristics, even if their interest in them is strictly of a legal nature. Meanwhile, the descriptions provided by foreigners who came into direct contact with the situation in the Romanian principalities make the occupation and way of life of the different categories of Gypsies the principal, if not the only, criterion for their classification. In this way it has been observed that there were always both sedentary Gypsies working on the estate or at the residence of their masters as well as nomadic Gypsies who wandered the countryside. Both groups were slaves. In the first half of the nineteenth century, Mihail Kogălniceanu divided the princely slaves into four categories: rudari or aurari, who were engaged in the collection of gold from river beds; ursari, who wandered the countryside, leading a bear, whom they would encourage to dance the tanana (a Gypsy dance) for paying spectators; lingurari, who made wooden spoons or other household objects; and lăieşi, whose main occupation was as blacksmiths, but who also worked as stonemasons, comb-makers etc. In addition, some of them made a living from stealing. None of the aforementioned had fixed dwellings; instead, they lived in tents and travelled the countryside in search of fresh ways to make

21  N. Grigoraş, op. cit., pp. 59–70.

The Gypsies in the Romanian Lands During the Middle Ages

991

a living.22 As slaves, they paid a sum of money to the Crown, which varied from category to category and according to their specific situation. There were, however, also princely slaves in the towns and at the princely court, where they worked as slaves and craftsmen. Kogălniceanu divided the private slaves, who belonged to the boyars and the monasteries, into two categories: lăieşi and vătraşi. The lăieşi belonging to private owners, like those belonging to the prince, wandered the country, under the obligation to pay a sum of money to their masters. However, when the master opened a building site, the slaves were used as labourers. The vătraşi, meanwhile, had fixed dwellings and, at the time when Kogălniceanu was writing, were already assimilated into the local population; they had forgotten their mother tongue and could not be distinguished from Romanian peasants. There were two types of vătraşi, divided according to their occupation: ţigani căsaşi or de curte (manor Gypsies) and ţigani de ogor or de câmp (estate or field Gypsies). Manor Gypsies served at the boyar’s manor house, carrying out various tasks. The majority of manor Gypsies were craftsmen: blacksmiths, farriers, locksmiths, stonemasons etc. They had a better situation than the other Gypsies. The estate Gypsies were used for agricultural labour; among the privately owned Gypsies the latter were the most numerous and the most heavily worked.24 Thus was the structure of the Gypsy population of Moldavia and Wallachia in the first part of the nineteenth century, shortly before emancipation. It is not a complete model, as there were other categories of Gypsies that are not found in the schema produced by Kogǎlniceanu but which are mentioned by other authors writing in the eighteenth century or at the beginning of the nineteenth century, or in later ethnographic studies of the Gypsies. In fact, over time the categories of the Gypsies have undergone various transformations. Even Kogǎlniceanu observed that the profession of rudari or aurari, once a very profitable business, was by his time in decline, indicating the start of an occupational shift in this category of Gypsies, who would re-orientate themselves as producers and sellers of wooden household objects. It is in this state that we find the rudari in all ethnographic studies devoted to them in this century. The Gypsies had an occupational dynamic that moved in accordance with the general economic changes that affected Romanian society during the 22  M. Kogălniceanu, Esquisse sur l’histoire, les mœurs et la langue des Cigains, Berlin, 1837; reproduced in Opere, vol. I, ed. A. Oţetea, Bucharest, 1946, pp. 572–574. (Future citations will refer to this edition.) 23  Ibid., p. 575. 24  G. Potra, Contribuţiuni, pp. 62–65.

992

Achim

Middle Ages and the modern era. The Gypsies were forced to adapt to new situations. Over time, the evolutions in their situation have led unquestionably to a gradual shift to a sedentary way of life. However, the process of transfer to a sedentary existence and the changes in occupation have not led to the shattering of the old divisions within the Gypsy population. 3

Slavery under the Romanian Ancien Regime

Slavery was an integral part of the social system of the Romanian principalities from their beginnings until the middle of the nineteenth century.25 The terms “slave” (rob) and “slavery” (robie)26 appear as such only at a relatively late stage. Slavery is mentioned as such for the first time in a deed dating from 30 September 1445, in which the prince of Moldavia Stephen II makes a gift to the bishop of Roman a Tatar “from among our Tatars at Neamt”. In the deed, it is specified that the bishop may do as he pleases with the slave; if the slave had been freed, he would have been able to “live there freely according to Romanian law and nobody should dare to remind him of his former slavery (a po holopstvo) or try to enslave him”.27 In 1470, the term “slave” is attested to for the first time, in a document issued by Stephen the Great “to our Tatar and slave (holop) Oană, who fled from Poland”.28 Documents written in the Slavonic language used the term holop for slave. After 1600, in documents written in Romanian, we find the term rob, used initially in parallel with the term of Slavonic origin, only later to replace it altogether. Still, the term şerb / şarbă was also used very frequently in place of holop. Until the second half of the sixteenth century, however, the term rob and its variants were used rarely in Moldavia and not found at all in Wallachian documents. Slaves are indicated almost exclusively by their ethnic origin, either as “Gypsies” or “Tatars” (the latter found only in Moldavia). In the language of the medieval Romanian 25  Works specially devoted to the institution of slavery in the Romanian territories: I. Peretz, Robia (lithographed course), Bucharest, 1934; B. Th. Scurtulencu, Situaţia juridicoeconomică a ţiganilor în principatele române, Iaşi, [1938]; N. Grigoraş, “Robia în Moldova (De la întemeierea statului până la mijlocul secolului al XVIII-lea)”, Anuarul Institutului de istorie şi arheologie “A. D. Xenopol”, Iaşi, IV (1967), pp. 31–79 (I); V (1968), pp. 43–85 (II). See also V. Costăchel, P. P. Panaitescu, A. Cazacu, Viaţa feudală în Ţara Românească şi Moldova (sec. XIV–XVII), Bucharest, 1957, pp. 143–164. 26  For a treatment of this problem, see I. R. Mircea, “Termenii rob, şerb şi holop în documentele slave şi române”, Studii şi cercetări ştiinţifice, Iaşi, I (1950), pp. 856–873. 27  DRH, A, I, p. 367. 28  DRH, A, II, pp. 239–241.

The Gypsies in the Romanian Lands During the Middle Ages

993

chancelleries, the term “Gypsy” (in Moldavia also “Tatar”) always had in addition to its ethnic sense, a social value, indicating a slave.29 Slaves formed a separate category within the social organisation of the Romanian principalities. They made up the lowest rung of the subjugated classes. What defines their social condition is not the absence of personal freedom, since in feudal society the serfs (known as rumâni in Wallachia, vecini in Moldavia and iobagi in Transylvania) were also subjugated, but the fact that they had no status as legal persons. The slave was wholly the property of his master, figuring among his personal property. The master could do as he pleased with the slave: he could put him to work, he could sell him or exchange him for some other good, he could use him as payment for a debt or he could mortgage or bequeath him. The possessions of the slave (consisting mainly of cattle) were also at the discretion of the master. Masters were constantly abusing their rights, as slaves could at any time be punished with a beating or with prison without the need for the intervention of the state authorities. The master did not, however, have the power of life and death over the slave. His sole obligation was to clothe and feed those slaves who worked at his manor. The master was in no way responsible for those slaves who wandered the countryside in order to earn a living, paying the master an established sum of money.30 In the Romanian principalities, there was a slaves’ law. It is mentioned in the Moldavian document of 1470 in which Stephen the Great frees Oană, a Tatar slave who had fled from Poland, as well as his children from slavery. The prince freed them from the obligations that resulted from one’s status as a slave: “let them never pay anything according to the law of the slaves and the Tatars (holopskym[i], tatarskym[i] pravom[i])”; they would be allowed to live in the country “as do all Romanians according to Romanian law (voloskym[i] zakonomi)”.31 Romanian law and slaves’ law were two different entities. Slaves had their own legal status that was different from that of the Romanian population. Slaves’ law consisted of a number of norms that referred chiefly to the obligations of slaves to their master and to the State, to the punishments they were liable to if they failed to fulfil their obligations or if they were found guilty of any crime, as well as the authorities that were put in place to judge them. There were equally norms that regulated relations between slaves and freemen, as well as the authorities that assured that the norms were respected.32 29  I. R. Mircea, op. cit., pp. 860–862. 30  N. Grigoraş, op. cit., (I), pp. 67–70; Istoria dreptului românesc, vol. I, Bucharest, 1980, pp. 486–489. 31  DRH, A, II, pp. 239–241. 32  N. Grigoraş, op. cit., (II), p. 43.

994

Achim

Slaves’ law is an ancient institution, which dates from before the foundation of the Romanian states. Slavery in all its forms falls under common law. For a long time, there were no written laws relating to slaves. When later on in Wallachia and Moldavia there appeared a tendency to invoke Byzantine law, the provisions of Byzantine legislation were adopted with regard to slaves. The large number of Greek and Slavonic texts of legal nature with regard to slaves in Wallachia and Moldavia33 is an indicator of the importance attached to the harmonisation of the de facto situation with the canons of Byzantine law. Legislative documents printed in Romanian in the mid-seventeenth century—namely the Pravila de la Govora (Law Book of Govora) of 1640 (in Wallachia) and the two legal codes, Vasile Lupu’s 1646 Cartea românească de învăţătură (Romanian Book of Teachings) (Moldavia) and Matei Basarab’s 1652 Îndreptarea legii (Improvment of the Law) (Wallachia)—record the legal norms mostly of Byzantine origin but also with some norms that came under common law in use up until then, relating to slaves.34 In practice, however, the status of slaves was dependent exclusively on common law. Official documents relating to slaves issued by the Prince Chancellery or by other state institutions, including those of legal nature, always make reference to customary laws (the “tradition of the land”), not to the written law of Byzantine inspiration. The obligations incumbent on the Gypsies were fixed by tradition. Official documents enable us to gain a more intimate knowledge of the obligations that princely Gypsies had to the State, their mode of organisation, the exemptions and the other privileges from which the different categories of princely Gypsies benefited. Private slaves were in principle exempt from any obligations to the State. We find that in the fifteenth century, both in Moldavia and Wallachia, it was forbidden for princely officials to impose royal service on private slaves.35 Princes would make reference to such provisions whenever they gave confirmation of previous awards of slaves or at the request of their masters. On some occasions, however, the Crown, in search of fresh sources of revenue, sought to impose at least partially on private Gypsies the same regime as that applied to princely Gypsies. At least from the first half of the seventeenth century, if not from the second half of the previous century, we find that slaves belonging to monasteries and boyars’ estates are also obliged 33  For the old legislation with regard to slaves used in the Romanian principalities, see I. Peretz, Curs de istoria dreptului român, vol. II, part I, Bucharest, 1928, pp. 187–440. 34  For their content, ibid., pp. 315–140; J. Peretz, Robia, pp. 83–119. 35  DRH, A, I, p. 23 (31 October 1402, Moldavia); DRH, B, I, pp. 201–202 (5 March 1458, Wallachia).

The Gypsies in the Romanian Lands During the Middle Ages

995

to pay certain taxes and carry out certain tasks for the State. In Moldavia, we find them paying tax as early as the 1620s. Later on, still more obligations were imposed upon them: they were obliged to pay tithes on beehives (desetină de stupi) and on boar (goştină de mascuri) (if they owned such goods), to place post horses (cai de olac) at the disposition of the Crown and provide transport (podvoadă) (if they had horses), to pay the mucarer,* to carry out certain labour tasks for State etc.36 These petty obligations were numerous and varied from era to era. Freemen and princely Gypsies were required to carry them out unconditionally, while private Gypsies carried them out only on in exceptional cases. Furthermore, when the obligations were applied to private Gypsies, the slaves of some monasteries and boyars could be exempted from carrying them out by special deed issued by the Crown. Such obligations were not only a heavy burden for the slaves, but also for their masters, who were in fact responsible for ensuring that their slaves carried out their obligations. There are many cases where the increased exploitation of Gypsy slaves resulted in the fleeing of Gypsies from one estate to another or even from one country to another. The most radical measure of this kind was the introduction of ţigănărit (Gypsy tax) in Moldavia at the beginning of the eighteenth century. It was introduced by Nicolae Mavrocordat, probably in 1711, as an exceptional tax “for the needs of the country”, and was abolished in 1714. Mihail Racoviţa reapplied it in 1725, but withdrew it as a result of the intervention of the boyars and the monasteries. Under the terms of the tax, boyars and monasteries were required to pay two ducats for each Gypsy in their possession.37 In fact, the tax was an extension to all the Gypsies of the dajdie (tax) paid by princely Gypsies. There were special officials appointed to supervise the princely Gypsies. In time, a network of princely officials was organised on a hierarchical and territorial basis, which dealt with every aspect of relations between the Gypsies and the State and those between the Gypsies and the inhabitants of the country. Special officials collected the taxes owed by the Gypsies to the Crown.38

*  Desetină de stupi, goştină de mascuri and mucarer were all taxes that those subject to tax were required to pay to the State. Cai de olac (post horses) were horses that villages situated on major roads were obliged to make available to officials so that the latter could continue their journey at maximum speed. Podvoadă or cărătură was the obligation of transport to be provided to the crown. An explication of these obligations can be found in Instituţii feudale din ţările române. Dicţionar, Bucharest, 1988. 36  N. Grigoraş, op. cit., (I), pp. 70–76. 37  See G. Potra, Contribuţiuni, pp. 69–71; N. Grigoraş, op. cit., pp. 73–74. 38  See below pp. 61–62.

996

Achim

Disputes that arose among the Gypsies, with the exception of manslaughter cases, were dealt with by their leaders: village leaders, sheriffs and Gypsy leaders. The master and his clerks in the case of boyars’ and monastery Gypsies, and the special officials in the case of princely Gypsies were invested with the power to punish and fine Gypsies. Cases of manslaughter and disputes between Gypsies and other inhabitants of the country fell under the jurisdiction of the state judiciary. Slaves found to be counterfeiting money and those who committed crimes out of the ordinary were judged by the State Council itself. Slaves did not have the right to defend themselves before a tribunal and at the same time they could not be held legally responsible for damages caused to freemen, for which their masters were accountable. However, as part of the compensation they were required to pay, masters could renounce their ownership of the guilty slaves in favour of the injured party. If a slave killed a slave belonging to another owner, as a rule the killer, although condemned to death, was not executed but given in exchange for the dead slave. The decision to annul the implementation of capital punishment was taken by the prince with the approval of the family of the deceased. Owners of slaves did not have the right to punish their slaves by putting them to death. Generally speaking, any free person who killed slaves with premeditation was liable to receive the death penalty. Such cases were judged by the country’s supreme court and presided over by the prince. There are, however, no known cases in which a boyar who killed one of his slaves suffered the same punishment. In cases where a master killed somebody else’s slave, the former offered the injured party a slave in place of the one he had killed.39 Generally speaking, the law was lenient on Gypsies. The law book of Vasile Lupu made the following provision: “If a Gypsy, his wife or his offspring should steal once, twice or three times a hen, goose or other small thing, the theft shall be forgiven: if they should steal again, they shall be punished as would common thieves.”40 The explanation for this state of affairs lies of course not only in the fact that the slave has no legal status, but also in the large number of crimes that took place among this marginal social category, as the application of those punishments resulting in imprisonment or execution to which the slaves were subject was not in the interest of their feudal masters. The work provided by the slave was as a first or last resort addressed to his master.

39  N. Grigoraş, op. cit., (II), pp. 43–52; Istoria dreptului românesc, loc. cit. 40   Carte românească de învăţătură. Ediţie critică, Bucharest, 1961, p. 68.

The Gypsies in the Romanian Lands During the Middle Ages

997

Slaves were allowed to marry, but only with the approval of their master.41 In cases of marriage between Gypsies belonging to two separate masters, the Gypsies were required to obtain the approval of both masters. In most cases, a preliminary financial settlement was agreed by the two masters: under the terms of such an agreement, it was settled that either one master purchased from the other the Gypsy set to move to his estate as a result of the marriage, or a compensatory exchange of slaves would be carried out, in which the master providing another slave of equal value in exchange for the Gypsy he was to obtain. Marriages performed without the prior approval of the masters were common. Particularly those cases involving a princely Gypsy and a Gypsy from a boyar’s estate or a monastery were recorded in contemporary documents. If in such situations an arrangement could not be reached, the two Gypsies were separated by force by their owners and the children resulting from the marriage were divided between the owners without parental approval.42 Such situations were commonplace. In the eighteenth century, modifications were applied to the ancient law. The resolution of cases of marriage between Gypsies belonging to different masters was no longer left to the discretion of their masters. The indissolubility of marriage was decreed regardless of the conditions under which it was contracted. The “establishment” of Constantin Mavrocordat of March 1743 in Moldavia forbade owners of slaves from separating married Gypsies belonging to different masters. In such situations, the masters were only allowed either to divide up the children resulting from the marriage or to carry out a compensatory exchange for the Gypsy (or Gypsy woman) and the children to which the respective parties were entitled.43 This first intercession into the old tradition relating to slavery was the work of the Phanariot ruler Constantin Mavrocordat who, in the spirit of the age, introduced both in Moldavia and Wallachia a series of reforms that aimed at ensuring the social, administrative and fiscal modernisation of the principalities. The most important of these reforms was the one that abolished serfdom in Wallachia in 1746 and in Moldavia in 1749, giving the peasants back their

41  For the problem of the marriage of slaves see B. Th. Scurtulencu, op. cit., pp. 19–26; N. Grigoraş, op. cit., pp. 70–75; Gh. T. Ionescu, “O anaforà din domnia lui Constantin Hangerli privind un interesant caz de eliberare de robie prin căsătorie”, Analele Universităţi Bucureşti, ser. Ştiinţe sociale, Istorie, XVII (1968), pp. 155–168. 42  N. Grigoraş, op. cit., (I), pp. 45–46. 43  Gh. I. Brătianu, “Două veacuri de la reforma lui Constantin Mavrocordat 1746–1946”, Analele Academiei Române, Memoriile Secţiunii Istorice, s. III, t. XXIX (1946–1947), p. 450.

998

Achim

freedom. As for the status of slaves, the modifications carried out at this time were limited to the problem of marriage. The same regulations were applied in both principalities. Subsequent laws made strict provision for the way in which owners of slaves were to carry out transfers of slaves. The Sobornicescul hrisov (Ecumenical Charter) of 1785 (Moldavia) established a fixed price for such transfers: fifty lei for a Gypsy woman and seventy lei for a Gypsy man. A slave skilled in a particular trade, however, was worth more, as to the price per person was added the “price for the trade of the Gypsy or Gypsy woman”. Likewise, it was no longer permitted for children to be taken from their parents. It was compulsory for those children to whom in theory the master was entitled to be bought back by the master who retained the family. Children over the age of sixteen were paid for like adults, while those under the age of sixteen were worth half price.44 It was at this time that the principle that the family was to remain whole was established. Gypsy slaves acquired the right not to have their families split up, that is to say that they acquired the right for children not to be donated or sold separately from their parents nor siblings to be donated or sold separately from one another.45 The most important new element to be introduced in the eighteenth century was that regarding mixed marriages (i.e., a Gypsy with a Romanian woman or a Romanian with a Gypsy woman). Until then, the rule had been that by marrying a slave, the free husband would also enter into the state of slavery, together with children born out of their union. The “establishment” of Constantin Mavrocordat stipulated that a Romanian man or woman who married a Gypsy could no longer be made a slave. The freeman retained the social status he or she held before the marriage, while the slave remained a slave. Children born out of their union were to be free.46 It appears that in the period immediately before the adoption of this measure, mixed marriages were fairly frequent. For the boyars, mixed marriages were a means of increasing the number of slaves. For the State, however, mixed marriages represented a loss, with the peasant transformed into a slave becoming exempt from the payment of tax and other public obligations. For this reason, we believe that the relinquishing of the common law that allowed the enslavement of Romanians was also done for fiscal reasons. In Wallachia, the measure was respected, but in Moldavia opposition from the boyars to this innovation led to the gradual restricting of its scope and finally to its abandonment. In 1766, such marriages 44  T. Codrescu, Uricariul, vol. I, 2nd ed., Iaşi, 1871, pp. 320–327. 45  N. Grigoraş, op. cit., pp. 64–66. 46  Gh. I. Brătianu, loc. cit.

The Gypsies in the Romanian Lands During the Middle Ages

999

were outlawed, with priests prevented from officiating at such weddings. In cases where such a marriage did take place, the spouses were to be separated. The children of such marriages, however, “remained among the Moldavians”, i.e. free.47 The Sobornicescul hrisov (Ecumenical Charter) of 1785 completely outlawed marriages between Moldavians and Gypsies and declared such marriages to be invalid. Children born out of such marriages were considered to be Gypsies. This meant a return to the “centuries-long tradition” of the past, in other words, to the annulment of the reform. In Moldavia, the phenomenon of enslavement through marriage continued to exist until a late stage. In Wallachia, the Pravilniceasca Condică (Legal Register) of 1780 stipulated that marriages between Gypsy men and free women were immediately split up, and the children born out of their union became freemen.48 In the final years of slavery, the Organic Regulations introduced in the two principalities in the years 1831–32 had a similar content when it came to the matter of marrying slaves. Marriages between freemen and slaves were forbidden. A freeman who married a Gypsy woman without knowing her to be so was allowed to buy back her freedom. The same applied to a Romanian woman married with a Gypsy. Any person who married a Gypsy in full knowledge of what they were doing was required to pay the price of the Gypsy woman to the almshouse. Children from such marriages were free.49 The reverse phenomenon, of release from slavery, also existed. In certain situations, for example for particular services carried out during the life of the master or after his death through his testament, the master could “release” a slave from slavery. The latter would thus obtain personal freedom and joined the ranks of Romanians. As freemen, these Gypsies could own land, while in towns they could own property. There were numerous cases of Gypsies who ended up selling themselves to a boyar or to a monastery in order to escape punishment, to escape their debts or to avoid dying of starvation.50 Even if medieval documents provide us with sufficient examples of free Gypsies, release from slavery was carried out only in exceptional circumstances. Only towards the middle of the nineteenth century, in the conditions of the appearance of a new attitude towards slavery and the abolitionist movement did gestures of this nature become relatively common. One of the princes of 47  I. Peretz, Curs de istoria dreptului român, vol. IV, Hrisoavele domneşti, Bucharest, 1931, pp. 47–54. 48  B. Th. Scurtulencu, op. cit., pp. 25–26. 49  Ibid., pp. 24–25. 50  For free Gypsies, see G. Potra, Contribuţiuni, pp. 43–44, 59–62; N. Grigoraş, op. cit., (I), pp. 36–38, 76–77; (II), pp. 76–80.

1000

Achim

Moldova, Ştefan Răzvan, was himself of Gypsy origin.51 In a text originating from Michael the Brave,52 we learn that Răzvan was the son of a princely slave woman from Wallachia. He managed to become a boyar, was sent in delegation to Constantinople, became hetman in the Cossack and Polish armies and finally occupied the throne of Moldavia for a short period of time (April to August 1595). The rise of Ştefan Răzvan is indicative of the social mobility that existed in the Romanian society, where some slaves could obtain their freedom and in exceptional cases could accede to the status of nobles and obtain high positions. At the end of the eighteenth century, regulations regarding the way in which slave owners could carry out the release of their slaves were introduced, as in the case of the Sobornicescul hrisov of 1785. In Wallachia, release from slavery was also carried out through marriage. A slave married to a freewoman with her knowledge and with the permission of his master (including verbal permission) would become a freeman and the marriage would not be annulled, while their children would also remain free.53 The new elements that appeared in the regulations relating to slavery in the eighteenth century referred almost exclusively to instances of marriage and the social consequence of such marriages. Collections of laws of the Phanariot rulers from the end of the eighteenth century and the beginning of the nineteenth century, namely the Pravilniceasca Condică (Legal Register) of Alexandru Ipsilanti (1780) and the Legiuirea Caragea (Caragea Legislation) (1818) in Wallachia, the legal code compiled by Andronache Donici in 1814 and the Condica civilă a Principatului Moldovei (Legal register of the Principality of Moldavia) (also known under the name of Codul Callimach) (Callimach Code) (1817) in Moldavia, all contain special chapters relating to slaves.54 Influenced by Western laws, these legal codes attempted to introduce certain elements of natural law into the treatment of slavery. Codul Callimach speaks of slavery as being “against the natural law of man”, but the institution is justified by the fact that it has been followed since ancient times. The code tries to define a modern concept of slavery.55 In essence, however, slavery did not undergo 51  See A. D. Xenopol, Istoria românilor din Dacia traiană, vol. III, Iaşi, 1890, pp. 194, 216–218, 481. 52  N. Iorga, “Originea lui Stefan Răzvan”, Analele Academiei Române, Memoriile Secţiunii Istorice, s. III, t. IX (1930), pp. 157–163. 53  See Gh. T. Ionescu, op. cit., pp. 162–166. 54  See these texts in I. Peretz, Curs de istoria dreptului român, IV, Hrisoavele domneşti, pp. 169–216. 55  “[…] the slave is not reckoned completely to be like an object, where his deeds, rights and obligations affect others (although not his master), he is reckoned to be like a person;

The Gypsies in the Romanian Lands During the Middle Ages

1001

modification. The Ştate did not intervene in relations between master and slave. In any case, slavery as an institution could not be reformed. It remained as such until abolition. Nor were the restriction of nomadism and the settling of the Gypsies who wandered the countryside of concern to lawmakers in this new era. Nevertheless, in accordance with the Enlightenment, a new spirit had begun to manifest itself, which, if it did not question the institution of slavery itself, tended to regard slaves as human beings. The arguments of the Metropolitanate and the bishops of Moldavia in 1766, when it was forbidden for families of Gypsies to be split up, are indicative of this new spirit: “if they are Gypsies, being part of God’s creation, they can by no means be shared out as if animals”.56 A parallel can be drawn between this attitude and certain initiatives undertaken by enlightened members of the church linked to the conversion of groups of nomads to Christianity. 4

The Social and Legal Situation of the Gypsies in Transylvania

With regard to the situation of the Gypsies during the Middle Ages in Transylvania,57—a Romanian province then belonging to the medieval Kingdom of Hungary, becoming an autonomous principality in the mid-sixteenth century only to pass under the dominion of the Habsburg Empire at the end of the seventeenth century—their social and legal status was marked by the manner in which the Gypsies appeared in the region and by the regional specificities that prevailed there. We have already seen that the first evidence of the presence of the Gypsies in Transylvania, which originates from the official records of Wallachia, refers to the land of Făgăraş, where around the year 1400 a boyar was recorded as being in possession of several villages and seventeen Gypsies dwelling in tents.58 This area, which borders Wallachia, was for a long time (from the second half of the fourteenth century until the end of the fifteenth century) under the control of the prince of Wallachia, who held the title of fief there. As a consequence of its being part of Wallachia and the perpetuation of the old Romanian mode of organisation there, the social institutions in Făgăraş were identical to those in thus is the slave subject to earthly laws and is protected by them.” Cf. Istoria dreptului românesc, vol. II/I, Bucharest, 1984, p. 244. 56  I. Peretz, Curs de istoria dreptului român, IV, Hrisoavele domneşti, p. 49. 57  There is a work devoted to this subject, which should, however, be treated with caution: A. Gebora, Situaţia juridică a ţiganilor în Ardeal, Bucharest, 1932. 58  See above p. 14.

1002

Achim

Wallachia, i.e. the Gypsies were the slaves of the boyars of Făgăraş. In the fifteenth century, the princes of Wallachia confirmed the dominion of the boyars of Făgăraş over the Gypsies in their possession.59 The slavery of the Gypsies in Făgăraş continued as a legacy of Wallachian rule even after reincorporating of the region into the Transylvanian voivodate (later principality). Transylvanian documents attest to this special social regime in force in Făgăraş that differed from the rest of the country until the end of the seventeenth century with the establishment of Habsburg rule.60 For example, in 1556 Queen Isabella issued confirmation of the possessions of the boyars in Recea, which included several families of Gypsies. Mihály Apafi, prince of Transylvania, confirms the deed later on, in 1689.61 There are also indicators that the Gypsies were also enslaved in regions of Transylvania that were temporarily under the authority of the princes of Moldavia in the Middle Ages. For example, Petru Rareş purchased from the mayor of Bistriţa, a family of Gypsies composed of husband, wife and six children for fifty Hungarian florins and a horse.62 There are also cases of Moldavian boyars purchasing slaves from Transylvania.63 It appears that the Gypsies of Bran Castle also had the status of slaves. They are mentioned for the first time only in 1500, when the king of Hungary, Vladislav II, ordered the voivode of Transylvania to prevent his civil servants from arresting or judging Gypsies (certi Egiptii seu Cigani) who ab antiquo belonged to Bran Castle as only the castellan had the right to do so.64 From a list of the castle’s accounts from 1504, we learn that the obligations of the Gypsies to the castellan consisted in the annual payment of a money tax and in certain services. It appears that King Vladislav II subjugated the Gypsies to the town of Braşov in 1498 together with the award of Bran Castle. In this way, they became the town’s serfs. The rights over them formerly held by the castellan at Bran Castle were passed to the town.65 How can we explain the fact that Bran Castle owned Gypsies, a case that was unique both in Transylvania 59  In 1441, the boyar Stanciu Moenescul of Voila received confirmation of ownership of his Gypsies Manea, Pascul, Cazac and Micul (DRH, B, I, pp. 160–162), while in 1476, in a deed issued by Basarab the Old, we see that the boyar Şerban of Şinca was master over the Gypsies Radul, Lalu, Curchea, Mujea and Costea (ibid., pp. 253–256). 60  See D. Prodan, Boieri şi vecini în Ţara Făgăraşului în sec. XVI–XVII, in Din istoria Transilvaniei. Studii şi evocări, Bucharest, 1991, pp. 22–27, 149. 61  I. Puşcariu, Fragmente istorice despre boierii din Tara Făgăraşului, Sibiu, 1907, pp. 416–422. 62   Documente privind istoria României, A, Moldova, veac XVI, vol. I, Bucharest, 1953, p. 409. 63  See N. Grigoraş, op. cit., (I), p. 37. 64  Hurmuzaki, XV/l, p. 152. 65  Maja Philippi, op. cit., pp. 142–143.

The Gypsies in the Romanian Lands During the Middle Ages

1003

and the other Romanian lands? Bran Castle, built in 1377, was ceded by King Sigismund of Luxembourg to Mircea the Old, prince of Wallachia, probably in 1406. It remained in the possession of Wallachia until 1419. It is possible that the attention paid by the Wallachian prince to the castle, which controlled the main route into Transylvania and which held an important strategic and commercial role, could have manifested itself by an award (or awards) of Gypsies. These Gypsies were undoubtedly princely slaves who had been allocated to the castle and placed under the direct control of the castellan. The 1500 Gypsies owned by Bran Castle at the beginning of the sixteenth century is unusually large. When the castle and its estate were awarded to Braşov, the Gypsy slaves found themselves under the authority of the town. The social status accorded to the Gypsies under their new masters was that of serfs, in line with the social system in place in Transylvania at the time. Their de facto situation, however, their dependence and regime of obligations, remained unchanged in comparison with the period of Wallachian dominion in Bran. A small part of the Gypsies from medieval Transylvania lived under conditions of slavery. The majority of them were a kind of “royal serfs”, directly dependent on the king. It was the king who accorded to different groups of Gypsies the freedom to live in the country, while the only obligations imposed on them were those they had to the Crown: they were required to pay certain taxes and to provide certain services for the State. The first deed relating to the Gypsies issued by the royal authorities in Transylvania is a grant of 1422 in which King Sigismund grants the voivode Vladislav the right to travel the country freely with his band of Gypsies.66 It would appear that the legal status that would for a long time apply to the Gypsies in the voivodate of Transylvania and the later in the Transylvanian principality was established during the reign of Sigismund. That is, they were placed under the protection of the Crown. In the medieval Kingdom of Hungary, with a few exceptions, the Gypsies were not the possession of any feudal master, whether secular or religious. Their admission to and settlement on private estates was done solely with the approval of the king. Thus, in 1476 King Matthias Corvinus granted permission for the town of Sibiu to make use of Gypsies living on the edge of town for labour where required.67 The status of the Gypsies was quite distinct from that of the rest of the population. In fact, the Gypsies had the status of one of the autonomous ethnic groups that existed in the kingdom at that time. It is significant that they were not placed under the jurisdiction of the authorities; instead they were allowed 66  Hurmuzaki, I/2, p. 527. 67  A. Gebora, op. cit., pp. 13–14.

1004

Achim

to remain under the authority of their leaders, known as “voivodes”. In theory, the authorities did not have power over the Gypsies, who were directly dependent on the king. In the deed of 1476, Matthias Corvinus ordered that voivodes and deputy voivodes should not dare to attempt to take the Gypsies settled in Sibiu under their jurisdiction, leaving them instead under the jurisdiction of the town.68 The grants to the Gypsies are a constant reminder of this exceptional situation. They attest to the Gypsies’ freedom to travel the country freely, to sojourn on lands belonging to the Crown, the internal autonomy of the bands of Gypsies, the regime of obligations to the State (fewer than those of the sedentary population), the absence of military obligations, the authorities’ tolerance of the Gypsies’ non-adherence to Christianity. All of the above made up a system of wide-ranging privileges that functioned in the medieval Kingdom of Hungary until its collapse with the occupation of Buda by the Ottomans in 1541, and which even continued to function after this time in the autonomous Transylvanian principality. In the sixteenth century, in Transylvania a Gypsy voivodeship was created, an authority led by a noble holding the title of voivode. The voivode held fiscal, judicial and administrative responsibilities and managed all aspects of the Gypsies’ relations with the State.69 The Gypsies were a source of revenues to the State. Under the Transylvanian principality, each tent-dwelling Gypsy was required to pay a qualification of one florin—fifty dinars on Saint George’s day (24 April) and fifty dinars on Saint Michael’s day (29 September)—at the headquarters of the county in which he was then staying, while the money was collected by a civil servant in the service of the Gypsies’ voivode.70 During this period a serf was required to pay a qualification of two florins per year. At the time, society was tolerant towards the nomadic lifestyle of the Gypsies, while craftsmen, especially blacksmiths, were integrated in the rural economy of the time. The Crown had a vested interest of a fiscal nature in the existence of these social—ethnic categories, hence the preoccupation that their status as “royal serfs” be maintained. While the different groups of nomadic Gypsies benefited from the aforementioned privileges, a part of the Gypsies of Transylvania with time settled into a sedentary lifestyle, settling on some nobles’ estates, in villages of serfs or at the edges of free villages and towns. In both cases, the respective Gypsies lost their privileges, if not at once, then after a relatively short period of time. The Gypsies who settled on the nobles’ estates became serfs or landless peasants, 68  Ibid. 69  See below pp. 63–64. 70  D. Prodan, Iobăgia în secolul al XVI-lea, vol. I, Bucharest, 1967, p. 415.

The Gypsies in the Romanian Lands During the Middle Ages

1005

while those who settled in towns and Saxon villages preserved their personal freedom, living as second-class inhabitants on the edges of those localities. The process of linguistic and cultural assimilation took place more particularly in the case of Gypsies settled in villages of serfs and less so in the towns and the Saxon villages. 5

The Position of the Gypsies in the Economy of the Romanian Lands

P. N. Panaitescu, the author of a study that looks specifically at this problem, considered that “the history of the Gypsies in Wallachia and Moldavia forms part of the economic history of the two principalities”.71 He shows that the Gypsies acquired an important role in the Romanian economy after the era of economic prosperity of the fourteenth and fifteenth centuries (when the Romanian principalities participated to the full as transit countries on the East-West trade routes linking Europe to the Black Sea and the East) had come to an end following the occupation by the Ottomans of the Romanian ports on the Danube and the Black Sea in 1484 and 1540. The ensuing economic transformations led to the disappearance of the great farms. These great estates needed craftsmen of all kinds. Since there were no Romanian craftsmen, while the foreign craftsmen who had come to the principalities together with the East-West trade of the previous centuries, had disappeared along with the decline of the trade, the Gypsies were a welcome source of skilled labour. They had been craftsmen since ancient times and possessed the extra advantage of being able to adapt quickly to the economic needs of the country. Panaitescu draws a connection between the Gypsies’ status as slaves and the boyars’ interest in guaranteeing a supply of this precious source of labour, thereby preventing their flight from the principalities. In this way, the great estates are shown to have been the cause of slavery.72 This theory is debatable from several points of view. However, there is no doubt that the function performed by the Gypsies in the Middle Ages in the Romanian lands is that of craftsmen, especially in the principalities. As in the entire Central and Eastern European area, town dwellers and craftsmen were originally overwhelmingly of foreign ethnic stock. In Wallachia and Moldavia, the history of the oldest towns is linked to the communities of German craftsmen and merchants who settled to the south and the east of the Carpathians at the end of the thirteenth century and during the fourteenth 71  P. N. Panaitescu, “Le Rôle économique”, p. 936. 72  Ibid., pp. 938–940.

1006

Achim

century. Romanian society, an agricultural society par excellence, was in need of foreign craftsmen. At the same time, there were also Romanian craftsmen. A little later on, even in the urban centres initially populated by foreigners, the vast majority of craftsmen and merchants were Romanians. In the villages, a relatively broad range of crafts had always been practised on the farmsteads, whilst in some regions there were villages specialised in a particular craft. The Gypsies who arrived from the south of the Danube added to the numbers of craftsmen present in the country. They were specialised (or became specialised) in crafts that could not be provided by rural Romanian craftsmen. Filling in this absence they were thus able to find a purpose in their new homeland. In the second half of the eighteenth century, an observer of the contemporary situation in Moldavia and Wallachia commented that here “all the mechanical crafts are in the hands of the Gypsies or of foreigners from neighbouring countries”.73 Documentary information enables us to gain a better understanding of the occupations of the Gypsies, especially in the eighteenth century and the first half of the nineteenth century. There is a paucity of information of this kind for the first centuries of the Gypsies’ presence in the Romanian principalities. We can suppose that over time an evolution took place in this field, with different groups of Gypsies adapting to situations that existed in different periods, leading eventually to the accentuation of the differentiation in their occupations. Generally speaking, however, the bands of Gypsies preserved their characteristic pursuits for a long time, with this situation continuing in some cases almost until the present day. A general characteristic among the Gypsies was the passing of a craft from generation to generation within a family. Furthermore, crafts did not pass from one clan (or category) to another, but remained within that particular clan. It is for this reason that observers have spoken of the division of the Gypsies into “natural guilds”.74 The Gypsies’ preferred craft was that of blacksmith. Throughout the Middle Ages on the Romanian territories the working of iron was an occupation reserved almost exclusively for them. The oldest literary reference to the presence of the Gypsies, the Ruinae Pannonicae by Christian Schesaeus, published in 1571, alludes to this occupation of the Gypsies.75 Foreigners familiar with Romanian realities constantly attributed the profession of blacksmith to the 73  I. St. Raicewich, Bemerkungen über die Moldau und Wallachey in Rücksicht auf Geschichte, Naturprodukte und Politik. Aus dem Italienischen, Vienna, 1789, p. 126. 74  I. Chelcea, Ţiganii din România. Monografie etnografică, Bucharest, 1944, p. 102. 75  Cf. Şt. Bezdechi, “Christian Schesaeus despre români”, Anuarul Institutului de Istorie Naţională, IV (1926–1927), p. 448.

The Gypsies in the Romanian Lands During the Middle Ages

1007

Gypsies, such as Charles de Peyssonnel in 1765.76 The production, whether for the boyars’ estates, for peasants or for the State, of tools made of iron— horseshoes, nails, even armour—was one of the chief occupations of the Gypsies. There were certain categories of Gypsies who specialised in this occupation. The lăieşi, who wandered the countryside, were exclusively engaged in this occupation until the beginning of the nineteenth century. The ursari also produced small objects from iron such as knives, axes and locks. Slaves working as blacksmiths were indispensable to any feudal economy and they existed in quite large numbers.77 Some Gypsy blacksmiths practised their craft either in workshops at their master’s residence or elsewhere on his estate, or in the towns, in workshops that belonged to themselves. The majority of them, however, were itinerants, moving from place to place and working with rudimentary tools. Bands of Gypsy blacksmiths, whether princely, boyars’ or monastery Gypsies, wandered throughout the country plying their trade. A band, or more probably a family or two, would stop for a time in a village and carry out, either in exchange for goods or money, all the iron work requested from them and sell the iron goods that they produced during their stay there. The Gypsies would regularly return to the village, and in time some would settle on the edge of the village where they would set up their workshop. Initially, they would build a shelter and then later a house of the kind built by the villagers. However, Gypsy families’ transition to a sedentary way of life in the villages only took place at the beginning of the nineteenth century, when there was an increasing need for their services and when the authorities actually encouraged the process. Ethnographic research carried out in the first half of the twentieth century found that in the former principalities there was not a single peasant farmstead that did not own pieces of ironwork produced by Gypsies.78 As a consequence of their monopoly in this field, “Gypsy” came to mean “blacksmith” in the Romanian villages. There were also Gypsy locksmiths, knife makers and sword makers, coppersmiths and goldsmiths. Among the crafts practised by the Gypsies, there is also reference to the manufacture of sieves, stone working, brick-making, 76  “Dans la Valaquie, la forge este l’unique occupation des Bohémiens” (Ch. de Peyssonnel, Observations historiques et géographiques sur les peoples barbares qui ont habité les bords du Danube et du Pont-Euxin, Paris, 1765, p. 111). 77  N. Grigoraş,, op. cit., (II), p. 67. 78  M. Block, Die materielle Kultur der rumänischen Zigeuner. Versuch einer monographischen Darstellung, comp. and ed. J. S. Hohmann, Frankfurt am Main, Bern, New York, Paris, 1991, p. 131.

1008

Achim

the production of spoons and saddles, pottery and milling. There were also Gypsy slaves who worked as cobblers, harness makers, cooks, innkeepers etc.79 Gypsies were used to extract salt in salt mines, namely at Ocnele Mari, the largest salt mines in Wallachia. From the time of Mircea the Old until almost the mid-eighteenth century, salt extraction at Ocnele Mari was carried out by Gypsy salt miners, slaves of the monasteries of Cozia and Govora. Later on, they were joined by workers recruited from among the peasantry. Slaves from the two monasteries worked there until emancipation. They were paid wages in return for their labour, like free workers.80 There were also other occupations that fell to the slaves to perform. Gypsy craftsmen were also present in the towns. Some lived at the residence of their master or in the Gypsy quarter of the monastery. Others, even though they were slaves, had their own homestead and made a living from a particular craft. In Wallachian and Moldavian towns there existed a certain division of labour between the Gypsies and craftsmen of Romanian or other origin organised into guilds. Meanwhile, Gypsies living in towns in Transylvania had a well-defined role in the life of the town. From the tax records of Braşov from the sixteenth century, for example, we learn that Gypsies were required to provide certain services to the town, such as repairing the gates of the town as well as its roads, the manufacture of weapons including cannons, keeping the streets clean, sweeping the market place and emptying sewers and latrines. They provided the town with its gravediggers, dogcatchers and executioners.81 In Sibiu, the Gypsies settled at the edge of town performed different tasks and services as ordered by the town and were responsible for taking letters to their destination. From an order issued by Matthias Corvinus in 1487, we find that the Gypsies of Sibiu were obliged to carry out certain tasks to help with the defence of the town.82 In some Transylvanian towns, over time it happened that the activities of the Gypsies came to affect the craftsmen who belonged to the town guilds. In Braşov in the years 1685–86, it was forbidden for the Gypsies to own sheep and they were deprived of the right to engage in commerce. The only activities permitted to them were horse-trading, the manufacture of nails and minor repair work.83 Such measures undertaken to protect other craftsmen from Gypsy craftsmen were not, however, introduced everywhere. In 1689, 79  N. Grigoraş, op. cit., pp. 68–69. 80  Aurora Ilieş, “Ştiri în legătură cu exploatarea sării în Ţara Românească până în veacul al XVIII-lea”, Studii şi materiale de istorie medie, I (1956), p. 180ff. 81  Maja Philippi, op. cit., pp. 143–145. 82  A. Gebora, op. cit., pp. 13–14. 83  Ibid., p. 29.

The Gypsies in the Romanian Lands During the Middle Ages

1009

the Transylvanian Diet introduced a new tax for Gypsy craftsmen, more specifically for Gypsy blacksmiths working with their own tools and for Gypsies working as gold-washers. They were required to pay fifty dinars per person (in the case of gold-washers, the sum was added to a special tax that they paid).84 Skilled slaves were very much in demand by princes, boyars and the monasteries. Since the Gypsies had a natural predisposition for craftsmanship, certain of their masters actually attached them as apprentices to master craftsmen. The Gypsies did not only practice crafts where they were regarded as having a kind of monopoly. In the last years of slavery, the boyars even used slaves in the new professions that were appearing at the time. Young Gypsies became proficient in their profession from a master craftsman, either on an estate or in the town. They became the cheapest craftsmen at the disposal of their owners. Similarly, the Gypsies also provided labour for private manufacturing enterprises in Moldavia and Wallachia. In the nineteenth century, we find that the activity of slave craftsmen is regulated and that the slaves are incorporated into guilds that were already in existence at the time. Gypsy craftsmen occupied a well-defined role in the economy of the principalities. Without doubt they constituted a source of wealth for the country. It was undoubtedly at a later stage that Gypsies took up occupations in agriculture. Moreover, these occupations never played an important role. It has been maintained that on the boyars’ estates, alongside the slaves employed at the boyar’s residence, there was also a category of agricultural slaves used in the fields and in the rearing of cattle that existed from the fifteenth century.85 However, it has been demonstrated that the only document that can be cited in this sense—a document from 1480 in which there appears the village “Goleşti and its Gypsies” (sixteen families)86—does not justify such a conclusion.87 The appearance in the sixteenth century of the large feudal estates specialised in the production of grain destined for Istanbul led to the use of Gypsy slaves in various forms in this kind of work, especially where there were acute labour shortages. Miron Costin states that rich inhabitants (boyars) of lower Moldavia worked the land with “purchased Gypsies”.88 In Transylvania, Giovanandrea Gromo shows that in 1564 the Szeklers were using Gypsies to work the land.89 84  Ibid. 85  V. Costăchel, P. P. Panaitescu, A. Cazacu, op. cit., pp. 152–153. 86  DRH, B, I, p. 275. 87  D. C. Giurescu, Ţara Românească în secolele XIV–XV, Bucharest, 1973, pp. 259–260. 88  Miron Costin, Istoria în versuri polone despre Moldova şi Ţara Românească (1684), translated by P. P. Panaitescu, Bucharest, 1929, p. 108. 89   Călători străaini despre ţările române, II, p. 322.

1010

Achim

Generally speaking, on estates that owned Gypsies, the Gypsies made up an auxiliary labour force. Neither the Gypsy slaves on the estates of the boyars and the monasteries in the principalities, nor the Gypsy serfs on the nobles’ estates in Transylvania were farmers in the medieval sense of the term, that is to say the owners of a piece of land which they cultivated under a particular regime of obligations to a master and to the State. It was the peasantry who worked exclusively as autonomous farmers. Until their de facto assimilation into the peasantry, which took place later on, even in the first half of the nineteenth century, in most cases the Gypsies were used for certain agricultural tasks that did not require any particular skill. In Wallachia and Moldavia, the boyars and the monasteries did not consider the need for a rational and economically profitable utilisation of their slaves until the era of emancipation. Only then did the slaves, forced to adopt a sedentary way of life, find themselves required to embrace agricultural occupations. Previously, the exploitation of the majority of the Gypsies had taken place through the dajdie (tax) they were required to pay to their master, with the Gypsies themselves earning their existence on their own, wandering the country and plying their trades. In the Romanian principalities, slaves provided the cheapest and most reliable labour force, their legal status tying them to the estate and their owner. For their masters, slaves constituted a source of income. For that reason they were highly sought after. Until the middle of the eighteenth century, the boyar class was engaged in a veritable race to obtain slaves. Boyars would buy slaves whenever the opportunity presented itself.90 The importance of their economic function also resulted from the high prices for which they were traded. An even more important reason for the high price of slaves was the prestige attached to the owning of a large number of slaves in medieval Romanian society. If private slaves generated an income for their master through the labour they provided or through the dajdie they paid, princely slaves paid tax to the state and were also subjected to other obligations. Since their numbers were fairly high, they generated substantial revenues for the Crown. In 1810, the revenue that the State generated from the 3427 families of princely slaves recorded in Wallachia totalled 700,000 thalers, 3000 gold ounces and countless payments in kind.91 In Moldavia in 1810, the Crown owned 1878 families of Gypsies. Prior to the Russian occupation of 1806, they generated annual revenues of 25,000 lei. In 1810, after several years’ experience of hiring out Gypsies to private

90  N. Grigoraş, op. cit., (I), p. 62. 91  T. G. Bulat, “Dregătoria armăşii şi ţiganii la sfârşitul veacului al XVIII”, Arhivele Basarabiei, VIII, 1936, no. 1, p. 6ff.

The Gypsies in the Romanian Lands During the Middle Ages

1011

persons, who paid up to 125,000 lei for them, obligations were established for the Gypsies whose total value was equal to the latter sum.92 The gold-washers formed a separate group among the Gypsies. They were skilled in the collection of gold from riverbeds rich in gold-bearing alluvia and diluvia. The technique they used consisted of washing the gold-bearing sands, which led to the Gypsies being given the name of “gold-washers” in Transylvania (in official documents in Latin the term aurilotores is used, while documents in German use the term Goldwäscher). In Moldavia and Wallachia, they were called aurari and rudari.93 The occupation was undoubtedly learned from the Romanians. Even though gold-washers were above all Gypsies, there were also Romanians who were skilled in the washing of gold. In Transylvania, in years of poor harvest, some needy peasants who lived close to areas with gold deposits occasionally practised this occupation. The entirety of the gold collected in Moldavia, the majority of the gold obtained in Wallachia (there were also Romanian aurari in the counties of Vâlcea and Argeş) and a substantial part of the gold of Transylvania were provided by the labours of this population. According to statistics from the year 1813, eight to ten măji of gold were collected from the rivers of Transylvania, while in 1837, seven to eight măji were collected.94 In Transylvania, there were two centres of gold washing, in the gold-bearing area of the Apuseni Mountains and in the area around the town of Sebeş. In 1781, in the Transylvanian principality there were 1291 Gypsy families registered as gold-washers.95 In the Banat in 1774 there were 84.5 families registered as being skilled in the washing of gold, a total of 244 people, while in 1801 there were 413 gold washers operating in

92  Idem, “Ţiganii domneşti, din Moldova, la 1810”, Arhivele Basarabiei, V (1933), no. 2, pp. 1–6. The total revenues of the State in 1810 came to 2,411,464.48 lei (idem, “Veniturile şi cheltuielile Ţărilor Româneşti între 1810–12”, Arhivele Basarabiei, VI (1934), pp. 135–136), which means that princely Gypsies contributed 5.18 per cent of revenues. 93  For the aurari and their occupation, see in particular C. Şerban, “Contribuţiuni la istoria meşteşugarilor din Ţara Românească: ţiganii rudari în secolele XVII–XVIII”, Studii. Revistă de istorie, XII (1959), no. 2, pp. 131–147; M. Acker, “Vechile spălătorii de aur din jurul Sebeşului”, Apulum, V (1965), pp. 647–658; C. Feneşan, “Date privind exploatarea aurului în Banat la sfârşitul sec. al XVIII-lea şi începutul sec. al XIX-lea”, Studia Universitatis Babeş—Bolyai, Series Historia, Fasciculus 1, 1967, pp. 55–64. See also H. Wilsdorf, “Zigeuner auf den karpathobalkanischen Bergravieren—montan-ethnographische Aspekte”, Abhandlungen und Berichte des Staatlichen Museums für Völkerkunde Dresden, 41 (1984), pp. 138–173; I. Chelcea, op. cit., pp. 143–149. 94  M. Acker, op. cit., p. 656. One majă = 56 kg. 95  See below p. 77 with note 160.

1012

Achim

thirty-eight villages recorded in the whole province.96 In Moldavia and Wallachia, such Gypsies constituted their own special guild of aurari or rudari. They were all princely Gypsies, totalling several hundred in each of the two principalities, and generated substantial revenues for the Crown.97 In 1810 in Wallachia, the principality’s treasury recorded a hundred Romanian gold miners and 400 Gypsy gold-washers.98 The authorities showed a special interest towards the Gypsy gold-washers. The revenues generated by the latter were substantial and were made up on the one hand by the tax that the Gypsies paid to the State and on the other hand by the gold that they were obliged to hand over to the tax office at the official price. The State, which had an interest in the continued existence of this group, protected them and instituted a privileged regime for them, under which they were exempted from any public tasks and sometimes also from certain obligations that they were otherwise obliged to provide to the owner of the estate on which they dwelt. In Transylvania from 1747 to 1832, gold-washing Gypsies were organised into their own voivodeship and kept separate from the other Gypsies. In Wallachia and Moldavia there was an official who dealt specially with the gold-washing Gypsies. In the last decades of the eighteenth century and at the beginning of the nineteenth century, both the Habsburg authorities in Transylvania and the Banat and the authorities in Wallachia and Moldavia undertook measures to regulate the activity of the gold-washers, establishing their status and their obligations to the State in a more organised manner, as well as the exemptions granted to them. These Gypsies led a nomadic way of life. The collection of gold could be carried out only in the warm season. Times of rain and flooding were particularly favourable to gold washing. The washers themselves lived in tents. In autumn and winter, they would as a rule move their tents to the plain or to some other suitable place where they could carry out other activities to earn their existence, such as the production and selling of pots and other objects made of wood. However, the washing of gold was not always a profitable activity. Contemporary reports illustrate the poverty in which these Gypsies lived. In spite of appearances, neither was the income they generated for the tax authorities always large. During the nineteenth century, the occupation entered into a decline until it virtually died out, in line with the dwindling of the gold deposits and the gradual settling of those Gypsies who practised it into a sedentary way of life. 96  C. Feneşan, op. cit., pp. 59–61. 97  See C. Şerban, op. cit., pp. 131–137. 98  T. G. Bulat, “Dregătoria armăşii şi ţiganii”, p. 6.

The Gypsies in the Romanian Lands During the Middle Ages

6

1013

Way of Life: Nomadism and Sedentarisation, Marginality

During the Middle Ages, on the Romanian territory, the Gypsies had their own habitat and social organisation, which differed from those of the Romanian population. The most distinctive feature was the nomadism of the Gypsies. Generally speaking, from the first attestations of the Gypsies in the fourteenth century and also later on, the Gypsies are referred to as living in sălaşe. For example, the Tismana monastery owned forty sălaşe of Gypsies, the Cozia monastery 300 sălaşe etc. Today, the term sălaş (plural: sălaşe) is usually translated in historical studies as “family”. In the Middle Ages, however, the term meant “tent”. In Latin texts, Gypsies living in these family units are referred to as Ciganus tentoriatus (tent-dwelling Gypsy), clearly expressing the contemporary understanding of the term. This understanding of the term as an improvised and mobile shelter exists even today in some regions of the country. The word is of Turkic origin, and can be found in the neighbouring Slavonic languages, as well as in Hungarian. It is an inheritance from the peoples of the steppe, who dwelt in this type of accommodation. Gypsy sălaşe as referred to in Romanian medieval documents were tents in which dwelt one or more families of Gypsies.99 The inhabitation of these mobile shelters was linked to the nomadic way of life of the Gypsies. In the first centuries of the Gypsies’ residence on the Romanian territories, almost all Gypsies led a nomadic way of life. The condition of slavery and belonging to a master, whether to the Crown or to a monastery or boyar, were not obstacles to the continuation of this way of life. The occupations practised by the Gypsies made it possible for their nomadic lifestyle to survive for a long period of time. The Gypsies were not farmers. Until late on, official documents that make reference to the occupations of slaves and the observations of foreign travellers in the Romanian countries make a clear distinction between the majority population of the country and the Gypsies when it came to their respective occupations. Sometimes the occupations of the two populations are even presented in opposition to one another, i.e. the peasants are engaged in agriculture (working the land and rearing animals), while the Gypsies practise certain professions. The use of Gypsy slaves to work the land and their consequent transition to a sedentary way of life belongs to a more recent era. Even as late as the second half of the eighteenth century, the vătraşi (sedentary) Gypsies, be they manor or estate Gypsies, were still few in number, not only among the ranks of this population but also among the slaves of most boyars and monasteries. Wherever there are 99  I. R. Mircea, op. cit., pp. 863–864.

1014

Achim

records of slaves on an estate, in most cases we find this kind of situation. A substantial proportion of private slaves were left to “feed themselves” by travelling the country, practising “their craft”. In return for this freedom, they paid their master a sum of money, the dajdie.100 Until the era of emancipation, most Gypsies were used by the monasteries and the boyars in such a way. As for the princely slaves, with the exception of the few slaves who worked at the princely court, almost all of them wandered the country in search of means of making a living, practising the occupations and crafts. Nomadic Gypsies, whether they were princely, monastery or boyars’ slaves, travelled the country in bands made up of a varying number of tents, carrying all their possession with them. They would set up their tents on the edge of a locality, where for a time they would practise their crafts, or, as in the case of the goldwashers and spoon-makers, who worked in isolated locations, they would set up their tents by rivers or in forests. Their horses and asses would graze on the land around the Gypsy encampment. In cases where blacksmiths, farriers, coppersmiths, locksmiths etc. were engaged in a constant movement from place to place in order to obtain work, Gypsies of this kind would practise seasonal migration, working and living in mountainous areas or in forests during the summer and descending to the plains for the winter, staying on an estate prepared to accept them. However, the nomadism of the Gypsies in the Middle Ages in the Romanian lands should not be understood in the strict sense of the term. Gypsies we call “nomads” lived on the estate of their master in winter (bands of princely Gypsies could, if they received the necessary permission, sojourn on the estates of monasteries or boyars), while in summer they wandered the land to earn their existence. In general, they would follow the same routes year after year. Gold-washers would stay in specific locations where they would collect gold, to which they would return every year. This seasonal migration was made possible by the social status and regimes of obligations of the Gypsies. Private slaves would return to their master’s estate in order to pay tax. The tax paid by princely Gypsies was collected by their leaders and handed over to state officials who were specifically responsible for the supervision of these slaves. Both private slaves and bands of princely slaves could equally be summoned by their master in certain situations or in order to carry out certain services. In Transylvania, twice a year, on Saint George’s day and Saint Michael’s day, bands of Gypsies would present themselves at the county seat to pay their tax. 100  N. Grigoraş,, op. cit., pp. 72–73.

The Gypsies in the Romanian Lands During the Middle Ages

1015

Therefore, it can be said that a limited and controlled nomadism existed. Of course, this nomadic way of life formed a contrast with the sedentary character of Romanian society, but it was not destructive or dangerous in nature. The way of life of the majority of the Gypsies was controlled by the public authorities and regulated in many ways. There was no conflict between the sedentary way of life of the autochthonous population and the nomadism of the Gypsies. In the Middle Ages, Romanian territories were sparsely populated. Until the nineteenth century, unoccupied lands were continually being populated with new settlements, often under the aegis of state policy. There was thus also room for the nomadism of the Gypsies in Romanian society. There was neither the demographic pressure nor the lack of spare land that existed in Central and Western Europe, which resulted in intolerance of the way of life of the Gypsies. Clearly, evolutions in the habitat of the Gypsies led to the gradual transition to a sedentary way of life. The domestic and agricultural occupations the Gypsies were with time forced to adopt tied them to stable settlement and a fixed dwelling. Settlements of Gypsy slaves appeared near some boyars’ residences and monasteries, where the servants and craftsmen of the master or even agricultural labourers worked. Such settlements were known as ţigănii (singular: ţigănie). Later on, even farming settlements made up of Gypsy slaves were created. The great monasteries in particular led a policy of capitalisation of their estates through the labour of the slaves. In the eighteenth and nineteenth centuries (although we presume that this was also the case earlier on), Gypsy settlements existed on the estates of the great monasteries. It is known that the largest concentration of Gypsies existed on the lands of the monasteries. Such settlements were, however, also created on the boyars’ estates. The way of life of the Gypsies was different to that of the Romanian population. The organisation of the Gypsies was also different. The regime of obligations imposed on the Gypsies was different to that of free peasants or serfs. Distinctions between the two societies, Romanian rural society and the society of the Gypsies, were great for a long time. Peasants and Gypsies were involved in certain economic relations, but they constituted two separate communities that functioned on the basis of different rules. For a long time the separation between the two communities was quite clear-cut. In medieval Romanian society, Gypsies had an inferior social status. Due to their social condition, they were considered to be living outside of society and were treated with the greatest disdain. This state of affairs can be discerned from contemporary documents. An English traveller wrote at the beginning of the nineteenth century: “Although the Gypsies form such an integral part of

1016

Achim

the community, they are regarded with the greatest disdain by the rest, whose behaviour towards them is scarcely better than towards animals; a man could more easily bear being called ‘thief’ or something similar than ‘Gypsy’.”101 Two centuries earlier, Laurentius Toppeltinus indicated that people avoided the Gypsies, refusing to greet them or show any sign of respect when they met them.102 In any case, their contacts with them were limited. It is indicative that when a group of Gypsies settled (either voluntarily or by force) in a Romanian locality (or in a Hungarian, Szekler or Saxon locality in Transylvania), their presence was only accepted on the edge of the settlement. In the Romanian principalities, Gypsies were not buried together with the other inhabitants, even though they were Christians. Instead, they were buried in a separate cemetery.103 Sources show that the Gypsies were treated with suspicion, particularly for the petty thefts that they carried out. The Gypsies’ wandering through the countryside to practise their trades was also regarded as suspect. Similarly, Gypsies constituted part of the ranks of beggars, vagrants etc.104 Even if the Gypsies were not persecuted in Wallachia and Moldavia on the same scale as in the countries of Central and Western Europe, we sometimes find that the local authorities would deal with groups of Gypsies. For example, whenever the first signs of an outbreak of plague appeared in Bucharest, usually the first measure to be taken, under the instructions of the prince, was the expulsion of the Gypsies from the town.105 The Gypsies were suspecting of spreading the epidemic via their itinerant lifestyle and wretched living conditions. Over time the distance between the agricultural population and the Gypsies began to decrease. Historical developments resulted in the increased dependency of the peasantry. From the sixteenth century, subjugated peasants in the Romanian lands (rumâni, vecini, iobagi) were tied to the land. They could not move anywhere else without the permission of the master of the estate upon which they worked, while the feudal lord could make use of them as he pleased, including the right to sell individual serfs, thereby separating them from their families. In this respect, the subjugated peasants were reduced to 101  W. Wilkinson, An Account of the Principalities of Wallachia and Moldavia, London, 1820, p. 175. 102  L. Toppeltinus, op. cit., pp. 59–60. 103  D. Dan, Ţiganii din Bucovina, Czernovitz, 1892, p. 17. 104  Nicolaus Olahus describes in 1536 a band of Gypsies, established near Şimand (in the west of present-day Romania), which lived exclusively from begging (Călători străaini despre ţările române, I, pp. 499–450). 105  For such measures see V. A. Urechia, Istoria românilor, Bucharest, vol. VI, 1893, p. 761 (from the years 1794, 1795 and 1796); vol. X/1, 1900, pp. 949–950 (from the year 1813).

The Gypsies in the Romanian Lands During the Middle Ages

1017

a state similar to that of the slaves. Slavery and serfdom, although states of dependence, were, of course, not identical to one another. Unlike the slave, who, together with his wife and children, had to serve his master in any way his master required on a daily basis, the obligations of a Wallachian or Moldavian serf were, in accordance with common law and the official law, far fewer and meant in effect only a certain number of days per year working the lands, in addition to obligations of transportation, guard duty and payments in kind. Furthermore, the serf’s family was exempt from the aforementioned work. The regime of obligations imposed on the serfs continued to worsen until the middle of the eighteenth century, although in this respect a distinction had always existed between a serf and a slave. However, the fiscal burden, the tax and an increasing number of other obligations to the State in reality rendered the situation for Wallachian and Moldavian serfs just as tough as that of the slaves. The latter had the greater obligations to their master, but in exchange they were exempt from tax obligations. In the seventeenth century and the first half of the eighteenth century, the subjugated peasants became considered the full property of their master, who could use them for any kind of work and could sell them, separating parents from their children, just as in the case of slaves.106 To a certain extent slavery and serfdom became confused, such that when in the middle of the eighteenth century the status of the peasant was regulated via the granting of personal freedom, the lawmakers felt it necessary to distinguish explicitly between serfs and slaves. The 1749 deed abolishing serfdom in Moldavia issued by Constantin Mavrocordat established that serfs are not slaves “because only Gypsies can have the status of slaves, serving together with their wives and children their masters every day. Among the serfs, only the men serve their master, with only one person per household required to work in this way, all the male offspring of the aforementioned will work as his assistants: the women do not serve the master; nor are serfs subjugated like slaves, because a serf is a free villager who owns no land.”107 During the same time, the transition to a sedentary way of life that the Gypsies were undergoing and their becoming tied to agricultural occupations brought them closer to the peasantry. The monasteries were particularly active in using Gypsy slaves in the exploitation of their estates. The regime of obligations imposed on these slaves evolved in the direction of that of the subjugated peasants, being required to pay terrage and carry out days of corvee (clacă). At the end of the eighteenth century and the beginning of the nineteenth century, many monasteries regulated the regime of working obligations imposed 106   Instituţii feudale, p. 411. 107  Gh. I. Brătianu, op. cit., p. 415.

1018

Achim

on their slaves. In some places, these obligations were fixed at three days per week, whilst in other places Gypsies were required to work a week out of every three for their masters. In other places, there were fewer obligations: sometimes the amount of corvee was fixed at twenty-four days per year, the same as it was for serfs in Wallachia.108 The transition to a sedentary way of life led over time to the disappearance of the ethnic character of a part of the Gypsy population. Living alongside the Romanian peasants, possibly practising the same occupations as the latter, living in isolation from their traditional way of life and constituting a small proportion of the entirety of the population, over the course of a few generations the Gypsies lost many of their cultural traits, chiefly their native language, but not only this. Mihail Kogălniceanu states that in 1837 the vătraşi, private Gypsies with fixed dwellings, remained Gypsies in name only, as they had completely forgotten their mother tongue and had lost the habits and customs of the Gypsies that were maintained only by the nomadic population. They could no longer be distinguished from Romanians.109 In 1800, the term “Gypsy” was in the first instance a social term, being an ethnic term only on a secondary level. Clearly, slaves who had been assimilated into the majority population were Gypsies only in the social sense of “slaves”, having already become Romanianised. When the Gypsies were emancipated, in the middle of the nineteenth century, some of those who benefited from emancipation (it is not possible to make an estimate of their number) had already lost their initial cultural traits, or at least the fundamental ones (i.e., language and customs) and had become Romanianised. After emancipation, their integration into Romanian society took place at a rapid pace. Changes of social nature affecting the peasantry and changes of habitat in the case of some Gypsies facilitated contacts between individuals belonging to the two populations and caused the old social barrier separating the Gypsies from peasants to fall in part. In this context, mixed marriages between Gypsies and Romanians appeared. In the seventeenth and eighteenth centuries, mixed marriages were not a rare phenomenon. For feudal masters they were a means of increasing the number of slaves. Enslavement through marriage seems to have been a widely used procedure, especially on the estates of the monasteries. In such situations, freemen would hand letters to the abbots declaring that they accepted the situation freely, that they would carry out the same work as other slaves and that they would never seek to regain their former legal status

108  Olga Cicanci, op. cit., pp. 160–162. 109  M. Kogălniceanu, op. cit., p. 575.

The Gypsies in the Romanian Lands During the Middle Ages

1019

via a tribunal. Likewise, the children resulting from such marriages would also remain slaves.110 The phenomenon whereby a number of peasants entered the ranks of the slaves is too large to be explained only by mixed marriages. The system of tax exemptions that existed in the Romanian principalities, according to which even private slaves were exempt from public duties, and the boyars’ and especially the monasteries’ pursuit of a labour force that was exempt from such public duties led to situations in which some feudal masters would declare certain of the peasants tied to their estates to be “Gypsies” in censuses organised periodically by the State to determine the number of taxpayers present on its territory. In this way, serfs were included among the ranks of those exempt from obligations to the State. The Austrian authorities noted at the beginning of their rule in Bukovina that on the estates of the monasteries lived a category of “Gypsies” who kept clean houses and who were well dressed etc., who were in fact peasants recorded in the monasteries records as slaves in order not to pay tax on them.111 Certainly, the number of Romanians in this situation was not great, but we should bear such arrangements in mind when we find that the censuses include some Romanians among the “Gypsies”. There is no doubt that in the eighteenth century there were individuals and families of Romanians who entered the ranks of slaves. We can speculate as to how many of the slaves lacking Gypsy characteristics referred to by Kogălniceanu were in fact Romanian peasants who through marriage or tax evasion had been added to the ranks of boyar or monastery slaves. At the same time, we can observe that through the medium of marriage an ethnic mixing of Gypsy slaves and Romanian peasants took place. When situations of this kind became relatively numerous, the authorities intervened to forbid mixed marriages. The Sobornicescul hrisov of 1785 forbade even Gypsies who had been released from slavery to marry Romanians; only the second generation of such Gypsies were allowed to marry Romanians.112 The ethnic mixing of Gypsies and Romanians also occurred as a result of freed Gypsies entering the ranks of freemen. They became members of rural and urban communities, were considered to be Romanians and quickly became assimilated in all respects. This phenomenon was, however, a minor

110  N. Grigoraş, op. cit., pp. 52–53. 111  R. Fr. Kaindl, Das Unterhanswesen in der Bukowina. Ein Beitrag zur Geschichte des Bauernstandes und seiner Befreiung (extract from Archiv für Österreich Geschichte, LXXXVI), Vienna, 1899, pp. 39–40. 112  T. Codrescu, loc. cit.

1020

Achim

aspect of the ethnic contacts between Gypsies and Romanians in the Middle Ages. Boyars very rarely freed their slaves. During the lengthy period of time when the Gypsies were slaves, the ethnic mixing between Gypsies and Romanians did not become a major phenomenon. The legal and social distinction and the different way of life meant that only in exceptional circumstances did such situations occur.113 Only in the nineteenth century, with the advent of emancipation did it become possible for a part of the Gypsy population, more specifically the former slaves who had changed their way of life, becoming peasants or craftsmen in villages or towns, to become fully integrated into Romanian communities. In the second half of the nineteenth century and during the twentieth century, the Romanianisation of a relatively significant part of the Gypsy population took place, which was noted by sociological studies and confirmed by demographic statistics. The process whereby a part of the Gypsy population lost its ethnic identity occurred throughout the entire country, not just in the principalities. In areas inhabited chiefly by other ethnic groups, the Gypsies adopted the language spoken there and were sometimes assimilated into the respective population. In Transylvania, sometimes even in the Middle Ages but to a greater extent in the nineteenth century, certain groups of Gypsies became Magyarised or Germanised. In Dobrogea, some Gypsies became completely integrated into Turkish or Tatar communities. 7

Social Organisation: The Leaders of the Gypsies

The Gypsies living in the Romanian states during the Middle Ages constituted their own microcosm that was separate from the local society. They also had their own social organisation. Contemporary written sources, interested almost entirely in the possessions of the Gypsies and their obligations to the State, paid no attention to the internal organisation of this population. Consequently information on this subject is scarce. On the other hand, the way in which communities of nomadic Gypsies were organised in the modern era 113  As for the rudari—who refuse to be known as Gypsies, considering themselves to be Romanians and speaking no language apart from Romanian, who distinguish themselves from other Gypsies through their customs and character—it is supposed that they are the result of a mixing of Gypsies and Romanians that took place some time during the Middle Ages (see I. Chelcea, op. cit., pp. 57–59). In our opinion, there are no arguments that would support this hypothesis. It is difficult to explain when and in what circumstances such a mixing could have taken place.

The Gypsies in the Romanian Lands During the Middle Ages

1021

is known. There is large body of evidence in this sense, which may be used as an indicator for the situation in earlier times. Ethnological studies provide us with interesting information on this subject. The most comprehensive study of this kind was carried out in the Romanian lands by Heinrich von Wlislocki at the end of the nineteenth century. He studied many different aspects of the life of tent-dwelling Gypsies in Transylvania and his writings are perhaps the richest in information of the social organisation of bands of Gypsies.114 The method used by the Transylvanian ethnologist is, however, questionable, and we may wonder to what extent data collected at the end of the nineteenth century is also valid for previous centuries. However, it is certain that some of the results produced by von Wlislocki are of use to historians. It was nomadic Gypsies who best maintained the cultural specificity of the Gypsies, preserving it almost until the present day. As a result of many factors, starting with the type of habitat in which they dwelt, nomadic Gypsies were far less affected by the impact of European feudal society than other Gypsies. The form of organisation of nomadic Gypsies displayed a remarkable longevity. The social evolutions that took place in Europe from the Gypsies’ arrival there until the modern era had little effect on the nomads. It is chiefly this category of the Gypsies that is meant when we consider the organisation of the Gypsies in the Middle Ages. Until the modern era most Gypsies were nomadic. The transition of a community to a sedentary way of life would lead sooner or later to its dissolution. The study of the social organisation of the Gypsies in the Romanian principalities during the Middle Ages will by its nature bring out both elements of specificity and continuity in the organisation of the Gypsies and the impact of local society and political factors on their organisation.115 The Gypsies lived in groups. The smallest of these groups was the family. Official documents from Wallachia and Moldavia used the term sălaş, while in Transylvania the term “tent” was often used. The family was composed of a Gypsy man, a Gypsy woman and their children. Thus, the form taken by the Gypsy family did not differ from that of the autochthonous population. 114  Particularly Vom wandernden Zigeunervolke. Bilder aus dem Leben der Siebenbürger Zigeuner. Geschichtliches, Ethnologisches, Sprache und Poesie, Hamburg, 1890, pp. 49–82; “Die Stamm- und Familienverhältnisse der transilvanischen Zelt-Zigeuner”, Globus, Braunschweig, 53 (1888), pp. 183–189; “Beiträge zu den Stammesverhältnissen der siebenbürgischen Zigeuner”, Ethnologische Mitteilungen aus Ungarn, I (1889), pp. 336–338. 115  Concise data on the organisation of the Gypsies in the Romanian lands: N. Grigoraş, op. cit., (II), pp. 52–58; G. Potra, Contribuţiuni, pp. 71–74; Instituţii feudale, pp. 412–413. For Transylvania see below p. 62–63 with note 124.

1022

Achim

Several families living in the same place made up a band. In Transylvania, this unit of organisation was usually referred to as a company. Generally speaking, the number of families that made up a band of Gypsies was small. From the relatively numerous statistics at our disposal from the second half of the eighteenth century, it emerges that a band was usually made up of between thirty and forty families. Bands were usually composed according to profession: gold-washers (aurari), bear-baiters (ursari), musicians (lăutari), spoon-makers (lingurari) etc. The whole band moved together as it travelled the countryside. It possessed a number of characteristics that distinguished it from other bands. Each band had its own leader, who was known as jude or giude in Wallachia and Moldavia and voivode in Transylvania. The Gypsies chose the leaders of the bands during an assembly attended by the entire group, which was carried out according to a certain ritual. Those chosen to become leaders were selected from among the men considered to be the strongest or wisest. The function of leader was held for life, but it was not hereditary. These leaders constituted a kind of aristocracy of the Gypsies. The voivode or jude became the all-powerful head of that particular community of Gypsies. He enjoyed the complete obedience of the band. Contemporary documents are able to tell us something of the prerogatives of the leader of the band. The leader’s first duty was the power to resolve disputes between members of the band. He had the power to pass judgement and to administer punishment. This aspect is revealed to us by certain grants issued by Hungarian kings and Transylvanian princes in the fifteenth and sixteenth centuries to bands of Gypsies. In the grant of 1422, King Sigismund grants (in fact confirms) the voivode Vladislav facultas judicandi over his Gypsy subjects (Ciganos sibi sujectos).116 The administering of justice among the Gypsies was the sole responsibility of their leader. Evidently, this is the case only for disputes within the community. In such cases the local administrative authorities had no prerogative. According to the verdict of 1476 given by King Matthias Corvinus in connection with the Gypsies living at the edge of the town of Sibiu, the voivode and vice-voivode are explicitly forbidden from intervening in the administering of justice among the Gypsies.117 However, cases where the Gypsies entered into legal dispute with somebody from outside the band came under the jurisdiction of the state legal system. In Wallachia and Moldavia, there is evidence of the legal autonomy of the bands of Gypsies only from the end of the eighteenth century and the beginning of the nineteenth century, when the Crown, concerned to halt the decline in the numbers of Gypsies in its 116  Hurmuzaki, 1/2, p. 527. 117   Urkundenbuch zur Geschichte der Deutschen in Siebenbürgen, VII, p. 113.

The Gypsies in the Romanian Lands During the Middle Ages

1023

possession, confirmed the rights of different bands that had been held in ancient times, but which had been violated. The charter of 25 March 1793 issued by Mihail Şuţu regarding ursari and lingurari Gypsies in Moldavia renewed their right to have problems that arose within the band resolved solely by their leader: “any legal dispute or implementation of legal judgements will be dealt with by them alone, according to their ancient custom, only their headmen have the right to try them. Prefects and other officials are not to interfere in their affairs, except in cases where the death of a man is involved.” The authorities intervened only in cases where the Gypsies entered into legal disputes with other inhabitants.118 This is the most important aspect of the autonomy enjoyed by communities of nomadic Gypsies, both in Transylvania and medieval Hungary and in Wallachia and Moldavia. Evidently, the judgements handed down by Gypsy leaders within their communities were given in accordance of the unwritten rules of that population. The second important duty of the Gypsy leader was the collection of taxes that the community as a whole or Gypsies individually owed to the State, the local authorities or in certain cases to the feudal master of the estate on which the band was staying. For fulfilling this task, the Gypsy leader was exempt from the payment of tax and other obligations. In this way, the Gypsy prince played the role of intermediary between the community and the authorities. We can, therefore, see that the Gypsy band had its own internal autonomy. There were no significant differences between the three Romanian principalities in this respect. Thus was the Gypsy community. In exceptional circumstances, two or more bands linked by ties of kinship or occupation might live together. However, there never existed an original master community of the Gypsies. From the beginning, there was no organisational structure of the Gypsies that included all Gypsies living on a region-wide or country-wide level. Organisational structures of this kind appeared later on. They were created by the State for fiscal reasons or in order to exert a more efficient control over the Gypsies. In Wallachia and Moldavia, several Gypsy bands living within a certain region and sharing the same occupation were placed under the authority of a Gypsy sheriff (vătaf). Together, the bands composed a Gypsy shire (vătăşie). The Gypsy sheriff was himself a Gypsy. From the eighteenth century he began to become known as the bulucbaşa or bulibaşa. The sheriff or bulibaşa was the head of a number of leaders from a particular region who also belonged to the same clan. As illustrated by a deed of 1753 issued to Iancul, he was appointed 118  T. Codrescu, op. cit., pp. 286–287. These privileges were renewed in 1804 (G. Potra, Contribuţiuni, pp. 327–331).

1024

Achim

bulucbaş over the leaders of the princely lingurar Gypsies of lower Moldavia by the prince of the principality. The document makes reference to the functions of this particular sheriff: to watch over all the leaders and their bands and to round up and bring back, together with the band leaders, Gypsies who had wandered away from their bands and entered other bands, whether princely, boyar or monastery bands.119 The vătaf or bulibaşa collected tax, presided over disputes between bands or between Gypsies from different bands and sometimes even in disputes between members of the same band. The vătaf would also represent the Gypsies under his authority before the official responsible for the Gypsies and before the authorities. Vătafi and bulibaşi were exempt from the payment of tax and from other obligations to the State. Such organisational structures were creations of the State. They were set up on a territorial and occupational basis and had nothing to do with the original social organisation of the Gypsies. With regard to the latter, it is possible that there may have existed some sort of tribal organisation of the Gypsies, but in the Romanian lands there is no evidence to confirm this. In any case, for Romologists, the extent to which a tribal organisation of the Gypsies actually existed is a question that remains unanswered. In Wallachia and Moldavia, in parallel with the sheriffs, a network of officials was set up to monitor princely Gypsies from a particular region, all Gypsies belonging to a particular clan within the whole country or a part of them. In the first document attesting to their existence, from Wallachia from 1458, these officials were given the name cnezi of Gypsies.120 Later on, these officials were known as sheriffs (vătafi) of Gypsies or vornici of Gypsies. At the beginning of the nineteenth century in Wallachia, they were known as zapcii of Gypsies and there were four of them appointed for the entire principality: one responsible for the counties west of the river Olt, two for the counties east of the Olt and one responsible only for aurari.121 Their responsibilities were chiefly fiscal in nature. In order to fulfil their responsibilities, they collaborated with the Gypsy bulibaşi and leaders. They were not subordinate to the administrative authorities, because the authorities did not have responsibility for the Gypsies. There was, therefore, a doubling of the authorities at the level of Gypsy bands. Alongside the Gypsy leaders drawn from the ranks of the Gypsy communities, the authorities also appointed officials in charge of the administering state authority over this population. The latter were not Gypsies. Some were low-ranking boyars. They carried out a service for which they were paid 119  G. Potra, Contribuţiuni, pp. 72–73. 120  DRH, B, I, pp. 201–202; 121  T. G. Bulat, “Dregătoria armăşii şi ţiganii”, p. 7.

The Gypsies in the Romanian Lands During the Middle Ages

1025

out of a certain quota from the taxes gathered from the Gypsies. Confusion can appear because at the time both Gypsy leaders and the state officials in charge with problems related to the Gypsies were known as sheriffs (vătafi). In Transylvania in the eighteenth century, there was a voivodeship of the Gypsies of the Făgăraş land. The prince of Transylvania, who had in its possession the Făgăraş estate, would appoint to this office a boyar from the region. That is how the situation appears in a deed of 1679, when Şerban Comşuţ was appointed to this office; the prince was responsible for the organisation of the Gypsy labour force, the collection of taxes and other contributions and for the fining of lawbreakers.122 A high official was responsible for the supervision of all princely slaves within a principality. In Wallachia, this function fell to the great armaş, while in Moldavia from the eighteenth century the function was appointed to the hetman (hatman). For this reason, princely Gypsies were also known as “the hetman’s Gypsies”. In the nineteenth century in Moldavia, the official responsible for the superior supervision of the Gypsies was given the title of nazâr or epistat over the princely Gypsies.123 In Transylvania during the sixteenth century, in the period of the autonomous principality, the function of voivode of the Gypsies of all the country was established.124 He was appointed by the prince from the ranks of the nobles of the principality. The first, succinct, reference to this function that has come down to us dates from 1541, when the general commander of Transylvania, Balthazar Bornemisza, confers the position of voivode of the Gypsies of all Transylvania to Mathias Nagy and Thomas of Aiud (Enyed).125 There is also the grant with which Queen Isabella grants the voivodeship of the Gypsies to Gáspár Nagy and Franciscus Balásfi, two nobles with close links to the palace. This document, together with two other documents also from 1557 relating to the voivodeship of the Gypsies, enables us to learn something of the responsibilities of the voivodes. They had authority over all the Gypsies living in the country, while the leaders of the Gypsy bands were subordinate to them. They were responsible for the collection of the taxes the Gypsies owed to the State. They were equally responsible for issuing and collecting fines. The two voivodes enjoyed complete freedom in the exercise of their duties. The leaders of towns, market towns and villages were ordered to give them their full 122  D. Prodan, Iobăgia în Transilvania în secolul al XVII-lea, vol. I, Bucharest, 1986, p. 110. 123   Instituţii feudale, pp. 412–113. 124  See H. M. G. Grellman, op. cit., pp. 138–147; A. Gebora, op. cit., pp. 17–18, 28; D. Prodan, Iobăgia în Transilvania în secolul al XVII-lea, vol. I, pp. 109–110. 125  A. Gebora, op. cit., p. 17.

1026

Achim

co-operation, with those attempting to obstruct them liable for prosecution by the royal court.126 In 1588, the Diet of Transylvania decided to abolish the office as a result of the abuses perpetrated by the holders of the office. Such abuses were also an annoyance to those nobles who had Gypsies under their authority. The tax that the Gypsies were required to pay to this leader of theirs was abolished together with the office itself. At the same time, it was decided that the masters of the Gypsies would be free to decide whether they would collect taxes from their Gypsy serfs or not, and that in cases where the boyars had many of this kind of serf, they were required to appoint a voivode to lead them.127 Later on, the voivodeship of the Gypsies was re-established. The Diet of Transylvania returned to this office on a number of occasions during the seventeenth century. The functioning of the office was subject to strict regulation, as was that of the problem of the Gypsies as a whole.128 The last voivode of the Gypsies was Peter Vallon, appointed by Prince György Rákóczi I. The voivode of all Gypsies dealt with all problems concerning their relations with the State. The voievode’s duties were predominantly fiscal in nature, as well as legal and administrative. In essence, we are dealing with the delegation of the sovereign’s authority with respect to the Gypsies. The voivode was in fact an official charged with the administration of the Gypsies on a particular territory, in this case the principality of Transylvania. Such voivodes also existed in areas of Hungary that had not been occupied during this period by the Ottomans and that were then under the authority of the Habsburg Empire. In total there were four such voivodes; one had his seat in Satu Mare. This innovation existed not only in Transylvania and Hungary: it also existed in Poland in the seventeenth and eighteenth centuries, where the leaders of the Gypsies, chosen by the authorities from among the nobility, were given the title of “king”.129 These voivodes of the Gypsies should not be confused with the Gypsy leaders who also bore the name of voivode. This organisational structure did not exist for all privately owned Gypsies. It is known that there were bands of nomadic Gypsies (lăieşi, lingurari, aurari etc.) who belonged to private owners and were led by their own heads (juzi). We can, therefore, presuppose that they would have been organised in the same manner as the bands of princely Gypsies. Some of these Gypsies had, however, settled into a sedentary lifestyle. In such cases, for a period there 126  B. Mezey et al., op. cit., pp. 78–79. 127  D. Prodan, loc. cit. 128  B. Mezey et al., op. cit., pp. 71–74. 129  J. Ficowski, Wieviel Trauen und Wege. Zigeuner in Polen, Frankfurt am Main, Bern, New York, Paris, 1992, pp. 21–34.

The Gypsies in the Romanian Lands During the Middle Ages

1027

existed a certain form of communal living, with its own autonomous organisation. Especially on the estates of the monasteries there were numerous communities of slaves. Supervision of boyar and monastery slaves fell under the authority of one of the master’s stewards, as a rule known as the jude of the Gypsies. Here the prince’s officials could not interfere. In Transylvania in the seventeenth century, we find voivodes of the Gypsies actually present on some feudal estates.130 In such cases, we are of course dealing with Gypsy leaders settled on these estates as craftsmen or agricultural workers. Families of Gypsies separated from their band or even entire bands carried out certain services on some estates. To begin with, no relation of legal dependency existed for the Gypsies in relation to the feudal master. As they were royal serfs, their employment by a private landowner was carried out with the permission of the sovereign, as we earlier saw in the case of Sibiu in 1486. With time, however, the Gypsies became serfs of their respective master. Since the Gypsies remained a separate social and occupational group even after they had acquired their new social status, relations with their master were carried out via the intermediary of their leader, the voivode. It is for this reason that in Transylvania some sedentary Gypsies also had voivodes. As we can easily observe, all the terminology used to refer to the leaders of the Gypsies has been borrowed from elsewhere. The Gypsyheads bore names acquired from other peoples, and not Indian names. Jude and vătaf are terms that belong to the administrative vocabulary of medieval Romania. “Voivode” (Romanian voievod) is a term of Slavonic origin, used, however, by the Romanians and the Hungarians. The Gypsies adopted this term in SouthEastern Europe, probably on Romanian territory. In Transylvania and Hungary, the leaders of the Gypsies were thus known from the beginning. The Gypsies who reached Central and Western Europe did not use it. In Wallachia and Moldavia, the leaders of the Gypsies were never known as voivodes. Bulucbaşa or bulibaşa is a Turkish term. Since in the Romanian principalities its sense was strictly reserved for that of the head of the Gypsies, we can presuppose that it originated from the Ottoman Empire together with the groups of Gypsies that used it in the Romanian lands for the first time (and this at a late stage) in the eighteenth century. However, this borrowed terminology covers a specific organisational structure of the Gypsies which, with some differences, we meet almost everywhere in the European countries where the Gypsies were present from the Middle Ages. The political powers of the different countries recognised this autonomous organisation of the Gypsies, regardless of the social status that they held there. 130  D. Prodan, Iobăgia în Transilvania în secolul al XVII-lea, vol. I, p. 109.

1028

Achim

In Romanian historiography the opinion has been advanced that the organisation of the Gypsies presents indicators of the Romanians’ nature of organisation during the period of crystallisation of the Romanian states.131 Certainly, the Gypsies adopted a great deal from the peoples alongside which they have lived, including the Romanians. However, when it is a question of the organisation of the Gypsies in the Romanian principalities, it does not differ in essence from the situation among the Gypsies of Transylvania, Hungary, Poland or other states. The Gypsies had nothing to do with the political organisation of the Romanians in the period prior to the formation of the principalities. When the Gypsies arrived in the area north of the Danube, the principalities had already become a fact of history. The Gypsies arrived on Romanian soil with their own form of organisation, which was tolerated by the political authorities, which incorporated it into the fiscal and administrative system already present in the country. 8

The Situation of the Gypsies in the Romanian Lands and in Other European Countries—A Parallel

In the previous sub-chapters, we have examined the legal status and economic and social condition of the Gypsies in the Romanian lands during the Middle Ages. To what extent, however, do these realities correspond to the history of the Gypsies in other countries of Europe during the same period? In these considerations the slavery of the Gypsies, the nomadism of the majority of them, the economic functions they fulfilled, their relations with autochthonous populations and the social evolutions they experienced can be included. Such aspects are characteristic of several centuries of the history of this population in the Romanian lands, and which may or may not have reoccurred in other European countries. From the beginning of the fourteenth century until the first part of the fifteenth century, the Gypsies who had left the Byzantine Empire scattered in all directions, reaching virtually all the countries of Europe. Groups of Gypsies have managed to live in various countries, each beginning a separate chapter 131  “Through their current organisation, the Gypsies show to us how the Romanians were organised in the thirteenth century: their heads, the voivodes, wore their hair long, wore reed slippers, they held the juzi under their authority; the names chosen by the Romanian princes were adopted by their slaves: Vlad, Dan, Radu. They have even preserved certain linguistic archaisms.” (N. Iorga, Locul românilor în istoria universală, ed. R. Constantinescu, Bucharest, 1985, p. 128.)

The Gypsies in the Romanian Lands During the Middle Ages

1029

in the history of the Gypsies. Once the Gypsies leave the Balkans, we in fact are no longer dealing with a single history of the Gypsies, but rather many histories of Gypsies, based around countries and possibly around larger geographical areas. What unites all these histories is the Gypsies’ extraordinary ability to conserve their cultural identity and their obstinate refusal to adapt to the values of European civilisation and to give in to assimilation. It is in this respect alone that historians believe it possible to speak of a single history of the Gypsies in the medieval and modern eras. From the beginning it is necessary to state that the Gypsies’ status as slaves marks out their history in Romania from the general history of this population. Slavery existed throughout Europe in the Middle Ages, especially in the countries bordering the Mediterranean and Black Sea basins, but also in other countries. Legal documents from these countries and the history of trade in the Mediterranean are full of information regarding this social reality. However, generally speaking it was confined to the early Middle Ages. In the Byzantine Empire, where there was continuity from ancient times with regard to slavery, the institution acquired a minor, domestic role, leading to its eventual disappearance.132 In medieval Hungary, the institution of slavery, which affected certain populations of oriental or non-Christian origin, disappeared in the thirteenth century.133 In Russia, where in the time of the Kievan Rus slavery was an institution that affected relatively large numbers of people and which was strictly regulated, in the fourteenth and fifteenth centuries the holops were gradually freed and transformed into serfs.134 In the fifteenth century, domestic slaves more or less disappeared from the countries of Southern Europe (Italy, France, Spain) where they had previously been present. Trade in slaves originating from the east of the Mediterranean basin and from the Black Sea basin was no longer practised in the Christian countries.135 In all of these countries, as well as in other European countries, for as long as it existed, medieval slavery was domestic in nature. The use of slaves in production was extremely limited. To a large extent, slavery was dependent on wars, on the capture of non-Christian slaves, on piracy and on slave trading from the Mediterranean basin. At a time when medieval slavery was disappearing from Western Europe and Russia, the slavery of this population of Indian origin arriving from the

132  See Helga Köpstein, op. cit. 133  See Al. Gonţa, op. cit., pp. 307–311, together with the references in the notes. 134  B. D. Grekov, op. cit., pp. 141–163 and 560–577. 135  For this problem see Ch. Verlinden, L’Esclavage dans l’Europe médiévale, vols. I–II, BruggeGhent, 1955–1977.

1030

Achim

Balkan Peninsula was being established in the Romanian lands. In the countries of Central and Western Europe, the Gypsies never experienced slavery. Why did this difference of social and legal status occur? Slavery in the Romanian principalities did not exist as a result, as has been claimed, of the “heavy dependence of the Romanian principalities on the East, where slavery was even more widespread”.136 It is true that for centuries the history of the principalities was linked to the Ottoman Empire. However, even if there were slaves in the Ottoman Empire,137 living under a regime similar to that in the north of the Danube, the social regime in the Romanian principalities had nothing to do with the East. Romanian feudalism was European in nature, with certain characteristics common to the area of South-Eastern Europe. In the case of the slavery of the Gypsies, there is no question of the adoption of some Eastern institution. This is confirmed when one considers that slavery existed in the principalities before any contacts were made with the Ottoman Empire. As we have seen, the appearance of slavery in the Romanian principalities can be explained by the way in which the Romanians to the south and east of the Carpathians replaced the Tatar domination that had ruled there after 1241. The creation of the Romanian states was the result of a series of battles that applied the rule, at that time in general use in Eastern Europe, according to which prisoners of war were to be enslaved. The Gypsies that arrived in the Romanian principalities from the south of the Danube, added weight to the institution of slavery. Slavery formed part of the social system of the Romanian principalities, undergoing certain modifications over time before being abolished in the nineteenth century. It is worth bearing in mind that the Gypsies were also enslaved in other countries in South-Eastern Europe. Neither in the Byzantine Empire nor the Ottoman Empire were the Gypsies a free people: there the Gypsies were in a situation similar to that of the princely slaves in the Romanian principalities.138 The regime of obligations and the legal status imposed on the Gypsies of Corfu in the fifteenth century during Venetian rule—as they appear in a decree of 1470—amounted, formally speaking, to a regime of serfdom.139 However, the regime was also very close to the regime of slavery, as it existed in the Romanian principalities. The presence of the Gypsies in Corfu is attested to from the 136   Istoria României, ed. M. Roller, Bucharest, 1947, p. 385. 137  See H. Inalcik, “Servile Labor in the Ottoman Empire”, in The Mutual Effects of the Islamic and Judeo-Christian Worlds: The East European Pattern, eds. A. Ascher, T. Halasi-Kun and B. K. Király, Brooklyn N.Y., 1979, pp. 25–52. 138  See above p. 29 with note 9. 139  G. Soulis, op. cit., pp. 164–165; A. Fraser, op. cit., pp. 50–51.

The Gypsies in the Romanian Lands During the Middle Ages

1031

second half of the fourteenth century, before the island fell into the hands of the Venetians (1386), and it is certain that the Gypsies had arrived there earlier. In the case of Corfu, we are very likely to be dealing with state slaves from the Byzantine period organised into a fief (feudum acinganorum) by the Venetians. The situation of the Gypsies in 1470 was in fact a continuation of the status imposed on them in the Byzantine era. The Gypsies were also enslaved in the medieval states of Bulgaria and Serbia. Direct documentary evidence of this is lacking due to the political history of the region at the time, but we can presuppose that the status imposed on the Gypsies in these countries was similar to that in the Romanian principalities. Following the disappearance of the Christian states to the south of the Danube, the Romanian principalities remained the only countries in Europe where the Gypsies were enslaved. When the first Gypsies appeared in the Romanian principalities in the second half of the fourteenth century, the political conditions of the time (namely the battles with the Tatars and the Ottomans) meant that the sense of confrontation between Christians and non-Christians, between the sedentary Romanian population and the foreign nomads, was still fresh. Due to their characteristics, their way of life and their organisation into bands, the Gypsies were regarded as enemies and enslaved. Contact between the peoples of Central and Western Europe and the Gypsies had a different character. Here the Gypsies were not identified as enemies. During the Gypsies’ first penetration into Western Europe, which took place in the years 1416–19, interest and tolerance was even shown to them.140 Consequently, there was no question of enslaving these Gypsies. The main reason why these countries did not seek to enslave the nomadic Gypsies was that at the beginning of the fifteenth century when the Gypsies first appeared, social structures in Western Europe had already been well established, with slavery absent from these structures, even slavery of a patriarchal nature, which admittedly existed there until the start of the second millennium. Only in the Mediterranean countries was this kind of slavery maintained for a time, but at a more or less insignificant level. Even if slavery did not exist as an institution, when the peoples of Western Europe realised that the bands of nomadic Gypsies made a living from fortune telling, sorcery, theft etc., measures were soon taken against the new arrivals. The history of the Gypsies in Central and Western Europe was until the eighteenth century an endless series of persecutions, expulsions, deportations, pogroms etc. The population was constantly hostile towards the Gypsies, while in 140  See F. de Vaux de Foletier, op. cit., p. 42ff.

1032

Achim

many places public authorities had an overtly anti-Gypsy policy.141 The situation that the documents of Ragusa (Dubrovnik) indicate is quite unique. Here, in the fifteenth century, the inhabitants of Ragusa treated Gypsies who practiced various trades or who were engaged in commerce as equals. There was no prejudice and no persecution of them.142 This case was, however, exceptional. In the West, the Gypsies never found their place in society. They remained always outside society, regarded as a foreign element. The situation that existed in the Romanian principalities, in which the Gypsies were accepted as part of the social structure as a servile and inferior class and occupied a position, however marginal, in the economy, did not exist in the West. Consequently, even if the Gypsies, who represented the bottom rung of society, were treated with disdain, measures were never taken against them along the lines of those taken in the West. Why did this difference in the treatment of the Gypsies exist? It is possible that the answer lies in the different stages of development that existed in the Romanian principalities in comparison to the countries of Central and Western Europe. In the Romanian principalities, the quasi-agrarian nature of the economy and the small number of craftsmen either of local origin or originating from neighbouring countries meant that Gypsy blacksmiths and those practising other crafts were able to find a place in society, being in demand both by the owners of the great estates and by the agrarian population. Furthermore, the sparseness of the population made it possible for groups of Gypsies seeking to earn an existence to practise a limited and controlled form of nomadism. In Central and Western Europe, however, the situation was completely different. In the economic and social system in place in those countries there was no room for populations who did not lead a sedentary lifestyle. Crafts were strictly organised into guilds of closed nature, while production was monopolistic in character and was subject to the power of cartels. Access to this system for foreign craftsmen was completely out of the question. The 141  For anti-Gypsy measures and policies in Western countries in the Middle Ages and at the beginning of the modern era, see A. Colocci, op. cit., pp. 67–107; J-P. Clébert, op. cit., p. 83ff; F. de Vaux de Foletier, op. cit., p. 58ff; M. Hasenberger, “Die Zigeuner in Europa mit besonderer Berücksichtigung des mittel- und südosteuropäischen Raumes. Ein historischer Abriß über die Reaktionen zwischen Wirtsvölker und Zigeuner” (doctoral thesis), Vienna, 1983, p. 22ff; R. Stangl, “Die Verfolgung der Zigeuner im deutschsprachigen Raum Mitteleuropas von ihrer Anfängen bis heute” (final-year undergraduate dissertation), Vienna, 1986, p. 18ff; J. S. Hohmann, Verfolgte ohne Heimat. Geschichte der Zigeuner in Deutschland, Frankfurt am Main, Bern, New York, Paris, 1990, pp. 17–26; A. Fraser, op. cit., pp. 85–190. 142  Dj. Petrović, op. cit.

The Gypsies in the Romanian Lands During the Middle Ages

1033

Gypsies, with their professions and the rudimentary goods that they produced had no place within such a system. Consequently, in the West, the Gypsies did not work as craftsmen; instead they forced to orientate themselves towards marginal occupations such as rearing horses and performing music, or to make a living from fortune telling, counterfeiting money, theft etc. Already by the mid-fifteenth century they were regarded as a burden on the local population. They were forbidden access to towns and villages, they were driven out of some areas and other measures were taken against them. For the peoples of Western Europe the Gypsies were an unwanted group and were treated as such.143 The process by which some of the Gypsies living as slaves on the estates of the boyars and monasteries in Moldavia and Wallachia, as well as Transylvania, settled into a sedentary way of life was repeated in corresponding forms in other parts of Europe. Already in the fourteenth century there were sedentary Gypsies living in villages in the Byzantine Empire, married to Greek women and more or less converted to Christianity, thereby separated definitively from their original communities.144 In the second half of the fourteenth century and the first half of the following century, there are a number of descriptions that attest to the existence of sedentary Gypsies settled in the Peloponnese, the western part of continental Greece and the Ionian islands.145 The process of sedentarisation was on a larger scale in Moldavia and Wallachia compared to neighbouring countries. In the eighteenth and nineteenth centuries, the vast majority of the Gypsies were tied to agriculture, so that at the time of emancipation, only certain groups of Gypsies, especially those belonging to the State, still practised nomadism. The process of sedentarisation was a natural one, but one that had more important consequences than the policy of forced assimilation practised in Hungary, Transylvania and the Banat in the second half of the eighteenth century by Empress Maria Theresa and Emperor Joseph II.146 The more open nature of Romanian society, where in spite of slavery, social barriers could be more easily overcome, also contributed to this phenomenon. As there were Gypsies who managed to liberate themselves from slavery to become freemen, so there were also Romanians from the ranks of the bounded peasantry who, through inter-marriage with Gypsies, became slaves.

143  R. Stangl, op. cit., pp. 16–17. 144  Ilse Rochow, K-P. Matschke, op. cit., p. 253. 145  G. Soulis, op. cit., p. 151ff. 146  See following sub-chapter.

1034 9

Achim

The Policy of Sedentarisation and Assimilation of the Gypsies Promoted by the Habsburg Authorities in Transylvania

The second half of the eighteenth century in Transylvania was for the Gypsies the time of the first concerted effort to convert them to a sedentary lifestyle. As a result of the major political changes that occurred in the middle Danube basin at the end of the seventeenth century and the beginning of the eighteenth century, the Habsburg Empire established its authority in Transylvania in 1688 and in the Banat in 1718. Habsburg rule over these two provinces, as for Hungary and the other territories that for one or two centuries had been under the direct or indirect rule of the Ottoman Empire, resulted in social and institutional modernisation. Aside from the preservation of certain provincial particularities, major transformations were wrought, in accordance with the spirit of the time and the House of Habsburg’s interest in increasing the fiscal and military capacity of the Empire. The reigns of Empress Maria Theresa (1740–80) and her son, Emperor Joseph II (1780–90), were particularly marked by a major intervention on the part of the State in the social organisation of the provinces. This was the era of enlightened despotism. The directed reformism of the Habsburgs targeted every aspect of social organisation. Regulation of the status of the bounded peasantry, institutional modernisation, religious tolerance and commercial and population policies were just some of the aspects of the reformist policy of the Habsburgs. The Gypsies, who were relatively many in number in the eastern countries and provinces of the Habsburg Empire, for a long time remained outside the great socio-economic transformations that the Empire was undergoing. They continued to live as they had done in previous times. There were many Gypsies concentrated in particular within the Kingdom of Hungary (which at that time included the western regions of present day Romania) and the Great Principality of Transylvania. A list of the number of families according to their tax status from 1772 in the Great Pricipalitiy of Transylvania recorded 3769 families of settled Gypsies and 3949 families of nomadic Gypsies.147 This gives a total number of families of 7718, which, if we estimate that there were an average of five persons per family, gives us a total of 38,590 Gypsies. If we compare the number of families of Gypsies to the total number of families registered in Transylvania at the time (302,986), it can be concluded that the Gypsies accounted for 2.55 per cent of the population of the principality. 147  A. Răduţiu, Populaţia Transilvaniei în ajunul călătoriei din anul 1773 a lui Iosif al II-lea, in Sabin Manuilă. Istorie şi demografie. Studii privind societatea românească între secolele XVI–XX, coord. S. Bolovan, and I. Bolovan Cluj-Napoca, 1995, pp. 76–77.

The Gypsies in the Romanian Lands During the Middle Ages

1035

The conscription of the population in the principality of Transylvania of 1776 records 3562 families of settled Gypsies and 3867 families of nomadic Gypsies, giving a total of 7429 families.148 Compared to the 271,852 families in total living in Transylvania, the Gypsies accounted for 2.73 per cent of the principality’s population. In the Banat, according to figures provided by Francesco Griselini in 1780, in the cameral districts there were 5272 Gypsies, while in the military districts in the south-east of the province, there were approximately 2800, meaning that in the whole of the Banat there were approximately 8072 Gypsies. The cameral districts had a total population of 317,928, while the total population of the Banat was approximately 450,000.149 This meant that the Gypsies accounted for 1.6 per cent of the population of the Banat. Aside from the differences regarding the status of the Gypsies from place to place, the revenues a Gypsy generated for the State were generally speaking of a limited nature; in any case, smaller than those generated by a serf or a craftsman. Nor were they a source of soldiers. On the other hand, the Gypsies created problems for the authorities due to the nomadic lifestyle led by many of them and the crimes that they committed. It was in these circumstances that the Habsburg authorities sought to integrate this population into the social structure of the country. In previous centuries and even in the first half of the eighteenth century until the reign of Maria Theresa, the authorities paid little attention to the sedentarisation of the Gypsies. It is true that over time some Gypsies settled on the edges of towns or on certain nobles’ estates. However, only a part of the Gypsy population was affected by sedentarisation and in most cases, it did not mean complete integration into the economic and social life of the respective communities. The authorities showed no intention to integrate either these relatively settled Gypsies or those Gypsies leading a nomadic lifestyle under certain privileges accorded to them in ancient times. The authorities’ interest in the Gypsies was strictly fiscal, and following the establishment of Habsburg rule in Transylvania and in other Romanian provinces on the other side of the mountains the Gypsies make frequent appearances in official documents. For example, there are instances of tax records at the level of the principality of Transylvania and the Kingdom of Hungary respectively, where Gypsies are registered under separate headings as well as tax records at the level of counties and other administrative units or towns, which record Gypsies together 148  C. Feneşan, Izvoare de demografie istorică, vol. I (Secolul al XVIII-lea. Transilvania), Bucharest, 1986, table XI. 149  Fr. Griselini, Încercare de istorie politică şi naturală a Banatului Timişoarei, translated by C. Feneşan, Timişoara, 1984, pp. 157–158.

1036

Achim

with their obligations, or even the amount of tax paid by them, particularly the gold-washers from the principality or the Banat. After the middle of the eighteenth century, this strictly tax-related interest in the Gypsies changed. A general and all-encompassing settlement of the problem of the Gypsies was undertaken during the reigns of Empress Maria Theresa and her son, Emperor Joseph II. A number of measures were adopted in succession with regard to the Gypsies; these were gradual in nature, indicating that we are dealing here with a clear policy elaborated by the Viennese Court in relation to this population.150 Maria Theresa issued four decrees relating to the sedentarisation and assimilation of the Gypsies.151 In 1758, she decreed that the Gypsies would be tied to one place and who would be required to pay taxes to the State and would be subject to a regime of obligations to their feudal masters; they would no longer be allowed to own horses and carts and would require special permission in order to leave the village in which they were settled. In 1761, the name “Gypsy” was replaced by decree with the name “new peasants” (neo-rustici, Neubauer) or “new Hungarians” (újmagyarok); according to the terms of the same deed, young Gypsies over the age of sixteen would be required to perform military service. The decree of 1767 abolished the jurisdiction of the voivode of the Gypsies, with the population now placed under the jurisdiction of the regular authorities; the use of Romanes was banned, as was the wearing of clothing and the practice of occupations specific to the Gypsies. In 1773, marriages between Gypsies were forbidden, while mixed marriages (with non-Gypsies) were strictly controlled. Gypsy children over the age of five were to taken away from their families and raised by nonGypsy families. These decrees and orders issued by Maria Theresa were targeted at the Gypsies from the Kingdom of Hungary, which at that time included western regions of present-day Romania. The administrative authorities also took them up at a county level. The measures issued by the Empress were not applied to the principality of Transylvania. Emperor Joseph II, who continued the policy of his mother with regard to the Gypsies, extended the measures introduced by her to cover Transylvania 150  For the policy of the two emperors with regard to the Gypsies, see: H. M. G. Grellmann, op. cit., pp. 143–151; I. H. Schwicker, Die Zigeuner in Ungarn und Siebenbürgen, Vienna und Teschen, 1883, p. 53ff; M. Tomka, “A cigányok története”, in Cigányok, honnét jöttek— merre tartanak?, ed. L. Szegő, Budapest, 1983, pp. 46–49; B. Mezey et al., op. cit., pp. 14–19; R. Stangl, op. cit., pp. 32–35; Gy. Szabó, Die Roma in Ungarn. Ein Beitrag zur Sozialgeschichte einer Minderheit in Ost- und Mitteleuropa, Frankfurt am Main, Bern, New York, Paris, 1991, pp. 69–75; A. Fraser, op. cit., pp. 157–159. 151  See I. H. Schwicker, op. cit., pp. 53–56.

The Gypsies in the Romanian Lands During the Middle Ages

1037

as well. On 12 September 1782, he gave an order regarding the Gypsies of the principality of Transylvania, known as the De Regulatione Zingarorum.152 This order contained the following requirements: – Gypsy children were to be sent to school; – they were no longer to be unclothed; – children of different sexes were to sleep separately; – Gypsies were to attend church on Sundays and public holidays; they were to follow the advice of priests; they were to follow the customs of the place where they lived with regard to what they ate, how they dressed and the language they spoke; – they were no longer to wear coats in which they could conceal stolen items; – no Gypsy was allowed to own a horse, with the exception of goldwashers; – the latter category of Gypsies were not permitted to trade in horses; village authorities were not to allow Gypsies to laze about; – Gypsies were required to work in agriculture; – where possible, the landowners who received Gypsies on their land were to provide them with a parcel of land, while anyone who refused to work the land was to be punished; – Gypsies would be allowed to perform music only when there was no work to be done in the fields. On 9 October 1783, Joseph II published the Hauptregulatio, a decree consisting of fifty-nine points regulating the status of the Gypsies in Hungary and Transylvania in all its aspects.153 This deed is a synthesis of all the previous measures adopted with regard to the Gypsies under the Habsburg monarchy. However, the regulations issued by the Emperor went even further. The following is a comprehensive tableau of the Habsburgs’ programme for the Gypsies in the second half of the eighteenth century: – Gypsies were forbidden to live in tents; – Gypsies previously under the authority of their voivode would from now on be under the authority of the village sheriff; – Gypsy children of four years and over were to be shared out among neighbouring settlements, at least every other year;

152  H. M. G. Grellmann, op. cit., pp. 147–150. 153  The text was republished in B. Mezey et al., op. cit., pp. 85–94. See I. H. Schwicker, op. cit., pp. 56–58.

1038

Achim

– nomadism was forbidden, and Gypsies already leading a sedentary way of life were permitted to go to market in another area only in cases of necessity and with special authorisation; – Gypsies were forbidden to own horses with the intention of selling them; Gypsy serfs were allowed to own horses, but only for use in agricultural work and they were not allowed to trade in them; – Gypsies were obliged to adopt the costume and language of the inhabitants of the village in which they are settled; – use of the Romanes was punishable by twenty-four lashes with a bat; – the same punishment would be applied to those found eating carcasses; – Gypsies were forbidden to change their names; – Gypsy houses were required to be numbered; – marriage between Gypsies was forbidden; – the local legal authorities would supply monthly reports on the way of life of Gypsies living in their district; – the number of Gypsy musicians had to be kept to the strictest minimum; – begging was forbidden; – it is compulsory for Gypsy children to attend school, with the priest responsible for ensuring their attendance; – landowners were required to make a parcel of land available to Gypsies in order to ensure their adoption of a sedentary way of life and an agricultural occupation; – anyone who abandoned their residence or occupation will be treated as a vagrant and returned to their residence. It can clearly be seen that the decrees and orders issued by the Habsburg emperors, as well as the measures taken on a local scale, constitute a carefully drawn-up policy that addresses every element of the Gypsy problem. It is difficult to estimate the effects of the policy following its introduction. Under the conditions of the autonomy and given the major differences from province to province within the Empire, decrees were as a rule applied either partially or even not at all. Only in certain places did the local authorities treat these instructions with utmost seriousness. It is reckoned that only on the western border of Hungary (today Burgenland, in Austria) did the measures taken by the local authorities match up more or less to the demands of the imperial decrees. However, there is no doubt that a part of the Gypsy population was settled into a sedentary way of life and tied to agricultural occupations. Contemporary sources note the existence of this rural population of “new peasants”, “new Hungarians” (in Hungary) or “new Banatians” (in the Banat). The 1780–83 census of the Gypsies living in the Kingdom of Hungary (together

The Gypsies in the Romanian Lands During the Middle Ages

1039

with Croatia—Slavonia), which did not include Transylvania proper, just happens to capture the moment when the policy of the Habsburgs with regard to the Gypsies was being put into practice.154 The number of the Gypsies recorded in 1780 was 43,609; in 1781—38,312; in 1782—43,772 and in 1783—30,241. The fact that in 1783 13,531 fewer Gypsies are recorded in comparison to the previous year is due to Gypsies no longer being registered as such but instead as sedentary “new peasants”. These changes in the figures reflect the process of sedentarisation of the Gypsies. The number of Gypsies that the census recorded in the counties and towns illustrates the scale of measures undertaken in this respect by the respective administrative units. In some counties the number of Gypsies fell, while in others it increased. If we examine some of the counties inhabited mainly by Romanians, we find that in Bihor the number of Gypsies fell during the period 1780–83 from 2289 to 1906; in Caraş, the number of Gypsies fell by 1008 in the space of a single year (1782–83); in the county of Arad, this number increased from 1135 to 1255; in Maramureş the number of Gypsies recorded fluctuated over the years 1780–82 from 446 to 903 and back to 717. Aside from the migration from one part of the country to another undertaken by the Gypsies during this period, the statistics of the census reflect the manner in which the measures ordered by Maria Theresa and Joseph II were applied from place to place. From this it follows that the process of sedentarisation did not take place on a large scale. Furthermore, in the following period, the “successes” of this time were lost, with some of the “new peasants” returning to their old way of life. The census of 1780–83 offers us a comprehensive tableau of the Gypsy population. It recorded all the elements that had a bearing on the policy of Vienna with regard to the Gypsies. Consequently, aside from general demographic data, the statistics reproduce for us the number of Gypsies living in houses and the numbers living in huts; the numbers of Gypsies with and without a fixed residence; among sedentary Gypsies, the number who owned a plot of land or part of it; those who wore normal dress (that is, similar to the locals) and those who wore traditional Gypsy costume. The census also tells us the distribution of occupations among the Gypsies: musicians, blacksmiths, other craftsmen, beggars; it tells us about those who obeyed the local legal authorities and those who did not; those who ate dead animal carcasses and those who did not; those who engaged in horse trading and those who did not, as well as the total fiscal obligations of the Gypsies. Special attention was paid in the census to the Gypsies’ children. It recorded by sex the number of children living with their parents and those who were taken from their parents and entrusted to other 154  I. H. Schwicker, op. cit., pp. 62–69.

1040

Achim

persons, the number of children who went to school and the professions for which they were trained. If we consider the data produced by the census of the Gypsies living in Hungary in 1780–83 as well as data from other official documents and contemporary testimonies, we can appreciate that the policy adopted with regard to the Gypsies did not lead to the desired effect. One explanation for this outcome might be that the policy was promoted at the highest level for too short a period. After the death of Joseph II, the Gypsies no longer represented a preoccupation of the Imperial Court. The three decades in which a concrete policy was applied to the problem of the Gypsies were not able to alter the destiny of a population that was both relatively numerous and in possession of a powerful sense of individuality. We believe that the causes of the policy’s extremely limited success lay principally in the fact that society was at that time not prepared to fully integrate the Gypsy population. Neither the nobility nor the local population were interested in the sedentarisation and assimilation of the Gypsies. Nobles had no interest in this because they were required to provide land for the newly sedentary Gypsies and to pay for the schooling of their children. The peasantry had no interest in the policy because the introduction of Gypsies onto the estates where they worked would have created extra pressure on them, as the parcels of lands that they worked barely enabled them to live as it was.155 As with other reforms introduced by the two emperors (especially those of Joseph II), imperial policy with regard to the Gypsies came up against hostility from the privileged classes, particularly from the nobility. Proof of this state of affairs is provided by the fact that after the death of Joseph II, the policy with regard to the Gypsies was abandoned, as was the case with other measures introduced by the reform-minded emperor. For a long time, the Imperial Court produced no new legislation with regard to the Gypsies. Another explanation for this state of affairs lies with the characteristics of the Gypsy population itself. Contemporary observers remarked on the difficulties faced by those who tried to implement the measures ordered by the Empire. When Gypsies were received on a noble’s estate, efforts to accustom them with agricultural work and with an orderly way of life as a rule failed to produce the desired results: the Gypsies refused to occupy the houses put at their disposal, preferring to live in huts; they refused to wear the same clothes as the other villagers etc.156 The policy of the Habsburgs aimed at more than just converting the Gypsies to a sedentary way of life and their integration into agricultural occupations; it aimed to assimilate the Gypsies, to erase their identity. The measures taken against the Gypsies were not, of course, con155  See R. Stangl, op. cit., pp. 34–35. 156  I. H. Schwicker, op. cit., pp. 59–61.

The Gypsies in the Romanian Lands During the Middle Ages

1041

ceived as measures based on ethnicity and race: rather, the Gypsies were treated as an asocial minority that ought to disappear. Perhaps this is what caused the Gypsies to demonstrate their powers of resistance, refusing to accept the extinguishing of their identity. The successes of the policy were limited and, to a certain extent, of a short duration. After 1790, some sedentarised Gypsies abandoned the houses and villages in which they had been settled and went back to their tents and huts or even to a nomadic existence. Children who had been removed from their families returned to their parents. Marriages among the Gypsies continued to take place, even in places where the ban on such marriages remained in force, evidently without any legal problems arising.157 In the Banat, the policy of sedentarisation was more successful than in other areas. In this province, which in the course of the eighteenth and nineteenth centuries was the most representative for the population policy of the Habsburgs, the authorities were greatly concerned with applying the measures stipulated in imperial decrees and orders. Although up until the reign of Maria Theresa almost all the Gypsies had been nomadic, already by the reign of Joseph II the majority of the Gypsies living in the province were sedentary. In the documents of the time they are the so-called Neubanater (new Banatians). According to the locality in which they settled, the Gypsies either adopted Romanian, German, Serbian or Hungarian as their native language. In the census of 1784, in Timişoara there were fifty Gypsy families, of which thirty were “German”. Out of the total number of Gypsies, thirty-six worked as musicians; out of these, thirty were “Germans”.158 In the Banat, only the gold-washers maintained a nomadic way of life for a time, but their freedom of movement was increasingly restricted. From the beginning of the nineteenth century, Gypsies from this category were also sedentarised, the majority of them in the military districts, that is, in the mountain villages in the south-east of the province. In Transylvania, the settlement ordered by Joseph II in 1783 was only partially implemented. Here, the status of the Gypsies came under the attributes of the Diet of the principality. A preoccupation with regard to the Gypsies on the part of the authorities in the principality dates from an earlier date: as early as 1747, the Diet had ordered that Gypsies who had fled their feudal masters were to be gathered and settled in a particular location. In 1791, the Diet renewed these measures.159 From a fiscal point of view, there were three categories of Gypsies living in Transylvania:

157  A. Fraser, op. cit., pp. 159–160. 158  I. H. Schwicker, op. cit., p. 61. 159  Ibid., p. 70.

1042 1.

2.

3.

Achim

Fiscal Gypsy gold-washers. These were authorised to engage in the collection of gold from riverbeds. Mining offices supervised the activity of these Gypsies, who were also under the jurisdiction of the offices. The occupation of such Gypsies was seasonal and they were nomadic. In 1781, 1291 families of gold-washing Gypsies were recorded. The revenues they generated for the State were relatively large. Fiscal taxed Gypsies. These Gypsies were so called because they paid an annual tax to the Treasury. They led a nomadic existence. They were organised into bands led by a voivode. In 1781, 1239 tents were recorded as well as twenty-six voivodes. The cameral tax was 933 florins and 8.5 kreuzers. Gypsies attached to the leading landowners and the towns. These were serfs or landless peasants tied to nobles’ estates, where they were chiefly employed as craftsmen, or they were living in relations of servitude to towns for whom they were required to carry out certain tasks. In the census of 1781, 12,686 heads of Gypsy families were recorded, together with their wives, giving a total of 35,539 people living in this state. A total of 8598 families had fixed dwellings, while 4088 were nomadş 10,947 were serfs, while 1739 were landless peasants.

The situation of these Gypsies, together with all the data mentioned above, was contained in a report produced in the year 1794, entitled Opinio. De domicilitatione et de regulatione Zingarorum.160 The report makes note of all Transylvanian legislation regarding the Gypsies starting with the year 1747 and in the spirit of this legislation and of imperial policy, presents the appropriate means of achieving the sedentarisation and assimilation of the Gypsies. As the structure of the Gypsy population in Transylvania was different from that of Hungary, the solutions to which this material refers, even if they are essentially the same, are appropriate to the realities in Transylvania. Pressure for the sedentarisation of the Gypsies, particularly the taxed Gypsies, was greater in Transylvania, which explains why at this time, in earlier periods and in the nineteenth century we come across the migration of the groups of the tentdwelling Gypsies from Transylvania into the Hungarian puszta, an area more suitable to a nomadic way of life.161

160  S. l., s.a., p. 17 in folio. The material was republished in Ethnologische Mitteilungen aus Ungarn, III (1893–1894), pp. 55–56, 114–116, 168–170, 210–212, 221–223. See also A. Gebora, op. cit., pp. 52–60. 161  I. H. Schwicker, op. cit., p. 70.

The Gypsies in the Romanian Lands During the Middle Ages

1043

The measures ordered by Maria Theresa and Joseph II gave impetus to the process of sedentarisation of the Gypsies in Transylvania, the Banat and in the west of present-day Romania, as in Hungary and on territories constituting present-day Slovakia. Even though it is necessary to recognise that there were cases of sedentarisation and assimilation of the Gypsies prior to this time, such cases were isolated and did not determine a major change in the way of life of the majority of the population. Only in the second half of the eighteenth century can we speak of a large-scale process in this sense, which affected the Gypsiy population at large. Now the Gypsies were given permission to build fixed dwellings and to own land, rights previously denied to them. Over a few decades, the majority of the Gypsies moved to a sedentary way of life. They settled in villages, working as farmers or craftsmen (especially blacksmiths). Their status underwent a fundamental change: from being members of a tolerated socio-ethnic group, they became inhabitants of the country, with a social and fiscal status identical to the other people living in the locality where they settled. In this way, they were integrated into society, even though they preserved, either as individuals or as a subgroup, certain distinctive features. Linguistic and ethnic assimilation did not take place necessarily, and when it did occur, it was the result of evolutions spread over two to three generations. From the end of the nineteenth century, the nomadic Gypsies became the minority. The proportion of nomadic Gypsies fell from generation to generation. Of course, it cannot be said that this process was a creation of the legislation introduced with regard to the Gypsies during the reigns of Maria Theresa and Joseph II. It is first of all the result of the natural evolution of society in this part of Europe, which, together with the reforms introduced in the middle and second half of the eighteenth century, offered fewer and fewer opportunities for the practice of a nomadic way of life. The assimilation policy promoted by the two monarchs with regard to the Gypsies belongs within this trend. It had the advantage of intensifying a natural process and leading it in the direction of the Habsburg population policy.

Serfs, Slaves, or Wage Earners? The Legal Status of Labour in Russia from a Comparative Perspective, from the Sixteenth to the Nineteenth Century* Alessandro Stanziani Since at least the eighteenth century, many comparative analyses of labour in Russia and the West have been carried out as if the dividing line between free and forced labour was universally defined in ahistorical terms. Free labour in the West was opposed to serfdom in eastern Europe, which in turn was held to be a form of forced labour, along with slavery and indentured service. Medievalists since Marc Bloch1 have produced more and more critical studies of western European ‘serfdom’.2 However, it has only been in the last two decades that historians have followed this example to show that, in Prussia, Lithuania, and Poland, the line between ‘free’ work and serfdom was fluid, negotiable, and negotiated, and that it was rooted not only in philosophical and political debates, but also in the way that social actors appropriated legal rules.3 These new interpretations of eastern Europe’s ‘second serfdom’ have partly reiterated the ways in which the historiography of slavery and indentured service has evolved over at least the last two decades, emphasizing the

Source: Stanziani, Alessandro, “Serfs, Slaves, or Wage Earners? The Legal Status of Labour in Russia from a Comparative Perspective, from the Sixteenth to the Nineteenth Century,” Journal of Global History, 3(2) (2008): 183–202. © Cambridge University Press, reproduced with permission. * I would like to thank the referees and the editors of the journal who corrected and suggested modifications in previous versions of this article. Thanks also to CNRS who provided a grant for this reseach. 1  Marc Bloch, ‘Serf de la glèbe: histoire d’une expression toute faite’, Revue Historique, 36, 1921, pp. 220–42. 2  Guy Bois, La Crise du féodalisme, Paris: Presses de Sciences-Po, 1980; George Duby, Les trois ordres ou l’imaginaire du féodalisme, Paris: Gallimard, 1978; Pierre Bonnassie, From slavery to feudalism, Cambridge: Cambridge University Press, 1991; Tom Scott, ed., The peasantries of Europe: from the fourteenth to the eighteenth centuries, London: Routledge, 1998. 3  Edgar Melton, ‘Population structure, the market economy and the transformation of Gutsherrschaft in East Central Europe, 1650–1800: the cases of Brandenburg and Bohemia’, German History, 16, 3, 1998, pp. 297–324; William Hagen, Ordinary Prussians: Brandenburg Junkers and villagers, 1500–1840, Cambridge: Cambridge University Press, 2002.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_034

Serfs, Slaves, Or Wage Earners ?

1045

shifting line dividing free labour from slavery.4 For example, until the middle of the nineteenth century, indentured service was viewed as a voluntary contract rather than as a form of servitude. The definition of indentured service, or the French contrat d’engagement, was based on ‘ordinary’ contracts, Masters and Servants Acts in the British Empire, or the renting of services in the French Empire.5 Similar conclusions might be reached regarding forms of servitude in the Ottoman Empire, although this case has been investigated to a much lesser degree.6 It is surprising, at first glance, that Russia has remained on the sidelines of this debate. Discussions of serfdom in Russia concern its creation by the state or by landlords,7 or its profitability.8 Questions about the nature of serfdom, or its very existence, surface in a few histories of ideas,9 but have seldom been touched upon in economic and social history. The aim of this 4  Stanley Engerman, ed., Terms of labor: slavery, serfdom and free labor, Stanford, CA: Stanford University Press, 1999; Robert Steinfeld, The invention of free labor: the employment relation in English and American law and culture, 1350–1870, Chapel Hill, NC: University of North Carolina Press, 1991; Michael L. Bush, Servitude in modern times, Cambridge: Polity Press, 2000. 5  Steinfeld, Invention. 6  Ehud Toledano, Slavery and abolition in the Ottoman Middle East, Seattle, WA: University of Washington Press, 1998; Omer Lutfi Barkan, ‘Le servage existait-il en Turquie?’, Annales ESC, 11, 1956, pp. 54–60. 7  Jerome Blum, Lord and peasant in Russia: from the ninth to the nineteenth century, New York: Atheneum, 1964; Alexander Gerschenkron, Economic backwardness in historical perspective, Cambridge, MA: Harvard University Press, 1962; Richard Hellie, Enserfment and military change in Muscovy, Chicago, IL: University of Chicago Press, 1971; Peter Kolchin, Unfree labor: American slavery and Russian serfdom, Cambridge, MA: Harvard University Press, 1987; Daniel Field, The end of serfdom: nobility and bureaucracy in Russia, 1855–1861, Cambridge, MA: Harvard University Press, 1976; Steven Hoch, Serfdom and social control in Russia: Petrovskoe, a village in Tambov, Chicago, IL: University of Chicago Press, 1986. 8  Ivan D. Koval’chenko, Russkoe krepostnoe krest’ianstvo v pervoi polovine 19th v. (The Russian serf economy during the first half of the nineteenth century), Moscow: Nauka, 1967; Ian Blanchard, Russia’s age of silver: precious metal production and economic growth in the eighteenth century, London: Routledge, 1989; David Moon, The abolition of serfdom in Russia, 1762–1907, London: Pearson Education, 2001; Tracy K. Dennison, ‘Did serfdom matter? Russian rural society, 1750– 1860’, Historical Research, 79, 203, 2003, pp. 74–89; Edgar Melton, ‘Proto-industrialization, serf agriculture, and agrarian social structure: two estates in nineteenth-century Russia’, Past and Present, 115, 1987, pp. 73–81; Edgar Melton, ‘Enlightened seignorialism and its dilemmas in serf Russia, 1750–1830’, Journal of Modern History, 62, 4, 1990, pp. 675–708; Evsey Domar and Michael Machina, ‘On the profitability of Russian serfdom’, Journal of Economic History, 44, 4, 1984, pp. 919–55. 9  Larry Wolff, Inventing eastern Europe: the map of civilization on the mind of enlightenment, Stanford, CA: Stanford University Press, 1994; Alessandro Stanziani, ‘Free labor—forced labor: an uncertain boundary? The circulation of economic ideas between Russia and Europe

1046

Stanziani

article is to show that the historical and institutional definition of serfdom in Russia poses a problem, and that the legal status of peasants and labourers is worth investigating further. I will study Russian legislation, and how it was applied in this area over the long term (sixteenth to nineteenth centuries). I will show that, contrary to generally accepted arguments, serfdom as such was never clearly introduced institutionally in Russia. Instead, over three centuries, we find regulations aimed at defining ‘nobles’ and those who were legally entitled to own and transfer inhabited estates. Forms of dependence were no doubt extreme compared with their counterparts in western Europe during the same period, but it was less a question of an opposition between free and forced labour than of gradations within a common world. I will also discuss the presence of slaves in Russia, and examine the association between certain forms of servitude (especially for debt) and chattel slavery. We find the presence of genuine slaves in the empire, as a result of the expansion of the Russian Empire towards the south, the east, and the west, and due to commercial relations with the Caucasus and Islamic empires. These forms of bondage will be compared with those existing at the same time in other situations, notably indentured service in western metropoles and colonies, debt servitude in India, and Islamic slavery. My work is based on sources from the period, especially materials in Russian archives, which contain many judicial decisions. In addition, I have consulted numerous collections of Russian laws, decrees, and ‘high’ jurisprudence (134 volumes in three series).10 From a comparative perspective, I have also had recourse to French archives and English and French case law of the eighteenth and nineteenth centuries. However, legal documents and legal definitions of status are not everything, for it is also important to know how norms were applied. I will show that, in Russia, not only nobles and the bourgeoisie but also peasants and labourers made widespread use of norms defining ‘genuine landowners’ and their rights. They used these rules to challenge a particular title to ownership, the rights over labour that were associated with the ownership of land, and hence their own obligations to their master. The rights of peasants were distinctly inferior to those of other social strata, but they did exist. Well before the reforms of 1861 that officially abolished serfdom, half the peasants on private estates had already acquired a different legal status, becoming state peasants or city from the 18th to the mid-19th century’, Kritika: Explorations in Russian and Eurasian History, 9, 1, 2008, pp. 27–52. 10   Polnoe sobranie zakonov Rossiskoi Imperii (Full collection of laws of the Russian Empire— henceforth PSZ), three series: I: 1649–1825, 46 vols., St Petersburg, 1830; II: 1825–81, 55 vols., St Petersburg, 1830–84; III: 1881–1913, 33 vols., St Petersburg, 1885–1916.

Serfs, Slaves, Or Wage Earners ?

1047

dwellers. Among the remaining private peasants, only about half still owed labour services. Conversely, after 1861, many legal and economic constraints on peasants were kept in place. These findings lead us to two more general conclusions. First, rather than opposing ancien regime systems to others and stressing institutional breaks such as the abolition of serfdom, it is preferable to take into consideration the slow evolution of markets, especially labour markets, and their institutions, which had both rigid and flexible characteristics. Second, between the seventeenth and the nineteenth century, not only in Russia but also around the globe, labour statuses and regulations were not those familiar to us today, which were not introduced until the early twentieth century.11 This approach is indispensable to avoid anachronism, and to understand how economies and institutions worked during the early modern period, and how they evolved towards current forms of labour control. I will begin by studying the presence of slaves and captives in the Russian Empire. I will then analyse kholopy: that is, forms of dependence ranging from domestic service to quasi-slavery, under contracts that were widespread in Russia until their abolition in 1725, but which had little impact in the countryside. In the third section, I look at the gradual introduction of legislation relating to rural ‘serfdom’. In the fourth section, we will see how regulations were applied, and ensuing changes in status. In the last section, conclusions will be drawn concerning the development of serfdom in Russia, and I will offer general observations on the ‘second serfdom’ of eastern Europe. In my conclusion, I attempt to place Russian forms of labour within a still more general framework, questioning the typology of servitude in the world and forms of so-called ‘free’ labour in the West during the period under study.

Captives at the Crossroads of Empires

The Russian Empire interacted with Islamic regions in which chattel slavery was common, as the only legitimate form of coerced labour under Islamic law. Muslim Tatars of the Crimea raided widely for Russian subjects, as well as Poles and Lithuanians, and exported most of their captives to the Ottomans.12 From the 1570s, about 20,000 slaves were sold annually in the port of Kaffa on the

11  Stanley Engerman, ‘Slavery at different times and places’, American Historical Review, 105, 2, 2000, pp. 480–4. 12  William G. Clarence-Smith, Islam and the abolition of slavery, London: Hurst, 2006, p. 13.

1048

Stanziani

Black Sea.13 Until the early seventeenth century, Russians, above all Cossacks, also sold captives to the Tatars, or directly to the Ottomans.14 Rules governed the criteria for redeeming Russian captives. Thus, Kalmyk Mongols agreed in 1661 to free Russians whom they had acquired through Tatars, and in 1678 to return Russians whom they themselves had taken captive.15 As late as the mid nineteenth century, the Russians were redeeming slaves from Turkistan.16 The Russian Empire also gradually incorporated areas where local populations had long practised forms of servitude and slave-trading.17 Many inhabitants of the Caucasus, especially Christian Georgians and Armenians, together with heterodox Muslim Circassians, were sent as slaves to the Ottoman Empire, whether overland or across the Black Sea. For the first three-quarters of the nineteenth century, the Ottomans imported between 16,000 and 18,000 such slaves on average every year.18 Some male slaves entered the servile administrative elite of the Ottoman Empire, while many women ended up in the harems of the rich and powerful. Circassian families at times sold their own children to intermediaries, who transported them to Ottoman territory. Under British pressure, the flow of slaves from the Caucasus was suspended in 1854, but it grew again after the end of the Crimean War. Moreover, the brutal Russian conquest of Circassia led to an influx of between half a million and a million refugees into the Ottoman domains between 1854 and 1865, of whom perhaps a tenth were of servile status.19 These massive arrivals increased numbers of agricultural slaves, relatively small beforehand.20 13  Halil Inalcik, ‘Servile labour in the Ottoman Empire’, in Abraham Ascher, Tibor HalasiKun, and Bela Kiraly, eds., The mutual effects of the Islamic and Judeo-Christian worlds: the east European patterns, Brooklin, NY: College Press, 1979, pp. 39–40; Yvonne Seng, ‘Fugitives and factotums: slaves in early sixteenth-century Istanbul’, Journal of the Economic and Social History of the Orient, 39, 2, 1996, pp. 136–69. 14  Alan Fisher, ‘Muscovy and the Black Sea trade’, Canadian—American Slavic Studies, 6, 4, 1972, pp. 582–93. 15   Materialy po istorii Uzbeskoi, Tadzhikskoi I (Materials for the history of Soviet Uzbekistan, Tajikistan, and Turkmenistan), part 1, Leningrad: AN SSSR, 1932, pp. 386–97, cited in Richard Hellie, Slavery in Russia, 1450–1725, Chicago, IL: University of Chicago Press, 1982, p. 25, n. 43. 16  Clarence-Smith, Islam, pp. 118–19. 17   Sbornik Imperatorskogo Russkogo Istoricheskogo Obshchestvo (Collected works of the Imperial Russian historical society), vol. 41, St Petersburg, 1884, pp. 42–3, 52–3, 104–7, 115– 21, 146–57. 18  Toledano, Slavery, p. 8. 19  Thomas Barrett, ‘Lines of uncertainty: the frontier of the north Caucasus’, Slavic Review, 54, 3, 1995, pp. 578–601; Clarence-Smith, Islam, pp. 13–14. 20  Toledano, Slavery, p. 81; Barkan, ‘Le servage’.

Serfs, Slaves, Or Wage Earners ?

1049

Captives also arrived in the Russian Empire from both Muslim and Catholic areas. According to the Russian laws of the period, captives were intended to serve the elites as administrative assistants or domestics. Their maximum term of service was supposed to last only until the death of the master to whom they had been entrusted. They could also be redeemed by an agreement between the Russian state and their country of origin. If they converted to Orthodox Christianity, they might be emancipated, although this was not mandatory. Several sources note the problems encountered by the Moscow authorities in ensuring compliance with these norms, and servitude for war captives persisted. After the Thirteen Years War (1654–67), nobles to whom Lithuanian and Polish captives were attributed did not register them, and tended, in practice, to treat them as genuine slaves.21 In 1655, Poles, Lithuanians, and others, both adults and children, were openly sold on the streets of Moscow.22 As a result of this war, many people were sold in Russia, at times becoming kholopy, as discussed below.23 In short, real slaves were present in Russia. As in other historical situations, they were typically taken in raids where boundaries were uncertain, or during military operations in the strict sense. From a geopolitical standpoint, these forms of slavery were linked to conflicts with the Islamic world, notably the Ottoman and Persian empires, as well as to the conflicts that tore Europe apart in the seventeenth century.24

Kholopy: Slaves, Serfs, or Indentured Servants?

We now have to determine whether there were other forms of slavery in Russia, particularly service for debts, which concerned not only foreign ethnic and religious groups but also Russians themselves. In the Russian language, from the fourteenth to the sixteenth century, the term krepost’ designated a legal document pertaining to a sale, ownership, or a loan. This document has 21  Hellie, Slavery, pp. 68–69. 22  Paul of Aleppo, The travels of Macarius: extracts from the diary of the travels of Macarius, Patriarch of Antioch, ed. by Lady Laura Ridding, London: Oxford University Press, 1936, pp. 28, 76. 23  Aleksandr’ L. Khoroshkevich, Russkoe gosudarstvo v sisteme mezhdunarodnykh otnoshenii kontsa XV—nachala XVI v.(The Russian state in the system of international relations towards the end of the fifteenth and the beginning of the sixteenth century), Moscow: Nauka, 1980, pp. 30–2. 24  David Brion Davis, Slavery and human progress, New York: Oxford University Press, 1984; Robert Crummey, The formation of Muscovy, 1304–1614, London: Longman, 1987.

1050

Stanziani

usually been identified with the proof of a landowner’s rights over peasants. As long as the document and the peasant’s obligations remained valid, the latter’s mobility was restricted. In addition to peasants with these obligations, others were considered to be starinnye (‘long established’), and hence without obligations of krepost’ to the noble landowner. They had to prove their status with a document issued by the administration.25 Krepost’ designated both the contract between peasant and noble, and the certification of its validity by the state, giving it at once a contractual and an administrative value.26 And yet, in the official regulations adopted from this period up to the famous Ulozhenie of 1649, which is generally considered to have institutionalized serfdom in Russia, the word krepost’ was not associated with peasants in general. Rather, it pertained to a particular category of people, the kholopy, and an associated type of contract, kholopstvo. The Ulozhenie of 1649 devoted an entire section to this topic.27 Richard Hellie initially translated kholopstvo as ‘bondage’, but later preferred the term ‘slavery’. Herbert Leventer objected to the latter translation, emphasizing that the status of Russian kholopy was not transferred to their children, that their servitude was temporary, and that they could accumulate and transfer property. He therefore thought that kholop corresponded instead to the English word ‘servant’. Hellie retorted that, in Russian, kholop was a synonym for rab (slave), and that, even if the conditions of the kholopy were different from those of slaves in antiquity and the Americas, they were perfectly compatible with those of other forms of slavery.28 Translating kholop as ‘slave’ is partly justified by the fact that, when Peter the Great abolished this status, the documents of the period associated the kholopost’ with slaves (rab). This association of ideas dates from the early 25  Mikhail F. Vladimirskii-Budanov, Obzor istorii russkogo prava (Summary of the history of Russian law), sixth edn, Kiev: Izdanie knigoprodstva N. Ia. Oglobina, 1909. 26   Entsiklopedicheskii slovar’ Brokgauz-Efron (Encyclopaedia Brokgauz-Efron), vol. 16, St Petersburg: Brokgauz, 1895, entry for krest’ianie (peasant), p. 681. See also Slovar’ russkogo iazika XVIII veka (Dictionary of the Russian language of the eighteenth century), vol. 10, St Petersburg: Sorokin, 1998, entry for krepostnoi. 27  Hellie, Slavery; Elena I. Kolycheva, Kholopstvo i krepostinichestvo, konets XV–XVI vek (The kholopy and enserfment, end of the fifteenth century to sixteenth century), Moskow: Nauka 1971; Viktor M. Paneiakh, Kholopstvo v pervoi polovine XVII veke (Kholopstvo in the first half of the seventeenth century), Leningrad: Nauka, 1984. 28  Richard Hellie, ‘Recent Soviet historiography on medieval and early modern Russian slavery’, Russian Review, 35, 1, 1976, pp. 1–36; Herbert Leventer, ‘Comments on Richard Hellie’s ‘Recent Soviet historiography on medieval and early modern Russian slavery’, Russian Review, 36, 1, 1977, pp. 64–7; Richard Hellie, ‘Reply’, Russian Review, 36, 1, 1977, pp. 68–75.

Serfs, Slaves, Or Wage Earners ?

1051

eighteenth century, however, and occurred in the special context of the reforms of Peter the Great. At the time, the word rab was used to designate a form of dependence with no specific legal value, insofar as slavery in the strict sense was prohibited, so that rab designated either a former slave, one mentioned in the Bible, or the symbolic relationship that the nobles maintained with the Tsar.29 Let us try instead to grasp the meaning and content of the word kholop through the centuries prior to its official abolition. From the fifteenth century at least, the word appeared in quite disparate sources: certain Sudebniki (law collections) and judicial cases, as well as private transactions, contracts, memoranda, estate accounts, registrations with solicitors, and so on. These documents never speak of kholopstvo in general, but qualify the word with another: starinnoe (‘hereditary’), polnoe (‘full’), dokladnoe (‘registered’), dolgovoe (‘obligated’, ‘indebted’), zhiloe (‘limited to a period of time’), dobrovol’noe (‘voluntary’), kabal’noe (‘limited to service’). The latter was by far the most widespread term, found in 80–92% of known contracts of kholopstvo, depending on the period.30 This multiplicity of qualifiers is significant because it indicates a set of contracts rather than a single formal personal status. Let us take the most widespread of these contracts, the kabal’noe kholopstvo, which appears in legislation, disputes, contracts between private individuals, wills, and estate inventories.31 All these documents mention the length of service and the possibility of transforming a six-month or one-year contract into a contract of unlimited service.32 However, the latter practice was prohibited in the early seventeenth century.33 The code of 1550 clearly emphasized that the kabal’nye were not dolgovye (‘indebted’). In subsequent years (1586 and 1597), 29  Marshall Poe, ‘What did Russians mean when they called themselves “slaves of the tsar”?’, Slavic Review, 57, 3, 1998, pp. 585–608. 30  Out of 2,499 documents with the words kholop or kholopostvo, 2,116 refer to the kabal’noe variety (Hellie, Slavery, p. 33). Examples of contracts are in the Saltykov-Shchedrin Library in St Petersburg, manuscript section, Obshchee sobranie gramot, nos. 1727, 1937, 1941, 2017, 2019, 2348, 2406, 2635, 2672, 3026, 3081, 3392, 3475, 3486. 31  L. V. Cherepnin and S. V. Bakhrushin, eds., Dokumenty i dogorovnye gramoty velikikh i udel’nykh kniazei XIV–XVI vv. (Documents and acts decreed by princes, fourteenth to sixteenth centuries), Moskow: Nauka, 1950, p. 409, n. 98. 32  Paneiakh, Kholopstvo; Viktor Paneiakh, ‘Ulozhenie 1597 g. o kholopstve’ (‘Ulozhenie of 1597 on kholopstvo’), Istoricheskie Zapiski, 77, 1955, pp. 154–89. 33  In 1609, this was reduced from six to five months, and was further reduced to three months in 1649: Akty istoricheskie, sobrannye i izdannye arkheograficheskoiu kommissieiu (Historical acts, collected and published by the Archaeographical Commission), 5 vols., St Petersburg, 1841–2, vol. 2, no. 85.

1052

Stanziani

new provisions confirmed that the kabal’nye could remain obligated only for the duration of the creditor’s life, and that the latter could not transfer the obligations to anyone, either in the form of a sale or an inheritance.34 This latter provision could be interpreted as the desire to maintain the kholop in a state close to slavery, but it is equally legitimate to interpret it as a provision aimed to exclude that form of dependence, and the link with the previous provisions would seem to confirm the latter interpretation. This conclusion is bolstered by all the contracts that have been found, which indicate the length of commitment, usually limited to one year.35 Moreover, the term translated by Hellie as ‘slavery for debt’ (dolgovoe kholopostvo) actually referred, according to the Sudebnik of 1550 and 1589, to labour services due by those who had been condemned to make a payment in compensation, but had found themselves unable to do so. However, that could only be in special cases, for, on 1 October 1560, creditors were prohibited from making debtors sign this type of clause (service obligations) in contractual commitments.36 It remains to examine the most extreme forms of kholopstvo. The ‘full’ (pol’noe) variety was very old, and had three main sources. The kholop himself or herself might ask to be included in this category, as a form of repayment of a debt to the authorities. Second, if a female kholop married a free man, without the authorization of the person entitled to the wife’s service, her husband became pol’noe kholop. The third source was domestic service contracts established for an unlimited length of time, but such contracts have been found only between 1430 and 1554, with none appearing after that date. The most widely accepted hypothesis is that this form of dependence tended to be transformed into other forms of kholopstvo of a temporary nature. The hereditary variety (starinnoe kholopstvo) seems to come closest to slavery in the strict sense. It expresses the condition of those whose parents were kholopy. It was possible to transfer such kholopy in wills, or as a dowry or gift. In the contracts examined by Hellie, there were 5,575 kholopy between 1430 and 1598, 483 of whom were hereditary. The kabal’nye knigi, at the end of the seventeenth century, mentions 418 hereditary kholopy out of a total of 2,168 registered at the time. The available sources do not allow us to say whether this higher percentage testifies to the poor economic situation of the time, or to a 34  Paneiakh, ‘Ulozhenie 1597’, p. 161. 35  Viktor M. Paneiakh, Kabal’noe kholopstvo na Rusi v XVI veke (Temporary limited servants in Russia in the sixteenth century), Leningrad: Nauka, 1967, pp. 127–8. 36  Mikhail F. Vladimirskii-Budanov, Khristomatiia po istorii russkogo prava (Compendium of the history of Russian law), St Petersburg: Izdanie knigoprodstva N. Ia. Oglobina, 1875, vol. 3, pp. 29–30, 41.

Serfs, Slaves, Or Wage Earners ?

1053

long-term trend, as this type of commitment was prohibited by the decrees of 1586 and 1593. To be sure, Iakovlev, and more recently Paneiakh, have found disputes and contracts concerning starinnye kholopy in the middle of the seventeenth century, decades after the official abolition of this type of contract.37 In other words, despite the official prohibition, several lords continued to impose forms of contractual servitude of a permanent and hereditary type. The authorities devoted much attention to what amounted to illegal slavery, and attempted to penalize transgressors; by banning this kind of servitude, the government sought to limit the power of nobles over peasants, and thereby strengthen state authority in relation to the owners of large estates. Furthermore, the kholopy were exempt from taxation, which reduced the revenue of the state. At the same time, when hereditary kholopstvo was prohibited, the krepost’ over the peasants was reinforced, in the sense that the latter were subject to legal forms of dependence on landowners and the state. This was a measure intended to strengthen small landowners, and to encourage their alliance with the state. Measures to eradicate hereditary servitude had important consequences. Rather than exclude part of the population from all civil rights, as in the case of slavery, the solution consisted in assigning highly differentiated rights to the various strata of the population, and dividing them into legally distinct groups. The peasants saw their rights severely restricted, while city dwellers were prohibited from subjecting themselves, even voluntarily, to any form of krepost’ or kholopstvo. Numerous provisions defined those entitled to sign, as creditors, a kholopstvo contract, as well as those who were entitled to enter into such relations as debtors. Thus, in 1641, the following were excluded from the category of creditors entitled to demand labour service: all tiaglye liudi (people subject to tiaglo, the unit of taxation), including peasants and artisans as well as other taxpayers, priests, artillerymen, and monastery servants.38 Conversely, starting in 1590, city dwellers subject to taxation (posad) were prohibited from offering these forms of labour service. In 1628, this prohibition was extended to include musketeers, soldiers, and all the intermediate ranks of the civil service and the military. The interpretation of these norms posed problems, as the categories were rather general. In the case of professions such as barbers, seamstresses, trappers, and small craftsmen, the question arose as to whether or not they could legitimately enter into kholopstvo contracts. The many petitions 37  Paneiakh, Kabal’noe; Aleksandr’ I. Iakovlev, Kholopstvo i kholopy v moskovskom gosudarstve XVII v. (Kholopstvo and kholopy in the Russian state in the seventeenth century), Moskow: Nauka, 1943. 38  Hellie, Slavery, p. 75; Iakovlev, Kholopstvo, p. 316.

1054

Stanziani

sent to the chancellery concerning such individuals demonstrate their involvement in these contracts, their desire to be able to continue being taken on as kholopy, and their use of the law to challenge the claims of their counterparts.39 From this point of view, the 119 articles of section 20 of the Ulozhenie of 1649 that were devoted to kholopy reproduced in large part the provisions of earlier legislation. For those who failed to meet their legal obligations (debts, penalties, fines, theft, etc.), the text indicated the amount of work required to repay their debt or, in general, to fulfil their obligation. Once the work was completed, the creditor brought the debtor back before an official, who released the debtor from all obligations. Section 20 of the Ulozhenie also mentions other conditions for release from kholopstvo. Various articles speak of both debts and krepost’, with the latter viewed as justifying the debt. The core provisions of section 20 of the Ulozhenie depart from the rules found in some slaveholding systems, although they are not very different from slavery in Islamic and Catholic areas. Kholopy were free to marry, and such an act was inviolable. The wife of a kholop was obliged to remain in residence until her husband’s debt was repaid, but, upon the death of the husband, his wife’s dowry passed to her family, and not to the landowner-creditor.40 The kholop could be called as a witness in a trial, which meant that legal personality was acknowledged. Diverging most from systems of slavery elsewhere, a master of kholopy had no obligation to feed or provide care for elderly kholopy, whereas this obligation formed part of a master’s commitment throughout the length of the contract itself.41 Overall, when Peter the Great abolished the kholopstvo status in 1725, it concerned 10% of the population.42 The 2,500 contracts and documents that have been recovered are almost all (92%) from the Novgorod region and were signed (in 80% of the cases) between 1581 and 1603. According to Hellie’s calculations, 23% of the cases involved single men, and 60.4% couples without children. The rest were couples with a minor child (1.6%), widowers (4%), widows (3.7%), married women (2.5%), and unmarried women (4.2%), while 39   Opisanie dokumentov i bumag, khraniashchikhsia v moskovskom arkhive ministerstva iustitsii (Inventory of documents and papers kept in the Moscow Archives of the Ministry of Justice), vol. 15, St Petersburg, 1908. 40  Petr Ivanovich Ivanov, Alfavitnyi ukazatel’ familii i lits, upominaemykh v boiarkikh knigach, khraniashchikhsia v l-m otdelenii moskovskogo arkhiva ministerstva iustitsii, (Alphabetical index of families and persons named in the boyari books, conserved in the first section of the Moscow Archives of the Ministry of Justice), Moskow: Ministerstvo Iustitsii, 1853. 41  Hellie, Slavery, p. 211. 42  Kolycheva, Kholopstvo; Paneiakh, Kholopstvo.

Serfs, Slaves, Or Wage Earners ?

1055

the status of the others was unknown. In the majority of cases, the kholopy were between ten and thirty-four years of age, but about 10% were between the ages of ten and fourteen, and the same percentage between the ages of five and nine. Finally, men made up at least two-thirds, and often virtually all, of the kholopy throughout the period under study, from the sixteenth to the late seventeenth century.43 Nearly all the kholopy were domestic servants, and they were rarely assigned to farm work. The available contracts show that about 20% of the kholopy concerned children between five and fourteen years of age, who were placed in service by their parents, under one-year contracts that were often renewable. Some contracts were for rather long periods. Such contracts were signed by the most disadvantaged among the city population, and the numbers rose at the turn of the seventeenth century, at a time of serious economic crisis. In a way, it meant placing children in service to ensure their survival. From this point of view, the kholopstvo contract for children sprang from the same motives as several contracts of this type that were widespread during the same period in France and England (servants in husbandry), albeit with different legal terms and institutional conditions.44 The other kholopstvo contracts referred to adults working as servants. Loans were sometimes the formal reason for these contracts, but the terms of the loans often suggest that these were really servants’ wages. Taking these elements into account, we can conclude that most of the aspects of kholopstvo, above all following the decline in its hereditary forms, resemble other types of indebtedness and limitation on mobility, such as forms of contractual servitude widely found in the same period among Hindu populations in India, and in parts of China. Temporary servitude fell within the scope of contracts that were considered ‘free’ and voluntary from a legal standpoint. Freedom of commitment did not exclude the renewal of contracts for up to several decades, or even throughout the lifetime of the ‘indebted’ person.45 However, the Russian situation differed from the one prevalent in the Islamic world, where sharia law forbade all forms of bondage for debt, crimes, and indigence, even if they occurred in practice under customary or sultans’ law.46 43  Hellie, Slavery, pp. 423–4. 44  Anne Kussmaul, Servants in husbandry, Cambridge: Cambridge University Press, 1981. 45  Gyan Prakash, ‘Terms of servitude: the colonial discourse on slavery and bondage in India’, in Martin Klein ed., Breaking the chains: slavery, bondage and emancipation in modern Africa and Asia, Madison, WI: University of Wisconsin Press, 1986, pp. 131–49; Harriet T. Zurndorfer, Change and continuity in Chinese local history: the development of Hui-chou Prefecture, 800 to 1800, Leiden: E. J. Brill, 1989, notably ch. 5. 46  Clarence-Smith, Islam, pp. 74–80; Toledano, Slavery.

1056

Stanziani

In virtually all the known Russian contracts, and increasingly so over time, the status of kholopy could not be transferred to descendents, and that is essentially what distinguished this system from slavery in antiquity and in the Americas. From this point of view, the contracts in question resembled indentured service in the British Empire. As Robert Steinfeld has shown, until the 1830s and 1840s, this contract was considered a form of voluntary commitment and, as such, it was the opposite of forced labour, which was identified with slavery and serfdom. That said, the voluntary nature of the commitment did not exclude quite harsh conditions of exploitation.47 It was no accident that, as indentures disappeared for Europeans emigrating to the United States by the mid nineteenth century, their use expanded for the Chinese, Indian, African, and other ‘coolies’ working in European colonies and Latin American countries after the formal abolition of slavery.48 In other words, by their very existence, forms of voluntary bondage testify to the variety of labour commitments, and to continuity rather than opposition between these forms, ranging from statutory and hereditary slavery to ‘free’ labour. Indeed, it would have been impossible to define voluntary bondage without the Master and Servant Act and its imperial variants, which assimilated the wage earner less to the one we know today than to the servant inherited from pre-industrial periods. The indentured labourer was a particular form of servant.49 Fugitives from the ranks of apprentices, domestics, and the indentured were caught by the state’s police forces, and were subject to criminal procedures. Such ‘penal sanctions’ applied equally to the Russian kholopy. However, unlike these workers in other parts of the world, the kholopy were seldom intended for farm work. One reason that slaves and kholopy were rarely found in Russian agriculture could be that masses of serfs performed such functions. Kholopy and serfs therefore appear to have been complementary, and this may have constituted one of the dominant features of Russian history.50 Another reason could be that, when slavery declines, serfdom increases. In this case, the Russian case would resemble the rise of indentured labour in Southeast Asia after the abolition of slavery.51 47  Steinfeld, Invention, p. 11; David Galenson, White servitude in colonial America: an economic analysis, Cambridge: Cambridge University Press, 1981. 48  Pieter C. Emmer, ed., Colonialism and migration: indentured labour before and after slavery, Dordrecht: Martinus Nijhoff, 1986; Klein, Breaking the chains. 49  Steinfeld, Invention; Galenson, White servitude; David Northrup, Indentured labor in the age of imperialism, 1834–1922, Cambridge: Cambridge University Press, 1995. 50  Hellie, Slavery; Hellie, Enserfment. 51  Emmer, Colonialism; Klein, Breaking the chains.

Serfs, Slaves, Or Wage Earners ?

1057

From the empirical standpoint, as we have seen, the kholopy developed from the sixteenth to the end of the seventeenth century, and disappeared in the early eighteenth century. Increasingly harsh peasant servitude was recorded during this period, which would appear to confirm the second hypothesis. However, before leaning towards this conclusion, we should take a closer look at the legal status and the conditions of the mass of peasants. What did it mean to be ‘enserfed’ in Russia at this time?

Serfs or Peasants?

It is necessary to examine carefully the legal conditions to which the rural population in Russia was subjected, starting with the limitations on their movements. Such restrictions were first laid down in 1455–62 for monastery peasants; the provisions were later entered in the sudebnik of 1550, and were thus appropriated by the state.52 They were extended to the peasantry as a whole as part of a particular operation, namely the attempt by Moscow statesmen to establish a land registry.53 Indeed, the adoption of a land registry law in 1592–3 was immediately followed by a prohibition on the peasantry moving, even during the winter months, so that properties and their resident population could be properly identified. The state oversaw compliance with these provisions, and yet authorized considerable movements of peasant families during the first half of the seventeenth century. The archives of lawyers, local institutions, and estates show that a large number of permissions to travel were granted throughout this period, to both individual and families,54 as well as for marriages outside a landowner’s estate.55 Peasants tended to move either from smaller estates to larger ones or to newly annexed regions. In the first case, the owners of large estates challenged the rights of small-estate owners over ‘fugitive’ peasants. In the second case, the tsarist authorities, both central and local, allowed peasant mobility, with a 52  Daniel Kaiser, The growth of law in medieval Russia, Princeton, NJ: Princeton University Press, 1980; Dmitri Grekov, Sudebniki XV–XVII vekov, Moskow: Akademia Nauk SSSR, 1952. 53  Archives of Ancient Russia (henceforth RGADA), pistsovye knigi (cadastral documents) in numerous collections, including: fond 1239, opis’ 3, chast 17, 69–72, 74, 76, 86–7; fond 396, opis’ 2, chast 5 (1616–1732); fond 1209, opis’ 1, chast 1–3, opis’ 2, chast 1–2, opis’ 16–72. 54  Hellie, Enserfment, p. 142; Daniel Morrison, Trading peasants and urbanization in eighteenth-century Russia: the central industrial region, London: Longman, 1987. RGADA, fond 294, opis’ 2. 55  RGADA, fond 615. John Bushnell, ‘Did serf owners control serf marriage? Orlov serfs and their neighbours, 1773–1861’, Slavic Review, 52, 3, 1993, pp. 419–45.

1058

Stanziani

view to promoting colonization. That is why the vast majority of landowners could only seek to take advantage of norms governing ‘fugitives’ in terms of the compensation due. They could not stop mobility but sought to be in a better position to negotiate the conditions of authorizations with other landowners and the tsarist authorities. Provisions concerning peasant mobility were therefore a compromise between the tax and military requirements of the Russian state and the interests of provincial landowners and non-noble elites in the administration.56 It was in this context that the famous Ulozhenie of 1649 was issued, with section 11 containing articles on fugitive peasants. Much of the text is devoted to the documentary evidence that nobles had to provide in order to have peasants returned. On this topic, the regulation does not refer in any way to ownership rights and titles over peasants, as would be the case for serfdom or slavery, but rather to land-registry certification concerning noble estates. As was the case for the regulations adopted from the sixteenth century, this text aimed first and foremost to impose upon nobles state certification of their ownership rights over land. Only on that basis could they have a right to transfer land along with the resident population, and be able to claim labour services and raise credit. This explains why the regulations placed less emphasis on punishing fugitive peasants than on sanctioning nobles who took them in.57 This preoccupation also explains why peasants who settled on an estate continued to sign a contract with the lord. This contract reflected a different legal status for master and peasant, which was a source of inequality, dependence, and a particular form of servitude. The fact that the norms defined the landowners and avoided mentioning ‘serfs’ did not mean that servitude did not exist, but gave it more flexible characteristics. This element was to play a crucial role in the workings and evolution of the Russian rural world. Within the framework provided by these provisions, there continued to be a widespread tendency for landowners to authorize both marriages outside the estate58 and 56  David Moon, ‘Peasant migration and the settlement of Russian frontiers, 1550–1897’, Historical Journal, 40, 4, 1997, pp. 859–93; Williard Sunderland, ‘Peasants on the move: state peasant resettlement in imperial Russia, 1805–1830’, Russian Review, 52, 4, 1993, pp. 472–85; Serguei I. Bruk and Vladimir M. Kabuzan, ‘Dinamika chislennosti i rasselenie russkogo etnosa, 1678–1917’ (‘Quantitative dynamics of Russian ethnic groups, 1678–1917’), Sovetskaya Istoriografiya, 4, 1982, pp. 9–25. 57   Ulozhenie, ch. 11, n. 10. 58  Thousands of certificates were delivered every year: RGADA, fond 615; Bushnell, ‘Serf owners’.

Serfs, Slaves, Or Wage Earners ?

1059

the emigration of entire families, the latter encouraged by the tsarist authorities to further colonization. Movements between the countryside and cities should also be taken into consideration. Here again, it would be a mistake to underestimate the amount of mobility. Quite often, when peasants moved from an estate, it was the individual owner who assigned some of his peasants to workshops and factories. These movements, which occurred rather frequently,59 supported the development of the proto-industrial and industrial sectors in Russia from the seventeenth to the nineteenth century.60 In addition to these recorded and authorized transfers, many forms of mobility were illegal and shadowy. Some peasants did not have an authorization from the landowner, the rural commune, or the rural authorities. Others had the necessary paperwork, but chose to settle in a city without registering or paying taxes. Although local tradesmen protested, the municipalities competed with each other to help workers and small merchants become established. They agreed not to register them officially, provided that they made a fixed payment.61 In short, the norms controlling peasant mobility attempted to accommodate the different interests of small- and large-estate owners, the state, the municipalities, agriculture, industry, and trade. In this context, the title of ‘genuine owner’, and hence the possibility of owning and transferring inhabited estates, became a crucial issue. In this regard, however, the Ulozhenie of 1649 was unable to solve the main problem for which it had been adopted, which was to introduce cooperative agreements among landowners. In reality, disputes intensified,62 and the authorities were hard pressed to enforce compliance with the norms and contracts governing peasant settlement.

59  Governors’ reports detail the regional specialization: RGADA, fond 1281, in particular of St Petersburg area: opis’ 6; Smolensk: opis’ 6; Moscow: opis’ 5; Vladimir: opis’ 4; Kaluga: opis’ 6. 60  Elena I. Indova, ‘O rossiskikh manufakturakh vtoroi poloviny XVIII v.’ (‘On Russian manufacturing during the second half of the eighteenth century’), in Istoricheskaia geografiia Rossii: XIX—nachalo XX v., Moscow: Nauka, 1975, pp. 248–345. 61  RGADA, fond 291, several files; also in Elise Kimerling Wirtschafter, Structures of society, Dekalb, IL: Northern Illinois Press, 1994, p. 181, n. 85. On trading serfs, see RGADA, fond 1287, opis’ 3. 62  RGADA, fond 615 (‘krepostnye knigi mestnyjh uchrezhdenii XVI–XVIII v’—‘register of the deeds of local institutions, sixteenth—eighteenth century’, opis’ 1; fond 294, opis’ 1–3.

1060

Stanziani

This explains why, throughout the eighteenth century, a great number of texts were still seeking to define who was entitled to own and transfer inhabited estates. Conditions were gradually tightened up. A decree of 1730 prohibited servants and peasants from acquiring and owning real estate, whether through inheritance or otherwise.63 Then, in 1739, soldiers and the lower echelons of the administration who did not own estates lost the right to acquire and transfer inhabited estates, or to establish krepost’ relationships with vol’nye liudi (free people). In 1746, this prohibition was widened to include the clergy, merchants, city guilds, Cossacks, and raznochintsy (people of different ranks).64 Finally, several regulations adopted between 1754 and 1758 prohibited non-nobles from owning inhabited estates. These restrictions led to numerous petitions, written by non-noble officers, manufacturers, and merchants. They asked that a distinction be made between the ownership of labourers (urban or rural) and that of entire villages, with only the latter being prohibited to non-nobles.65 Numerous judicial disputes concerning the validity of noble titles also pitted landowners against the administration, and landowners against each other. Such disputes arose when estates were to be transferred, dowries were constituted, or an inheritance needed to be settled.66 Disagreements also emerged when it was envisaged that the rank of ‘noble of the Russian Empire’ should be granted to landowners in newly annexed regions such as Poland. According to the statistics of the Ministry of Justice, in 1845 alone, 6,400 requests for confirmation of noble titles were addressed to it, only half of which were validated.67 Problems of definition of status, and of the permeability of social categories, also held for those who were the objects of noble rights, namely peasants and labourers. The definition of the latter was the mirror image of the definition of nobles and merchants. Hence, the same rules defining the ‘claimants to estate ownership’ were increasingly used not only by nobles and merchants but also by the tsarist administration and by peasants and labourers themselves.

63  PSZ, series I, vol. 8, no. 5633. See also Blum, Lord and peasant, pp. 358–62. 64  PSZ, series I, vol. 12, nos. 9332, 9367. 65  François-Xavier Coquin, La Grande commission législative 1767–1768: les cahiers de doléances urbains, province de Moscou, Paris: Publication de la Faculté des Lettres et Sciences Humaines de Paris-Sorbonne, 1972, pp. 110 and 161–3. 66  Moscow Archives (henceforth TsGIAM), fond 54. See also Wirtschafter, Structures, pp. 71–4. 67   Otchet ministerstvo iustitsii za 1845, St Petersburg, 1846, p. xix.

Serfs, Slaves, Or Wage Earners ?



1061

The Change of Legal Status: An Administrative Act versus Judicial Proceedings

The appropriation of the rules of law by a variety of economic and social agents has been the subject of an impressive number of works in history, law, sociology, and anthropology. The analysis of judicial conflicts has helped renew the history of indentured service68 and of slavery.69 Slaves, runaway slaves, and indentured servants made considerable use of the law, challenging the idea of unlimited dependence. Legal pluralism gave these systems more flexibility, and has led scholars to consider abolition less as a break with the past than as a step in a long-term process. For tsarist Russia, a growing interest in law and its use by economic and social groups has been almost entirely limited to the period from 1864, following legal reforms and the introduction of a genuine hierarchical system of justice.70 The analysis of the law during the period of ‘serfdom’, in contrast, remains virtually unknown, and this gap needs to be filled. The materials exist, as we have a significant number of documents concerning judicial disputes involving both kholopy and peasants. It is possible to distinguish two main tendencies in the disputes concerning kholopy, one involving several ‘claimants to title’ and the other between such claimants and kholopy. In cases involving several claimants, the main issue concerned those who claimed to have established a kholopstvo contract 68  Steinfeld, Invention. 69  Lauren Benton, Law and colonial culture, Cambridge: Cambridge University Press, 2002; Michael Craton, Empire, enslavement and freedom in the Caribbean, Kingston, Jamaica: Ian Randle, 1997; Marc Galanter, Law and society in modern India, Delhi: Oxford University Press, 1989; Douglas Hay and Paul Craven, Masters, servants and magistrates in Britain and the Empire, 1562–1955, Chapel Hill, NC: University of North Carolina Press, 2004; Alan Watson, Slave law in the Americas, Athens, GA: University of Georgia Press, 1989. 70  Wirtschafter, Structures; Virginia Martin, Law and custom in the steppe: the Kazakh of the Middle Horde and Russian colonialism in the nineteenth century, Richmond: Curzon Press, 2001; Jane Burbank, Russian peasants go to court: legal culture in the countryside, 1906–1917, Bloomington, IN: Indiana University Press, 2004; Kritika, special issues, 6, 1, 2005 and 7, 1, 2006. Russian historiography offers more on the pre-emancipation period: Natalia N. Efremova, Sudoustroistvo Rossii v XVIII—pervoi polovine XIXe v (The judicial organization of Russia, eighteenth century to first half of the nineteenth century), Moscow: Nauka, 1993; Ekaterina A. Pravilova, Zakonnost’ I prava lichnosti: administrativnaia iustitsiia v Rossii, vtoraia polovina XIX v.–oktiabr’ 1917 (Legality and the rights of the person: administrative justice in Russia, second half of the nineteenth century to October 1917), St Petersburg: SZAGS, 2000.

1062

Stanziani

in good faith with someone who had previously signed one with another master. Such an individual was legally a ‘fugitive’. In the early sixteenth century, the Russkaia pravda (article 118) stated that the first claimant to rights could recover the fugitive, but had to compensate a buyer who had acted in good faith. However, the Sudebnik of 1550 adopted the principle of caveat emptor: the buyer of a title over a kholop could not be usually compensated, especially if he had been negligent.71 Finally, the Ulozhenie of 1649 returned to the earlier principle. In every case, written documents were required to prove the validity of a plaintiff’s claims. There were also disputes between those who claimed rights over people and those in a situation of obligation, who might object to the original obligation, or to the terms of its cancellation. These conflicts were so numerous that a kholopii prikaz (chancellery) was set up in the seventeenth century to resolve issues of this kind.72 Among the most frequent disputes were those concerning types of kholopstvo. The prohibition against hereditary kholopstvo towards the end of the seventeenth century did not in fact put an end to this practice. Many cases were brought before the court at this time, by the kholopy himself, often by the children of a kholopy, or by new masters who were claiming their rights. These disputes confirm that it was not impossible for the kholopy to win a case, although the chances were slim compared with those of ‘claimants to title’. At the same time, this use of rights was possible because it intersected with the interests of other lords, other claimants over kholopy, or of the state itself, for the reasons mentioned above. These conclusions also apply to peasants, at least from the late eighteenth century. While it is true that peasants did not initially have the right to sue the landowner, this situation was changing well before the ‘great reforms’ of 1861–4. In 1770, peasant courts were set up. To be sure, these courts had limited powers, and were under the control of local nobles. Moreover, the peasants were still not allowed to sue nobles over matters of corporal punishment or work organization. However, the validation of ownership rights, and therefore of peasant obligations, was easier to establish, insofar as these disputes were usually multilateral, implicating other nobles, family members (heirs, cousins, and so forth), and the tsarist administration, which was itself interested in defining 71  Hellie, Slavery, pp. 194–8; Russkaia istoricheskaia biblioteka, 17, 1898, pp. 106–7, nn. 298–9. 72  A. K. Leont’ev, Obrazovanie prikaznoi sistemy upravleniia v russkom gosudarstve. Iz istorii sozdaniia tsentralizovannogo gosudarstvennogo apparata v kontse XV—pervoi polovine XVI v. (The formation of a chancellery system in the Russian state: history of the formation of the centralized state, fifteenth—sixteenth century), Moscow: Moskovskii Universitet, 1961, pp. 179–92.

Serfs, Slaves, Or Wage Earners ?

1063

ownership rights. These multiple interests explain the adoption between the end of the eighteenth century and 1861 of numerous laws aimed at facilitating the submission of cases to these courts, and challenges of ownership rights to ‘inhabited estates’. These provisions were initially designed to resolve disputes within noble families, as well as between different categories of nobles, or even between nobles and merchants, but they came to be appropriated by peasants and workers in workshops and factories. These developments testified to the peasants awareness of the numerous regulations governing the ownership of estates and serfs. For example, in many cases, the plaintiffs emphasized that the transfer of the estate took place through the intermediation of a non-noble, which was a prohibited practice. In other cases, the plaintiffs demonstrated that the landowner was not a noble but a merchant, and, as such, prohibited from ownership.73 The tsarist authorities even improved the ability of the common people to challenge the ownership rights of their masters, by adopting several norms favourable to their cause between 1801 and 1858. There were a number of reasons for this attitude: the need for political stability by avoiding peasant unrest and limiting conflicts between nobles; paternalistic criticism of the obligations imposed on peasants; and the desire of tsarist statesmen to facilitate land ownership by bourgeois elements and ‘service nobles’.74 Proceedings instituted by peasants became so numerous that, between 1837 and 1840, the senate even decided to call a halt to cases where serfs were still living with their master.75 Overall, between 1833 and 1858, the senate recorded 15,153 cases of illegal estate ownership, and thus of illegal ‘servitude’. These cases were recorded in the anthology of laws and jurisprudence, and the details were kept in the senate archives. For local courts, only partial estimates are possible at present, with 22,000 known cases of this type during the same period.76 In addition to this figure, it is necessary to include all the peasants

73  TsGIAM, fond 54 (Moskovskoe gubernskoe upravlenie), 1783–1917, opis’ 1: for example, delo 56, 284, 966, 1509. Rossiskie Gosudarstvennoie Imperialskie Arkhivi (Russian Imperial Archives—henceforth RGIA), fond 1149, opis’ 2, delo 20 and delo 44. Gosudarstvennie Arkhivi Rossiskoi Federatsii (State Archives of the Russian Federation—henceforth GARF), fond 109, opis’ 3, delo 1885. 74  Moon, Abolition; Blum, Lord and peasant; Hoch, Serfdom; Daniel Saunders, Russia in the age of reaction and reform, 1801–1881, London: Longman, 1992. 75  RGIA, fond 1149, opis’ 2, delo 90. See also Wirtschafter, Structures, p. 84. 76  PSZ, series II, vol. 20, no. 19283, vol. 22, no. 20825; RGIA, fond 1149, opis’ 3, delo 125.

1064

Stanziani

who changed status following a unilateral act by the landowner.77 It is difficult to gauge the significance of these figures, as there is no systematic study available that is broken down by province. According to estimates of the period, at the ninth reviziia (tax census) in 1851, in twelve provinces, eleven thousand meshchane (merchants) were former private peasants.78 Such favourable outcomes should not make us forget the problems that peasants encountered when they tried to institute proceedings against a real or self-proclaimed noble. Quite often, local courts handed down totally different rulings. Several judges considered peasant petitions to be inadmissible, and refused to grant them recourse to an appeal.79 Many cases of noble landowners bribing judges were also recorded. Finally, pre-trial investigations were lengthy, and it was often a decade before a case came to trial. Measures aimed at changing this state of affairs were not adopted until the end of the 1840s, when a new law facilitated judicial proceedings for all those who considered that their obligations to an estate owner were illegal.80 In other words, eighteenth- and nineteenth-century Russia was far from resembling the Britain and France of the nineteenth- and twentieth-century; but nonetheless it was also not the ‘land of serfdom’ that is usually represented. This conclusion becomes all the more striking if we take into consideration the fact that changes in the status of ‘private’ peasants also took place by administrative action, notably through military service. Once a conscript had completed his military service of twenty years’ duration, he entered the category of raznochintsy (people of different ranks, belonging to the urban population). He could therefore move about freely, and settle in a city. Hoch and Augustine estimate that 433,750 peasants changed their legal status in this manner between 1833 and 1858 alone.81 Other forms of administrative reclassification of peasants from private estates took place for political reasons. Thus, the authorities reclassified peasants of nobles who took part in acts of ‘sedition’ in Poland, Ukraine, Belarus, and the ‘western provinces’ between 1838 and 1849. Taking these regions as a whole, an estimated 264,000 peasants were seized, includ77   Svod zakonov rossiskoi imperii (Collection of laws of the Russian Empire), St Petersburg, 1832, vol. 9, art. 674–80, 1833, art. 1148–84, 1857. 78  P. V. Keppen (P. V. Köppen), Deviataia reviziia: issledovanie o chisle zhitelei v Rossii v 1851 godu (The ninth census: study on the population of Russia in 1851), St Petersburg, 1857, pp. 6, 7, 21, 88, 95–100, 127, 142–4, 152, 159. 79  RGIA, fond 1149, opis’ 2, delo 20. See also Wirtschafter, Structures, pp. 79, 119. 80  Law of 1847, in PSZ, series II, vol. 22, no. 20825. 81  Steven Hoch and Wilson Augustine, ‘The tax census and the decline of the serf population in imperial Russia, 1833–1858’, Slavic Review 38, 3, 1979, pp. 403–25.

Serfs, Slaves, Or Wage Earners ?

1065

ing 72,500 who became peasants of the state. Finally, between 1803 and 1858, numerous laws were passed to facilitate the change of legal status of peasants for more strictly economic and social reasons. Changes occurred in cases of nobles without heirs, mortgaged estates, purchase of land by peasants, and so forth. In all, about a million peasants changed categories between 1800 and 1858. Half of private peasants were re-assigned to other categories, to such an extent that, in 1858, only 40% of peasants were classified as ‘private’ peasants, and only half of those were still subject to labour obligations.82

Legal Status, Labour, and the Dynamics of the ‘Second Serfdom’

These elements lead us to more general conclusions, first of all concerning the reforms of 1861. From our perspective, these reforms should be viewed as part of a long-term process. Rather than marking a sharp break, passing from serfdom to free labour, they can be seen as just one step in a long process of transforming the legal and economic status of labour in Russia. The reforms of that year in many ways reflected measures already adopted towards peasants in the western regions of the empire, along with the regulations of 1803 and 1841, aimed at emancipating the peasants from any obligation to landowners, while granting them a plot of land.83 The main difference was that the reforms of 1861 were imposed administratively, and across the board. At the same time, the state financed a good portion of these operations, while contributing to setting a relatively high price for the land granted to peasants, through the much-debated work of local arbitration commissions. Significant continuities also emerged in the labour market. While peasants were formerly not ‘serfs’ in the traditional sense of the term, after 1861 they were not free wage earners in the classic sense of the term. Internal passports were still required, and failure on the part of wage earners and peasants to comply with their contractual commitments was punishable by the criminal code. Finally, labour obligations continued to be heavy when peasants obtained land or entered into loan contracts, such as labour services in exchange for an advance of seeds, or the loan of tools.84

82  Hoch and Augustine, ‘Tax census’; Moon, Abolition. 83  Field, End of serfdom, pp. 77–83. 84  Hoch, Serfdom, Peter Gatrell, The tsarist economy, 1850–1917, London: Longman, 1986; Moon, Abolition; Paul Gregory, Before command: an economic history of Russia from emancipation to the first five-year plan, Princeton, NJ: Princeton University Press, 1994.

1066

Stanziani

One important implication of this conclusion is that the schema for breaking down forms of coerced labour was quite similar to ones experienced in other historical contexts, as was also the case for ‘entering into serfdom’ and the ways in which ‘serfdom’ operated in Russia. This evolution entailed a sequence of legal and legislative decisions, as well as an appropriation of legal rules by workers themselves. Russian peasants contributed to their own emancipation by relying on norms theoretically designed for other purposes. The tsarist autocracy was thus a far more flexible system than is usually asserted, even if it was neither egalitarian nor democratic. These conclusions coincide with those recently established with regard to other parts of central and eastern Europe, notably Brandenburg and Bohemia. Contrary to traditional interpretations, the period from 1650 to 1800 is now seen as having been distinguished by a transformation of the Gutsherrschaft and of the legal status of peasants, well before the reforms of the early nineteenth century.85 ‘Unlimited labour service’ was restricted in significant ways, and gave rise to numerous disputes between peasants and lords, as trials in local courts beginning in the seventeenth century testify. The enlightened autocrat Frederick II (1740–86) adopted measures to reduce or even eliminate hereditary forms of dependence, widened access for peasants to courts of justice, reduced labour obligations, and recorded the obligations of peasants towards lords in contracts that could, on occasion, be used as evidence.86 Similar points can be made about other regions of central and eastern Europe. For example, about half of Polish households towards the end of the sixteenth century had servants.87 In Schleswig-Holstein, Mecklenburg, Swedish Pomerania, and Lusatia, numerous forms of personal dependence were widespread.88 At the same time, the line between ‘free peasants’ and ‘serfs’ was qualified at the time when quitrent replaced labour services in Prussia. As in Russia, the authorities

85  Edgar Melton, ‘The decline of Prussian Gutsherrschaft and the rise of the Junker as rural patron’, German History, 12, 1994, pp. 334–50; William Hagen, ‘Village life in East-Elbian Germany and Poland, 1400–1800’, in Tom Scott, Peasantries, pp. 145–90; Sheilagh Ogilvie, ‘Communities and the second serfdom in early modern Bohemia’, Past and Present, 187, 2005, pp. 69–119. 86  Hagen ‘Village life’, p. 149. 87  Robert Frost, ‘The nobility of Poland-Lithuania, 1569–1795’, in Hamish Scott, ed., The European nobilities in the seventeenth and eighteenth centuries, vol. II: Northern, Central and Eastern Europe, London: Routledge, 1995, pp. 183–222; Hagen, ‘Village life’; W. Hagen, ‘Capitalism and the countryside in early modern Europe: interpretations, models, debates’, Agricultural History, 62, 1988, pp. 13–47. 88  Hagen, ‘Village life’, p. 175.

Serfs, Slaves, Or Wage Earners ?

1067

at the time encouraged a change in the legal status of peasants.89 Judicial disputes grew in number and, already during the first quarter of the eighteenth century, it was not uncommon for peasants to submit cases to the Berlin Court of Appeals.90 These observations lead us to the following general conclusion, that the ‘second serfdom’ is a phenomenon that deserves to be reassessed from both an institutional and economic standpoint. Far from being a mere copy of the serfdom of Europe’s Dark Ages, these forms of restriction on mobility were part of a context in which the constraints weighing upon a large percentage of the population (especially in rural communities), although considerable, nevertheless allowed some leeway to those involved. The legal statuses of peasants, labourers, and noble landowners were mutually defined, which meant that their respective rights and obligations were not absolute, but changed over time according to the particular estate and region concerned. But if that was the case, then where should we draw the line separating the Russian labour system from the forms of labour encountered in western Europe and its colonies?

Conclusion: Forms of Servitude and Institutional Dynamics

Russia was not exceptional from the point of view of slavery in the strict sense. Slavery was rarely inflicted on Russians themselves, but it was widely found in relation to prisoners of war and ethnic minorities, some of whom in turn practised slavery within their own societies. Russia was a society with slaves, but not a slave society. The kholopy were numerous, and subject to clearly marked forms of dependence, but we would have trouble identifying them as slaves, at least once the hereditary element had been eliminated. These forms of dependence often came to resemble indentured service and debt bondage. There was hardly any difference between loan and apprenticeship contracts and kholopstvo. Even if the real conditions of these labourers could hardly be distinguished from those of wage earners in the strict sense, or even slaves, their different legal status was important. It brought out complex relationships of continuity, rather than 89  Hartmut Harnisch, ‘Bäuerliche Ökonomie und Mentalität unter den Bedingungen den ostelbischen Gutsherrschaft in den letzten Jahrzehenten vor Beginn der Agrarreformen’, Jahrbuch für Wirtschaftsgeschichte, 24, 3, 1989, pp. 87–108. 90  Hartmut Harnisch, Kapitalistische Agrarreform und Industrielle Revolution, Weimar: Böhlau, 1989.

1068

Stanziani

breaking with the past, and challenges any simple opposition between servitude, slavery, and ‘free’ labour. This general principle was also valid for the British and French cases. British indentured service was defined using ‘ordinary’ contracts, Masters and Servants Acts, and apprenticeship; while French engagement rested on the renting of services and apprenticeship. These contracts also provided for criminal sanctions, the recovery of ‘fugitives’ by the forces of the state, and a different legal status being applied to master and servant. Hence, British indentured service and French engagement formed two variants of one and the same legal notion of labour that embraced the kholopy. If neither kholopy nor slaves constituted a Russian ‘specificity’, should we look for this in serfdom? In reality, the disappearance of kholopy, or rather their merger with the peasants after 1725, in no way entailed the enslavement of the latter. Serfdom and servitude did not correspond to the stereotypical image that we have had of them since the eighteenth century. The ‘serfs’ were never defined as such in tsarist law, except in the years immediately preceding the 1861 Emancipation Act. The system that has been called the ‘second serfdom’ in eastern Europe and Russia was therefore a form of dependence that was at once statutory and contractual, although never institutionalized as such, and which gradually weakened in the eighteenth and nineteenth centuries. These forms of dependence were no doubt harsher than those experienced by peasants in western Europe: a good portion of the peasants on Russia’s private estates never changed estates, and lived under hard conditions of submission. Nevertheless, these peasants did have rights, however reduced they may have been: they could sue the landowners in a court of law, and challenge their right to own land. With the help of the Tsarist authorities, the legal and economic possibilities emerging from these rights increased during the first half of the nineteenth century. Russia might therefore be defined as an extreme form of a more general model. This consisted in limiting the movements of labourers and peasants, subjecting them to penal sanctions, and imposing on them a different legal status from that of their masters and employers. Such a model was widespread in the western world, including its colonies, from the seventeenth to nearly the end of the nineteenth century. Many categories of workers had fewer rights and legal and economic assets than ‘free’ labourers, who in turn definitely had fewer rights at the time than those attributed to them in the twentieth century. Not only the indentured servant but also the domestic servant had a different legal status than that of the employer. Russia may have constituted an extreme case, but this was in a world in which, from the standpoint of ‘freedom’, the range of forms of labour expressed a continuity. There were subtle gradations, rather than an outright opposition, between free and forced labour.

I Make Him My Dog/My Slave Brett Rushforth On April 12, 1680, a Belgian monk-turned-missionary named Louis Hennepin tinkered with a canoe on the banks of the Mississippi River. As two French servants boiled a wild turkey for his lunch, Hennepin surveyed the strange and beautiful country before him. His party had traveled the Mississippi for eleven days without incident, but, as he awaited his meal, Hennepin “suddenly perceived … fifty bark canoes, conducted by 120 Indians, entirely nude, who descended this river with great speed.” Hennepin called out to them, twisting his tongue around rudimentary Algonquian to assure the Indians of his good intentions. “Mistigouche,” he cried, using the Algonquians’ name for his people to identify himself and his servants as their friends. As Siouan speakers, the approaching war party did not understand his words. But, unfortunately for Hennepin, they got the message: these bearded foreigners were allies of the Algonquian-speaking peoples of the Mississippi Valley, the very peoples the Sioux had come to attack.1 Source: Rushforth, Brett, “I Make Him My Dog/My Slave,” in Brett Rushforth, Bonds of Alliance: Indigenous and Atlantic Slaveries in New France, Chapel Hill: University of North Carolina Press, 2012, 35–72. Published for the Omohundro Institute of Early American History and Culture. Copyright © 2012 by the University of North Carolina Press. Used by permission of the publisher. www.uncpress.unc.edu. 1  “J’apperçus tout d’un coup … cinquante Canots d’ecorce conduits par six vingt Sauvages tous nuds, qui décendoient d’une fort grande vitesse sur ce Fleuve.” Louis Hennepin, Nouvelle découverte d’un très grand pays situé dans l’Amérique, entre le Nouveau Mexique et la mer glaciale (Utrecht, 1697), 314–315. Catherine Broué discusses problems with Hennepin’s credibility, concluding that the 1697 edition is largely reliable and highly valuable. Broué, “En filigrane des récits du Père Louis Hennepin: ‘trous noirs’ de l’exploration louisianaise, 1679–1681,” Revue d’histoire de l’Amérique française, LIII (1999–2000), 339–366. For the negative view, see Jean Delanglez, Hennepin’s Description of Louisiana: A Critical Essay (Chicago, 1941). Perhaps the foremost authority on seventeenth- and eighteenth-century Sioux history and culture, Raymond J. DeMallie concludes that Hennepin’s writings, if evaluated carefully, “present valuable ethnographic detail” about the eastern Sioux. DeMallie, “The Sioux at the Time of European Contact: An Ethnohistorical Problem,” in Sergei A. Kan and Pauline Turner Strong, eds., New Perspectives on Native North America: Cultures, Histories, and Representations (Lincoln, Nebr., 2006), 243.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_035

1070

Rushforth

Hennepin and his party quickly realized the danger and scrambled to evade an impending assault. Ditching the turkey in the brush, the servants ran to the canoe, joining Hennepin in a hasty retreat. Within seconds, Sioux canoes surrounded them. Raising ceremonial war cries, the attackers boarded Hennepin’s canoe and took him captive. “We offered no resistance,” Hennepin later recalled, “because we were only three against so great a number.” Now using signs because he “did not know a word of their language,” Hennepin tried to urge the Sioux on to their original target, but to no avail. Next he offered bribes, first tobacco from Martinique, then two wild turkeys they had saved for dinner. This pleased his captors, and their demeanor seemed to soften, but by nightfall Hennepin and the Frenchmen still feared for their lives. The servants resolved to die fighting like men, but Hennepin was more resigned to his fate, whispering a vow that he would “allow them to kill me without resistance in order to imitate the Savior, who gave himself voluntarily into the hands of his executioners.”2 Rather than dying a martyr, Hennepin lived the next eight months as a captive among the Sioux. For nineteen days he and his companions were forced to row their overburdened canoe against the Mississippi’s strong current. Reaching the northern edges of navigable waters, the Sioux destroyed Hennepin’s canoe to prevent his escape and then marched the prisoners over half-frozen marshlands toward their villages. The prisoners faced daily threats to their lives, enduring “hunger, thirst, and a thousand outrages … marching day and night without pause.” When Hennepin’s hunger and fatigue caused him to lag, his captors set fire to the meadows behind him, forcing him to push

 Throughout this study I use the term “Sioux” rather than the recently fashionable “Dakota” because the latter term excludes those Sioux who are Lakota or Nakota and because “Sioux” is a much more widely recognized term among Anglophone readers. French sources from the seventeenth and eighteenth centuries do not allow a clear distinction between various Sioux bands, so the broader term also better reflects the historical record. Several modern tribal organizations in the United States use “Sioux” in their official names, but none uses “Dakota” except to designate their location. See Raymond J. DeMallie, “Sioux until 1850,” in Handbook, XIII, Plains, part 2, 718; and, for a different perspective, Gary Clayton Anderson, Kinsmen of Another Kind: Dakota-White Relations in the Upper Mississippi Valley, 1650–1862 (Lincoln, Nebr., 1984). 2  “Nous ne faisons aucune resistance, parce que nous n’êtions que trois contre un si grand nombre,” 316; “Je ne savois pas un mot de leur langue,” 320; “J’avois resolu de me laisser tuer sans resistance afin d’imiter le Sauveur, qui s’étoit remis volontairement entre les mains de ses bourreaux,” 319: Hennepin, Nouvelle découverte.

I Make Him My Dog / My Slave

1071

ahead. In short, according to Hennepin, “The insults that these barbarians committed against us during our journey are beyond all imagining.”3 When they reached the Sioux villages, the prisoners faced another wave of humiliating assaults. They were stripped naked, their bodies were painted, and they were forced to sing and dance as they drummed a rattling gourd. The warriors stood the prisoners in front of tall stakes, set in the ground and surrounded by straw and wood, erected to burn incoming captives. Then they began to negotiate Hennepin’s fate. A few urged good treatment to curry favor with the French, but their voices were overwhelmed by those arguing for his execution. They began to torture him but were cut short when an influential war chief named Aquipaguetin stepped forward and claimed the priest as a replacement for his son, who had fallen at the hands of Hennepin’s allies. In Aquipaguetin’s charge Hennepin entered the chief’s village, injured and demoralized but glad to be alive.4 As he made the transition to life in captivity, Hennepin found it difficult to make sense of his position in Sioux society or even to find the right word to define it. His captors themselves said that “they considered [him] a slave that their warriors had captured in their enemies’ territory.” Yet, because of his ceremonial adoption as Aquipaguetin’s son, he expected a level of independence and respect he never achieved. He was beaten. He faced repeated death threats. He performed forced labor, farming with Aquipaguetin’s wives and children on a nearby island. And he was under almost constant surveillance: “The more I hid myself, the more I had Indians after me … for they never stopped watching me.” But it was hunger that troubled him the most. To keep him weak and dependent, Hennepin’s newly adopted kin fed him only five or six meals a week, giving him just enough wild oats and fish eggs to keep him alive. “I would have been very content had they given me something to eat, as they did their children,” he remembered. “But they hid [their food] from me … conserving what little fish they had to feed their children.” Despite the metaphorical kinship conferred by his adoption, he and his captors understood the difference between real and fictive sons. “They thus preferred the lives of their children to mine,” a distinction that even Hennepin admitted was only reasonable. If the priest could not fully grasp his experience as a Mississippi Valley

3  “Le faim, la soif, et mille outrages … marché jour et nuit sans delai,” 342; “Les Insultes, que ces Barbares nous firent pendant nôtre route, sont au dessus de toute imagination,” 322: Hennepin, Nouvelle découverte. 4  Ibid., 355.

1072

Rushforth

captive, he was eager for it to end: “It must be said that it is a sweet and pleasant thing to come out of slavery.”5 Looking back from the twenty-first century, historians share Hennepin’s dilemma, finding no easy shorthand to encapsulate so complex and unfamiliar an experience. Not only did Indians leave no written records of their own, but slaves’ marginalization also made them less visible to outside observers, especially in societies like these that did not rely on slavery as a dominant mode of social or economic production. Even those who, like Hennepin, survived the ordeal of enslavement had only a fragmentary understanding of its meaning to their captors. Full of silences, their narratives often raise as many questions as they answer. This is particularly true of published accounts, like the Jesuits’ insightful but carefully scripted Relations, which functioned as public performances for a particular French audience and thereby obscured many insights that modern scholars wish to glean from them. Reconstructing the practice of indigenous slavery requires a wider reconnaissance, gathering scattered shards of surviving evidence and relying on context to fill in the gaps that remain. Using archaeology, ritual, and linguistics to supplement the observations of captives and other colonists, it is possible to trace the contours of indigenous slavery among the central Algonquian and Siouan peoples of the Pays d’en Haut. These sources take us beyond sketchy European impressions of Indian captivity and adoption to see the patterns of enslavement, domestication, and forced integration by which enemy outsiders were made into subordinate domestics. A rich body of linguistic records, in particular, reveals how Indians explained the place of slavery in their own societies. In the late-seventeenth century, Jesuits compiled and translated thousands of pages of Algonquian phrases, recorded from conversations with Native men and women and 5  “Elles me consideroient comme un Esclave, que leurs Guerriers avoient fait dans le pays de leurs Ennemis,” 362; “plus je me cachois, plus j’avois de Sauvages à ma suite … car ils ne me quittoient point de veuë,” 320–321; “J’aurois esté fort content, s’ils m’eussent donné à manger, comme à leurs enfans. Mais ils se cachoient de moy … conservoient le peu de poisson, qu’elles avoient, pour en nourrir leurs enfans…. Elles préféroient donc la vie de leurs enfans à la mienne. En quoy il est certain, qu’elles avoient raison,” 362; “Il Faut avoüer, qu’il est bien doux et bien agreeable de sortir de l’Esclavage,” 463: Hennepin, Nouvelle découverte. For mistreatment and labor, see Louis Hennepin, Description de la Louisiane, nouvellement découverte au Sud’Ouest de la Nouvelle France (Paris, 1683), 246; Hennepin, A New Discovery of a Vast Country in America, Extending above Four Thousand Miles, between New France and New Mexico (London, 1698) (Wing H1451), I, 109. William Henry Foster similarly argues that the power of Native women to control slaves “came not from physicality but from the hearth.” Foster, The Captor’s Narrative: Catholic Women and Their Puritan Men on the Early American Frontier (Ithaca, N.Y., 2003), 9.

I Make Him My Dog / My Slave

1073

corrected over several years by Native informants. As transcriptions of Indians’ own speech in their own language, these manuscripts are the nearest thing to contemporary Algonquian sources available, providing far greater clarity than the French observations that have to this point guided our understanding of captivity in the Pays d’en Haut. Because the words and phrases have a linguistic integrity and logic of their own—and because there are hundreds of examples of their connotations and contextualized meanings—they provide unprecedented depth to our understanding of indigenous slavery.6 The Indians who would engage in a century-long slave trade with French colonists brought their own complex and evolving practice of slavery to the colonial encounter. The act of enslavement dominated and defined Natives’ thinking about slavery far more than the long-term status of those they enslaved. Indigenous slaves lived under a wide range of conditions, some dehumanizing and others nearly familial, and a particular slave’s place in the community could change over time. But all of them had to pass through the ritualized system of enslavement designed to strip them of former identities and forcibly integrate them into the capturing village. In war dances and diplomatic ceremonies, through the binding and marking of bodies, and with a sophisticated language of dominion and ridicule, the Native peoples of the Pays d’en Haut articulated an elaborate idiom of slavery as a form of human domestication that reduced enemy captives to the status of dogs and other domesticated animals. Simultaneously expressing and seeking power, enslavement involved a series of scripted acts of physical and psychological dominion designed, in the words of several Algonquian and Siouan languages, to tame and domesticate captured enemies. In so doing, captors harnessed the enemy’s power to serve the needs of their own people. Although slaves were defined by their place in Native war culture, they also labored in agriculture and performed other useful tasks. For men, this work often violated gender norms, as they were compelled to perform traditionally female tasks like hoeing in the fields or carrying baggage on hunting expeditions. Female slaves often became subordinate wives, adding their reproductive and domestic labor to the households that incorporated them. Full of possibilities for social integration, enslaved women’s work also carried many dangers, including the potential for sexual violence that seems to have been a hazard unique to their slave status. Some of enslaved individuals’ most important labor was performed in the area of diplomacy. As both agents and objects of intercultural relations, indigenous slaves mediated between the 6  For a fuller discussion of the origin and nature of these dictionaries as well as a list of translated Algonquian terms relating to slavery and captivity, see Appendix A.

1074

Rushforth

violent impulses that led to their enslavement and the alliance building that their bodies facilitated as symbols of generosity. As a regionally and temporally specific system of human bondage, Algonquian and Siouan slavery differed in important ways not only from European chattel slavery but also from other forms of Indian captivity in North America.

The General Custom of the Country Our Outagami are fully resolved not to treat The Nadouessi more humanely than they are treated by them. Claude Allouez, Jesuit, 1672

The Pays d’en Haut was a biologically and climatically diverse region loosely bordered by Lake Huron on the east, the Minnesota River on the west, and the confluence of the Missouri and Mississippi Rivers on the south, comprising roughly the modern states of Minnesota, Wisconsin, Michigan, Indiana, and Illinois as well as parts of eastern Iowa and Missouri and western Ontario (see Map 1). Linked by a network of rivers and lakes—especially the Mississippi River and Lakes Michigan and Superior—the Native peoples in this region had a long history of interaction through trade, intermarriage, and warfare that bound their villages in a regional political economy that balanced the imperatives of trade with the harsh realities of intergroup competition and warfare. Slavery and slave raiding had a deep history in the Pays d’en Haut, expressed most fully and brutally by the people of Cahokia, who dominated the midMississippi from the tenth to fourteenth centuries. Centered on the eastern bank of the Mississippi River across from modern Saint Louis, Missouri, Cahokia had a population estimated between twenty and forty thousand, managing a system of subordinate alliances that drew goods and peoples from all across the midcontinent. Intensely hierarchical, Cahokia practiced slavery on a large scale. The presence of massive public earthworks demanding an ample supply of manual laborers, combined with evidence of ritual human sacrifice, indicates a tradition of slavery that supported the ruling elite and sustained the hierarchies that ordered this world and the next. Many slaves were killed for burial with their masters, interred with other prestige goods to honor and comfort the dead. In one dramatic example (discovered by archaeologists in the late-1960s), an elite man was buried with more than twenty thousand shell beads, flanked by a burial pit containing fifty-three murdered women and four men with severed heads and hands. The confluence of monumental public works with “theatrical” human sacrifice points, in the words of

I Make Him My Dog / My Slave

Map 1

1075

The Native Peoples of the Pays d’en Haut. Drawn by Jim DeGrand.

one archaeologist, to “the paramount importance of human labor and the ability to coordinate, control, and sacrifice it.” A long-standing tradition of slave raiding and trading permeated Mississippian cultures well beyond Cahokia too, running the length of the Mississippi Valley and into the mound-building chiefdoms of the Southeast.7 Echoes of Cahokian slave raiding reverberated long after the city’s collapse in the early fourteenth century. Archaeological remains suggest that, from the decline of Cahokia to the mid-seventeenth century, the Native villages of the Pays d’en Haut both suffered and conducted a nearly unbroken succession of 7  Timothy R. Pauketat, Ancient Cahokia and the Mississippians (Cambridge, 2004), 92–93 (“paramount importance”); Pauketat and Thomas E. Emerson, eds., Cahokia: Domination and Ideology in the Mississippian World (Lincoln, Nebr., 1997); Jon Muller, Mississippian Political Economy (New York, 1997); Emerson, Cahokia and the Archaeology of Power (Tuscaloosa, Ala., 1997); Christina Snyder, Slavery in Indian Country: The Changing Face of Captivity in Early America (Cambridge, Mass., 2010), 13–45.

1076

Rushforth

raids and counterraids, what anthropologists flatly describe as low-level endemic warfare punctuated by periods of heightened violence. Of all burials uncovered from the late-precolonial period, for example, a significant proportion of all deaths were caused by war instruments like clubs, stone daggers, and arrowheads. Skeletal remains show signs of torture that correspond with later descriptions of Native captive rituals: scalping, finger mutilation, multiple blows from clubs, and live burning. Most of these torture victims were also adults and more often male than female, following later patterns of women’s and especially children’s being kept alive while men were killed. Skulls from several sites in the Pays d’en Haut exhibit healed cranial fractures consistent with nonfatal blows from war clubs. In some samples females with these injuries outnumbered males as many as four to one, whereas males had far more fatal injuries. Relatively few child remains exhibit any signs of torture or violent death. Evidence of warfare beyond the Pays d’en Haut confirms what later sources describe as a long history of raiding into bordering regions, especially the eastern Great Plains. Villages in the Missouri River valley, which would be the target of Illinois, Fox, and Sioux raids in the seventeenth and eighteenth centuries, suffered frequent raids in preceding generations, especially during periodic resource shortages traced to climate change. Summarizing a regionwide review of archaeological data from the fourteenth through sixteenth centuries, one team of researchers concluded that “outbreaks of violence occurred regularly, and each attack resulted in only a few deaths…. Internecine conflict … was a pervasive element in everyday life.” Bones are poor guides to the cultural meanings of this violence, but some form of slavery evidently played an important role.8

8  For forensic markers of slavery, see Richard D. Wilkinson and Karen M. Van Wagenen, “Violence against Women: Prehistoric Skeletal Evidence from Michigan,” Midcontinental Journal of Archaeology, XVIII (1993), 190–216; Debra L. Martin, “Ripped Flesh and Torn Souls: Skeletal Evidence for Captivity and Slavery from the La Plata Valley, New Mexico, ad 1100– 1300,” in Catherine M. Cameron, ed., Invisible Citizens: Captives and Their Consequences (Salt Lake City, Utah, 2008), 159–180; Douglas B. Bamforth, “Climate, Chronology, and the Course of War in the Middle Missouri Region of the North American Great Plains,” in Elizabeth N. Arkush and Mark W. Allen, eds., The Archaeology of Warfare: Prehistories of Raiding and Conquest (Gainesville, Fla., 2006), 66–100. The quality and range of the archaeological evidence from the mid-Missouri River sites represent, according to Bamforth, “some of the clearest evidence in the world for the existence of tribal warfare prior to Western contact” (66). For widespread, low-level violence, see George R. Milner, Eve Anderson, and Virginia G. Smith, “Warfare in Late Prehistoric West-Central Illinois,” American Antiquity, LVI (1991), 581–603, esp. 589 (“outbreaks”), 594 (“Internecine conflict”).

I Make Him My Dog / My Slave

1077

Because these traumatized bones lay beside a diverse collection of trade goods, it is also clear that the region’s endemic violence coexisted with peaceful exchange relationships that facilitated broader alliances, trade, and cultural mixing. Violence and slavery, in other words, did not define the region’s political economy. The peoples of the Pays d’en Haut responded to the region’s ecological and climatic diversity with a creative range of subsistence practices, spanning the spectrum from intense reliance on hunting and fishing in the north to a much greater dependence on agriculture in the south. Allies traded grain for meat and leather to diversify their diets and to demonstrate friendship. Tools and decorative items often drew upon external resources, with goods like obsidian, turquoise, pipestone, porcupine quills, shell beads, and animal hides circulating broadly within the region and linking it to areas far beyond. Given the historical association between trade and kin-based alliances, this commerce represents a counternarrative to violence and slavery that simultaneously highlights the restraints on warfare and the imperatives of military alliance that drove slave raids and the captive trade.9 Although we know very few particulars, in the fifteenth and sixteenth centuries the search for security drew many peoples together into several fluid but relatively stable alliances. Broadly defined by linguistic affinity, five core ethnic clusters coalesced in the Pays d’en Haut: Miaimi-Illinois, Fox-Sauk-Kickapoo, Sioux, Cree-Monsoni-Assiniboine, and Ojibwa-Ottawa-Potawatomi. Linked by geographic proximity, intermarriage, language, and culture more than by political unity, these broad groupings remained more or less stable through the seventeenth century, negotiating familial, commercial, and diplomatic alliances with neighboring villages. These coalitions have been mistaken for “tribes” or “confederacies” when nothing so politically structured held them together. Instead, they were collections of closely allied villages speaking dialects of a shared language, intermarrying, and cooperating in trade and diplomacy.10 9  As Milner, Anderson, and Smith argue: “It is not surprising that evidence for both cooperative and antagonistic intergroup relationships occurs in the same archaeological context because these contradictory forms of behavior coexist among the inhabitants of hostile social environments…. The establishment and maintenance of contacts among members of different communities presumably were motivated in part by a desire to stabilize some aspects of a volatile social setting, thereby diminishing the hazards of everyday life.” “Warfare in Late Prehistoric West-Central Illinois,”American Antiquity, LVI (1991), 592. See also Helen Hornbeck Tanner et al., eds., Atlas of Great Lakes Indian History (Norman, Okla., 1987), 18–23; Gilles Havard, Empire et métissages: Indiens et Français dans le Pays d’en Haut, 1660–1715 (Paris, 2003), 124–126. 10  Neal Salisbury, “The Indians’ Old World: Native Americans and the Coming of Europeans,” William and Mary Quarterly, 3d Ser., LIII (1996), 435–458. For Illinois social organization,

1078

Rushforth

The Indians of the Pays d’en Haut have often been depicted as “refugees” whose societies were “shattered” and scattered by a combination of warfare and disease in the mid-seventeenth century, particularly the aggressive Iroquois expansion of the 1640s and 1650s. Yet, aside from the Hurons, a tiny population of true refugees who fled to the heart of the Pays d’en Haut after the Iroquois routed their villages in the late-1640s, all of the region’s other peoples maintained ethnic and regional divisions broadly similar to those of the early seventeenth century. The Hurons’ dislocation was more of a geographic than a cultural reorientation, as their villages had a long history of trade and cultural interaction with the Anishinaabes and other Algonquian-speakers around Lake Huron. Although villages did move to ensure proximity to resources and allies—as they had always done—Indians’ sense of territoriality was rooted much more deeply than European colonizers first understood (or were willing to acknowledge). Far from “irrelevant,” as one scholar described their territorial attachments in the late-seventeenth-century Pays d’en Haut, Natives forcefully explained the boundaries of their territory to French observers and fought to maintain control of hunting, fishing, and agricultural sites against their neighbors’ pretensions. As Jonathan Carver observed long after Iroquois violence had passed: The most uncultivated among them are well acquainted with the rights of their community to the domains they possess, and oppose with vigour every encroachment on them. Notwithstanding it is generally supposed that from their territories being so extensive, the boundaries of them cannot be ascertained, yet I am well assured that the limits of each nation in the interior parts are laid down in their rude plans with great precision. Rather than scattering in defeat, these peoples responded to Iroquois and other warfare—as well as to new trade opportunities—by moving villages to more secure locations nearby or by forming new defensive alliances. In this atmosphere, conflict over new territorial claims and over access to limited resources produced a new wave of warfare and slave raiding.11 see Eric Hinderaker, Elusive Empires: Constructing Colonialism in the Ohio Valley, 1673–1800 (Cambridge, 1997), 9–13; Charles Callender, “Illinois,” in Handbook, XV, Northeast, 673–680; Callender, Social Organization of the Central Algonkian Indians (Milwaukee, Wis., 1962). 11  Jonathan Carver, Three Years Travels through the Interior Parts of North-America … (Philadelphia, 1789), 213–214. For Natives unmoored by Iroquois violence, see Richard White, The Middle Ground: Indians, Empires, and Republics in the Great Lakes Region, 1650– 1815 (Cambridge, 1991), 1–49, esp. 1 (“refugees,” “shattered”), 17 (“irrelevant”). For earlier

I Make Him My Dog / My Slave

1079

When they arrived in the Pays d’en Haut, French observers themselves misread Indians’ annual subsistence cycles as evidence that they had been unmoored by Iroquois warfare. Routine seasonal movement left agricultural settlements all but empty several months a year as residents left summer towns for smaller winter villages that might house only a handful of extended family groups. Generally more difficult to see in the historical and archaeological records, winter villages were essential to Native territoriality, despite being dismissed by the French as mere campgrounds. The presence of multiple ethnic and linguistic groups at strategic trading centers like Green Bay, Michilimackinac, and Sault Sainte Marie also led French writers to overestimate ethnic fluidity in the Pays d’en Haut. Similar to other multiethnic trading centers around the continent—like The Dalles on the Columbia River or the Mandan villages on the Plains—these sites succeeded as trade centers precisely because they drew from a wide array of places and peoples in the region. Stumbling upon these villages and carrying an inflated sense of Iroquois power, French authors interpreted Indians’ long-standing cross-cultural relations as Iroquois-induced chaos, even when the intermarriage and village sharing followed established linguistic and geographic patterns. What is more, the French exploited and exaggerated anti-Iroquois sentiments to bolster their own alliance-building efforts. To be sure, western Indians feared the Iroquois, and these fears were only heightened by the arrival of Huron refugees fleeing Iroquois raids. But, far from fundamentally reconfiguring their identities and political economies in response to the Iroquois threat, the peoples of the Pays d’en Haut drew on the strength of long-standing alliances and trade partnerships to protect them from this distant but formidable enemy. Slavery played expressions of the “refugee” interpretation, see Louise Phelps Kellogg, The French Régime in Wisconsin and the Northwest (Madison, Wis., 1925), 103; George T. Hunt, The Wars of the Iroquois: A Study in Intertribal Relations (Madison, Wis., 1940), 101; R. David Edmunds, The Potawatomis: Keepers of the Fire (Norman, Okla., 1978), 5. For late-seventeenth-century Native territoriality in the Pays d’en Haut, see Havard, Empire et métissages, 113–203; Heidi Rosemary Bohaker, “Nindoodemag: The Significance of Algonquian Kinship Networks in the Eastern Great Lakes Region, 1600–1701,” WMQ, 3d Ser., LXIII (2006), 23–52; Bohaker, “Nindoodemag: Anishinaabe Identities in the Eastern Great Lakes Region, 1600 to 1900” (Ph.D. diss., University of Toronto, 2006), esp. 12–16; William James Newbigging, “The History of the French-Ottawa Alliance, 1613–1763” (Ph.D. diss., University of Toronto, 1995), esp. 26–82; Andrew K. Sturtevant, “‘Inseparable Companions’ and Irreconcilable Enemies: The Huron-Wendats and Odawas of French Detroit,” Ethnohistory (forthcoming). For a similar discussion of the Southwest, see Juliana Barr, “Geographies of Power: Mapping Indian Borders in the ‘Borderlands’ of the Early Southwest,” WMQ, 3d Ser., LXVIII (2011), 5–46.

1080

Rushforth

a central role in this strategy, weakening outsiders and strengthening relationships with allied neighbors.12 From the confluence of the Missouri and Mississippi Rivers through the Illinois River valley to the southern shores of Lake Michigan, Illinois and Miami villages occupied the most fertile and temperate territory in the Pays d’en Haut. Speaking mutually intelligible dialects of Miami-Illinois (Algonquian), these peoples lived in relatively large agricultural settlements that also controlled key trade routes linking the Great Lakes with the Mississippi and the eastern Plains. People in Miami and Illinois towns traded with and married one another but had no formal political ties beyond the extended kinship cemented by marriage. Generally friendly with the Anishinaabes to the northeast and the Missouris to the west, they had longstanding quarrels with the Foxes and their allies as well as the Sioux to the northwest. Although they would encounter the Iroquois on several occasions as military enemies, Illinois peoples did not flee in terror, moving their villages westward in response to Iroquois warfare. On the contrary, modern archaeological evidence suggests that, if anything, Illinois settlements moved farther up the Illinois River valley, edging closer to the Iroquois in the mid-seventeenth century during the peak of the Iroquois wars in the eastern Great Lakes. Given the growing presence of European goods in the remains of these villages, archaeologists speculate that the Illinois might have repositioned themselves to gain access to European commodities. Whatever the reason, there is no evidence of a shattered political economy or of a widespread westward flight from the Iroquois.13

12  Duane Esarey, “Seasonal Occupation Patterns in Illinois History: A Case Study in the Lower Illinois River Valley,” Illinois Archaeology, IX (1997), 164–219; Bohaker, “Ninddodemag: Anishinaabe Identities,” 12–16. For an excellent discussion of Native seasonal movement in the Pays d’en Haut, see Michael J. Witgen, “An Infinity of Nations: How Indians, Empires, and Western Migration Shaped National Identity in North America” (Ph.D. diss., University of Washington, 2004), esp. 23–34, although Witgen draws different conclusions than I do here about the implications of ethnic fluidity. 13  Tanner et al., eds., Atlas of Great Lakes Indian History, map 6; J. Joseph Bauxar, “History of the Illinois Area,” in Handbook, XV, Northeast, 594–601; Thomas Emerson and James A. Brown, “The Late Prehistory and Protohistory of Illinois,” in John A. Walthall and Thomas E. Emerson, eds., Calumet and Fleur-de-Lys: Archaeology of Indian and French Contact in the Midcontinent (Washington, D.C., 1992), 77–128. For Illinois movement in the mid-seventeenth century, see Robert Mazrim and Duane Esarey, “Rethinking the Dawn of History: The Schedule, Signature, Agency of European Goods in Protohistoric Illinois,” Midcontinental Journal of Archaeology, XXXII (2007), 145–200; Esarey, “Seasonal Occupation Patterns,” Illinois Archaeology, IX (1997), 164–219.

I Make Him My Dog / My Slave

1081

Villages of the Foxes, Sauks, and Kickapoos, who spoke mutually intelligible Algonquian dialects, dotted the Wisconsin and Fox Rivers north of the Illinois country and west of Lake Michigan. Twentieth-century oral histories recall a period when these peoples lived much farther east, moving westward into the Great Lakes region before Europeans arrived. If these stories are read literally, placing them east of Lake Michigan, Fox westward movement occurred at least a century before the arrival of the French, as both the written and archaeological record places them in Wisconsin by the first half of the seventeenth century. Like the Illinois, these peoples practiced seasonal movement from agricultural to winter hunting villages throughout modern Wisconsin. Pressed between Sioux and Illinois enemies, the Foxes and their cousins compensated by gaining privileged access to Green Bay through an alliance with a collection of Potawatomi villages situated between the bay and the main body of Lake Michigan. Some evidence suggests that Sioux raids from the northwest had compressed the territory of the Fox-led alliance in the mid-seventeenth century, explaining lingering friction between the two groups at the time of French contact.14 The Sioux occupied a relatively large territory from the Minnesota River to the lake country in north-central Minnesota. Not yet the horse-mounted bison hunters of the nineteenth century, the eastern Sioux survived on a mixture of hunting, fishing, maple sugaring, and wild rice and oat gathering. They 14  Based on an overly literal reading of twentieth-century—origin stories, some scholars have argued the Fox-Sauk-Kickapoo grouping moved from modern Michigan to Wisconsin around the mid-seventeenth century, but it conflicts with other evidence showing a longer occupation of the Fox and Wisconsin River valleys. Oral history recalling life on the shores of great waters could also suggest a much shorter move, from the western shores of Lake Michigan to Lake Winnebago and the Fox and Wisconsin River valleys. No identifiable Fox sites have been found east of Lake Michigan, suggesting that the Iroquois wars had little to do with Fox village placement in Wisconsin. See Havard, Empire et métissages, 117; Tanner et al., eds., Atlas of Great Lakes Indian History, map 6; R. David Edmunds and Joseph L. Peyser, The Fox Wars: The Mesquakie Challenge to New France (Norman, Okla., 1993), 3–14; Charles Callender, “Fox,” in Handbook, XV, Northeast, 636–647, and “Sauk,” 648–655, and Callender, Richard K. Pope, and Susan M. Pope, “Kickapoo,” 656–667. Callender, “Fox,” 636, argues that the Foxes’ eastern placement relies on “vague traditions and early cartographic data,” which were notoriously unreliable. Edmunds and Peyser, Fox Wars, 9–10, esp. nn. 11–13, conclude that, at least for the Foxes, the evidence for seventeenth-century residence east of Lake Michigan is unpersuasive. Archaeologist Jeffrey A. Behm recently repeated the “refugee” interpretation of Fox village sites, dating Fox villages based as much on secondary historical accounts as on archaeological markers, in “The Meskwaki in Eastern Wisconsin: Ethnohistory and Archaeology,” Wisconsin Archaeologist, LXXXIX (2008), 7–85.

1082

Rushforth

cultivated great quantities of tobacco for their own use and for trade. Sioux settlements could be quite large at certain times of the year, generally organized as a cluster of smaller hamlets containing fifty or sixty extended family groups. In these more or less permanent residences the Sioux lived in framed bark lodges, but, because regular movement was essential to their subsistence strategy, they traveled with deerskin lodges for easier mobility. “This nation, which is very numerous,” wrote an early French trader, “is always wandering, living only by hunting.” Although he later qualified this claim by noting the presence of semipermanent villages where the Sioux would gather in spring and summer, he understandably compared the Sioux to more southerly peoples and found them less settled. Thriving on plentiful deer, elk, bison, fish, and wild grains, Sioux villages traded for corn with neighbors who occupied regions more suited to agriculture.15 Their Cree (Algonquian) neighbors, in alliance with the Assiniboines (Siouan), competed for control of the northern lakes and access to the northwestern shores of Lake Superior. Settled north and east of the Sioux, in modern northwestern Ontario, the Crees and Assiniboines lived well north of the best agricultural territories and thus relied exclusively on hunting, fishing, gathering, and trade for their subsistence. With access to excellent moose and beaver hunting grounds, however, they also played an important role in trade, linking Lake Superior to Hudson Bay, which sat within relatively easy reach to the northeast. Long-standing conflict plagued their southwestern borderlands, where Cree and Assiniboine territory abutted Sioux hunting lands and fisheries. At some point in the distant past this had, for the Assiniboines, been a fight among Siouan-speaking cousins.16 South of the Crees, arcing along the northern shores of Lake Superior and throughout much of modern northern Michigan, a wide array of Anishinaabe peoples lived in semipermanent villages and controlled key portage and trade centers. Speaking closely related dialects of Anishinaabemowin (Algonquian), the Ojibwa and Ottawa communities in this area lived on the Great Lakes’ 15  André Pénicaut, Relation of M. Penicaut, ed. Edward D. Neill, trans. A. J. Hill (Minneapolis, Minn., n.d.), 8 (“always wandering”); DeMallie, “The Sioux at the Time of European Contact,” in Kan and Strong, eds., New Perspectives, 239–260; DeMallie, “Sioux until 1850,” in Handbook, XIII, Plains, part 2, 718–760; Anderson, Kinsmen of Another Kind, 1–13; Douglas A. Birk and Elden Johnson, “The Mdewakanton Dakota and Initial French Contact,” in Walthall and Emerson, eds., Calumet and Fleur-de-Lys, 203–240. 16  Tanner et al., eds., Atlas of Great Lakes Indian History, map 6; David Meyer, “Timedepth of the Western Woods Cree Occupation of Northern Ontario, Manitoba, and Saskatchewan,” in Papers of the Eighteenth Algonquian Conference (1987), 187–200.

I Make Him My Dog / My Slave

1083

abundant fish, game, wild rice, and maple sugar, cultivating small fields of corn and legumes only rarely because of a short northern growing season. The Ottawa stronghold at Bawating (modern Sault Sainte Marie) provided access to Lakes Huron, Michigan, and Superior and thus had long been a site of trade involving most of the region’s peoples, but it remained under Ottawa control both before and after midcentury warfare quickened the usual pace of village movement. Like the Illinois, the Ottawas were, not a single, politically-unified tribe, but rather a collection of closely allied villages with a shared history and clan structure that bound them together over generations. Even in times of disruption and dislocation, these bonds of kin and clan allowed the Anishinaabes to maintain both a persistent sense of community and a remarkable degree of cultural integrity.17 Overlaying the cultural and linguistic particularities that separated Ottawas from Sioux, Illinois from Crees, Miamis from Ojibwas, a regional diplomatic culture provided mechanisms for cooperative trade and methods of controlled warfare with enslavement at its center. Reciprocal diplomacy was balanced by reciprocal violence, and over time the form of that violence developed into a widely shared set of practices that one seventeenth-century Jesuit described as the “general custom of the country.” According to Antoine Laumet de Lamothe Cadillac, all the region’s Indians conducted warfare according to “the same maxim, the same practice, the same mode, and the same manner of fighting.” Far from the irrational bloodlust that Europeans would see in these practices, they were, in fact, designed to discourage warfare because war councils knew that they risked their own people’s enslavement each time they authorized a raid. Slavery in the Pays d’en Haut was thus indigenous not only because it was practiced by native peoples but because the peoples that practiced it were indigenous to the slave system: they resided in the region where reciprocal slave raiding, trading, and negotiating took place. Living within reach of potential slaves and slavers placed powerful limits on whatever social impulses would argue for violence. Indigenous slavery was thus driven, not by a high demand for slaves, but by the political and cultural imperatives of enslavement.18 17  Tanner et al., eds., Atlas of Great Lakes Indian History, map 6; Johanna E. Feest and Christian F. Feest, “Ottawa,” in Handbook, XV, Northeast, 772–786, and E. S. Rogers, “Southeastern Ojibwa,” 760–771; Bohaker, “Nindoodemag: The Significance of Algonquian Kinship Networks,” WMQ, 3d Ser., LXIII (2006), 23–52; Newbigging, “History of the FrenchOttawa Alliance,” 25–41. 18  “La meme maxime, meme pratique, meme mode, et meme façon de combattre.” Antoine Laumet de Lamothe Cadillac, “Relation du Sieur de la Motte Cadillac,” MS, ca. 1693–1697, Newberry Library, Ayer MS 130; Jesuit Relations, LVIII, 67.

1084

Rushforth

This regional war culture was most fully articulated in the rituals of the calumet, which balanced the competing impulses of war and peace through slavery. “There is nothing among them either more mysterious or more revered,” wrote Jacques Marquette and Louis Joliet in 1681. “One does not give as much respect to the scepters of kings as they give to it; it seems to be the God of peace and war, the arbiter of life and death.” Widely recognized as the central diplomatic ritual performed in the seventeenth-century Pays d’en Haut, the calumet was believed to originate among the Pawnees in the Missouri River valley, having been “communicated from village to village as far as the Outaoüas [Ottawas]” by the time the French arrived. Practiced by all the peoples of the region, the ceremonies associated with the calumet facilitated trade and diplomacy among friends by expressing a commitment to violence against common enemies. At every stage of what often became a multiday ceremony, the calumet drew upon enslavement as the earthly expression of otherworldly powers harnessed to protect and unite the ceremony’s hosts and their guests.19 “This calumet,” explained Louis Hennepin, “is a type of large smoking pipe, with a head of beautiful, well-polished, red stone, with a shaft two-and-a-half feet long made from a very strong reed, decorated with feathers of all types and colors, mingled and arranged with great care, with several locks of women’s hair arranged in different ways with two wings.” The pipe embodied the mystical and practical impulses that converged to produce the ritual’s power. Its otherworldly symbolism was difficult for Europeans to read: according to one late-seventeenth-century writer, it was “the most mysterious Thing in the World.” However limited their understanding, all French observers agreed that Indians believed it carried sacred powers. The Jesuit Pierre Le Sueur, who lived among the Sioux at the end of the seventeenth century, judged the calumet “a true religious cult,” and trader and colonial official Claude-Charles de La Potherie wrote that the calumet was held “very sacred among these peoples.”20 19  “Il n’est rien parmy eux ni de plus mysterieux ni de plus recommendable, on ne rend pas tant d’honneur aux sceptres des Rois qu’ils luy en rendent, il semble estre le Dieu de la paix et de la guerre, l’arbitre de la vie et de la mort.” Jacques Marquette and Louis Joliet, Voyage et découverte de quelques pays et nations de l’Amérique Septentrionale (Paris, 1681), 23; Nicolas Perrot, Memoir on the Manners, Customs, and Religion of the Savages of North America, in Emma Helen Blair, ed. and trans., The Indian Tribes of the Upper Mississippi Valley and Region of the Great Lakes (Cleveland, 1911–1912), I, 186 (“communicated”). 20  “Ce Calumet est une espece de grande pipe à fumer, dont la teste est d’une belle pierre rouge bien polie, dont le tuyau long de deux pieds et demy, est une Canne assé forte, ornée de plumes de toute sorte de couleurs, meslées et ranges fort proprement, avec plusieurs nattes de cheveux de femmes, lassées de diverses manieres avec deux aisles”: Hennepin,

I Make Him My Dog / My Slave

1085

figure 4 Man Holding a Calumet. Louis Hennepin, Nouvelle découverte d’un très grande pays situé dans l’Amériques. Utrecht, 1697. Courtesy Special Collections Research Center, Swem Library, College of William and Mary.

Miami-Illinois—speakers called the calumet “apᴕagana,” and its shaft “apᴕacanti,” referring to the feathers with which the shaft was wrapped, drawing manit, or spiritual power, from these beings that lived between heaven and earth and therefore bridged the two worlds. The smoke, too, linked the sky world with land, drawing upon the power of the sun, which had grown the tobacco, by ingesting and then offering back the smoke from the plant. Indians sometimes made calumets in pairs, one painted green and the other blue to represent the earth and sky, emblematic of the calumet’s power to bring otherworldly power to bear on worldly matters. Alliances confirmed in a calumet Description de la Louisiane, 80; Hennepin, New Discovery, I, 74 (“most mysterious”). For Le Sueur, see William A. Turnbaugh, “Calumet Ceremonialism as a Nativistic Response,” American Antiquity, XLIV (1979), 688. “C’est un lien si Sacré parmi ces peoples”: [ClaudeCharles] Bacqueville de La Potherie, Histoire de l’Amérique Septentrionale (Paris, 1722), II, 15 [13] (“very sacred”).

1086

Rushforth

ceremony thus represented far more than practical political agreements. They were sacred bonds, and those who violated them risked disaster. “They are persuaded,” according to Hennepin, “that great misfortune would befall them were they to violate the faith of the calumet.” La Potherie went even further, claiming that they believed that “whoever violates a Calumet shall perish, and incur the indignation of the Gods who have allowed the power of the Sun to give light to the earth.”21 If the mystical power of the calumet came from another world, it derived much of its social authority from its physical creation. Because producing the pipe and the ceremonies that accompanied it demanded complex labor and coordination within the hosting village, calumet ceremonialism ensured broad participation—of both men and women—in diplomacy and broad support for the agreements reached. As each calumet took great effort to craft, only a limited number of calumets circulated. From killing ducks, birds of prey, porcupines, and sometimes rabbits to cutting cane reeds, quarrying, polishing, and boring out the pipe bowl without the advantage of metal tools, calumet production was a serious effort that drew upon—and drew in—all segments of society. Female artists decorated each calumet with distinctive marks to allow others to know its origins. According to one early account, “Every Nation adorns the Calumet as they think fit according to their own Genius and the Birds they have in their Country.” Pierre-François-Xavier de Charlevoix agreed: “They say that according to the Manner in which the Feathers are disposed, they immediately know what Nation it is that presents it.” Jacques Marquette experienced this aspect of calumet ceremonialism when he approached a Michigamea (Illinois) village downriver from the Peoria village that had given him a calumet. Sighting the French from a distance, the Michigameas readied for attack despite Marquette’s theatrical presentation of a calumet held above his head. Only when Marquette drew close enough for the Michigameas to recognize the pipe as a product of their Peoria allies did they stand down and 21  Jacques Gravier and Jacques Largillier, “Dictionnaire illinois-français,” MS, ca. 1690s, Watkinson Library Special Collections, Trinity College, Hartford, Conn., 46 (“apᴕagana,” “apᴕacanti”); Perrot, Memoir, in Blair, ed. and trans., Indian Tribes, I, 186. “Tous ces Barbares sont généralement qu’il leur arriveroit de grands malheurs, s’ils avoient violé la foy du Calumet”: Hennepin, Nouvelle découverte, 151. “Quicon’que viole un Calumet doit perir, et il s’attire en même tems l’indignation des Dieux qui ont laissé le pouvoir au Soleil d’éclairer la terre, et ne peuvent soufrir qu’un perfide fasse rien de contraire au Calumet qui est le gage de la Paix”: Bacqueville de La Potherie, Histoire de l’Amêrique Septentrionale, II, 14–15. For paired calumets, see Robert L. Hall, An Archaeology of the Soul: North American Indian Belief and Ritual (Urbana, Ill., 1997), 50–52.

I Make Him My Dog / My Slave

1087

welcome Marquette into their village. The labor of the Peorias who had produced Marquette’s pipe fixed their commitment to its recipient and provided a visual endorsement of the foreigner to others within their alliance.22 The calumet was also embedded in a ceremonial complex that transcended the pipe’s physical creation. Generally understood as an instrument of peace, the calumet’s power was expressed through the violent imagery of enslavement. The ceremony occurred in four stages, sometimes lasting as long as two or three days. In the first stage, the chief offering the calumet gathered his people to welcome the guest to whom the pipe was being offered, and some of them formed either a circle or square within which the ceremonies would take place. The chief held the calumet to the sun, took several deep draughts, and blew the smoke toward the sky. The guests were then invited to smoke the pipe as a sign of their unity and shared power. In the third—and by far the longest—stage of the calumet ceremony, individual warriors entered the circle or square and began to “dance the calumet.” After calling on his guardian spirits, the performer mimed the act of killing and capturing his enemies, using the pipe shaft as a simulated weapon against other dancers. This was done so gracefully, according to Jacques Marquette, that it “might pass for a very fine opening of a Ballet in France.”23 Each warrior then retold, one by one, the story of every slave he had taken since becoming a warrior. “The third Scene,” Marquette explained, “consists of a lofty Discourse, delivered by him who holds the Calumet … he recounts the battles at which he has been present, the victories that he has won, and the names of the Nations, the places, and the Captives whom he has made.” The message was clear: his manitous, whose feathers adorned the pipe, gave him power over his enemies and would continue to do so. Accept our friendship, he implied, or suffer the fate of those whose stories I have told. He then offered slaves as a gift to ratify the alliance, bodies of evidence that his people would be a powerful ally or, if necessary, a dreaded enemy.24

22  Hennepin, New Discovery, I, 74 (“every nation adorns”); “Chaque Nation l’embellissant selon son usage … particulier”: Hennepin, Nouvelle découverte, 150; and [Pierre FrançoisXavier de] Charlevoix, Letters to the Dutchess of Lesdiguieres; Giving an Account of a Voyage to Canada, and Travels through That Vast Country, and Louisiana, to the Gulf of Mexico (London, 1763), 134 (“immediately know”). For the Michigameas, see Hall, Archaeology of the Soul, 7. 23   Jesuit Relations, LIX, 135–137. 24   Jesuit Relations, LIX, 137.

1088

Rushforth

Produced by the hosting village, bearing symbols of otherworldly power, and accompanied by slaves and other gifts, the calumet generated its symbolic meaning face-to-face, limited to the relatively small region where it was recognized and practiced. The farther one moved from the village where the pipe was created, the less likely one would be to find others who would recognize its marks and therefore be able to know the lines of alliance it represented. Louis-Armand de Lom d’Arce, baron de Lahontan, experienced these regional limits firsthand. Having been given a calumet as a token of friendship, probably by an Illinois village, he headed up the Missouri River. Upon reaching the village of a people he identified as “Gnacsitares,” he produced the pipe, hoping it would signify his peaceful intentions and demonstrate his desire for alliance. His hosts did not react as he had anticipated. “Here ended the Credit and Authority of the Calumet of Peace,” he wrote, “for the Gnacsitares are not acquainted with that Symbol of Concord.” Underscoring the regional limits of this war culture, Lahontan’s experience provides a valuable reminder that, in the seventeenth century, calumet ceremonialism was practiced only in the western Great Lakes and Mississippi Valley.25 Simultaneously symbolizing war and peace, performing violence to cement friendship, the calumet was an apt shorthand for the larger forces that Great Lakes war culture worked to balance. Those who quarried its stone, decorated its shaft, or provided and prepared food for the accompanying feasts joined in both the collective expression of friendship and the collective memory of war. These significant investments in time and resources gave everyone a stake in the maintenance of the peace that the ceremonies established or in the support of the warfare they celebrated. As the narrative center of the calumet dance and the punctuation at its conclusion, slaves stood at the heart of both Natives’ regional war culture and the diplomacy that kept warfare from overtaking the region.

25  [Louis-Armand de Lom d’Arce], Baron de Lahontan, New Voyages to North-America (London, 1703), I, 121 (“Gnacsitares”). For a new interpretation of Lahontan’s journey up the Missouri, see Peter Wood, “Lahontan’s Letter XVI: Frenchmen on the Missouri River in 1688,” paper, Annual Conference of the Omohundro Institute of Early American History and Culture, Quebec, 2006. For the geographic limits of calumet ceremonialism, see Ian W. Brown, “The Calumet Ceremony in the Southeast and Its Archaeological Manifestations,” American Antiquity, LIV (1989), 311–331; Louis André, “Preceptes, phrases, et mots de la langue algonquine outaouaise pour un missionnaire nouveau,” ca. 1688, MS, John Wesley Powell Library of Anthropology, Smithsonian Institution, Washington, D.C. (the original is located in the Archives des Jésuites au Canada, Montreal).

I Make Him My Dog / My Slave



1089

Nit’aouakara—I Make Him My Dog / My Slave This reception is very cruel; some tear out the prisoners’ nails, others cut off their fingers or ears; still others load them with blows from clubs. Sébastien Rale, Jesuit, 1723

Indians of the Pays d’en Haut expressed their relationship to slaves through metaphors of domestication and mastery, comparing captives to dogs and other domesticated animals. More than a simple insult, the metaphorical domestication of enemy captives represented an elaborate cultural idiom that shaped the practice and defined the meaning of indigenous slavery. To shame and intimidate their enemies, Algonquians and Siouans treated their prisoners with great disrespect through symbolic acts of humiliation designed to strip them of their former identities and incorporate them as subordinate domestics. Beginning with demeaning abuse on the journey home, continuing through acts of torture as captives were received into the village, and culminating in ceremonial killing or forced incorporation, Indians designed their rituals of enslavement to demonstrate their mastery over weaker enemies and to secure the allegiance and passivity of those they would keep alive as slaves. Anishinaabe-speakers called their slaves awakaan, which meant “captive,” “dog,” or “animals kept as pets.” The earliest French lexicon of central Algonquian languages, recorded between 1672 and 1674 by Jesuit Father Louis Nicolas, included aouakan, meaning “slave or prisoner of war,” as one of eight essential nouns for missionaries to know to effectively teach western Indians. After living among the Anishinaabes of Michilimackinac and Sault Sainte Marie, Lahontan composed his own dictionary of essential Anishinaabe terms, listing Ouackan for “slave.”26 26  John D. Nichols and Earl Nyholm, A Concise Dictionary of Minnesota Ojibwe (Minneapolis, Minn., 1995), 14; Frederic Baraga, A Dictionary of the Otchipwe Language, Explained in English (Cincinnati, Ohio, 1853), 49–50, 453; C. Douglas Ellis, Âtalôhkâna nêsta tipâcimôwina: Cree Legends and Narratives from the West Coast of James Bay (Winnipeg, 1995), 55, 85, 159, 449. Although there are many variant spellings of awakaan, I use Nichols and Nyholm’s version as the most recent standardization of the orthography. For Nicolas, see Diane Daviault, ed., L’algonquin au XVIIe siécle: une édition critique, analysée et commentée de la grammaire algonquine du Père Louis Nicolas (Sainte-Foy, Que., 1994), 5, 34, 106–107; [Louis-Armand de Lom d’Arce], Baron de Lahontan, Voyages du Bon de Lahontan dans l’Amérique Septentrionale (Amsterdam, 1705), II, 321. Jonathan Carver echoed Lahontan’s spelling in his own word list more than sixty years later: “Esclave. Ouackan.” See Carver, Voyage dans les parties intérieures de l’Amérique Septentrionale pendant les années 1766, 1767 et 1768 (Yverdon, Switzerland, 1784), 312.

1090

Rushforth

The most-advanced French linguist of Anishinaabemowin in the seventeenth century was Louis André, a Jesuit trained in Latin linguistics and later a linguistics professor at Quebec’s Jesuit college. Living in several Native villages and working closely with Ottawa informants, he conducted a fourteenyear study of their language and produced an eight-hundred-page manuscript dictionary and phrase book designed to teach other missionaries the language. André recorded dozens of Anishinaabe terms and phrases relating to slavery, most of which expressed the metaphor of slaves as domestic animals. André wrote that the verb to enslave (nit’aouakara), for example, literally meant to make someone a dog. Often rendered in the first person possessive, it described enslavement as an act of animal domestication: to say “I make him my slave” was to say “I make him my dog.” One of the most intriguing variations of this verb was translated by another Jesuit among the Ottawas as “I make him my plaything, my slave.” Rendered in the diminutive form, it could be translated more literally as “I make him my little dog / puppy.”27 Miami-Illinois—speakers also equated slaves with dogs, their most common domesticated animal. One Miami-Illinois word for slave was aremᴕa, meaning, in the words of the Jesuit Jacques Gravier, “dog, all domestic animals, and, as a term of contempt, slave.” Gravier defined a female equivalent of this word, acᴕessemᴕa, as “heifer, bitch, female animal.” If dogs or slaves ran free, Miami-Illinois—speakers would say achaᴕe nitaïa: “my dog, animal, slave runs

 Linguists tend to distinguish between words’ historically recorded forms and their standardized spellings by placing those words quoted from historical sources in quotation marks and placing standardized non-English words in italics. I break from that convention here—by generally italicizing all Algonquian words—for two reasons. First, because the Miami-Illinois language lost most of its native speakers quite early, it has relatively few standardized words, and none of the Algonquian languages has standardized forms for much of the seventeenth-century vocabularies I discuss here. Second, the large number of Algonquian words in the text would make the use of quotation marks cumbersome, interfering with the narrative rather than clarifying its meaning. Algonquian words quoted from historical sources are spelled as they were recorded in the original manuscript or printed text. 27  “Je le fais esclave”: André, “Preceptes, phrases, et mots de la langue algonquine outaouaise,” s.v. “esclave.” “J’en fais mon joüet, mon esclave”: Pierre Du Jaunay, “Dictionarium gallico-outaouakum,” MS, 1748, copy in Smithsonian Institution Anthropology Library, s.v. “esclave.” Du Jaunay also confirms André’s translation of nit’aouakara: “nit’aᴕakan. mon [esclave] nit’aᴕakara. Je le fais esclave.”

I Make Him My Dog / My Slave

1091

around.” The word nitaïa—literally “my possession”—had one meaning that Gravier translated as “my domestic animal, my dog, my cat, also my slave.”28 The act of domestication—turning enemies into dogs—began even before the warriors left their village for the raid. Raids originated in communal, and quite often contested, discussions of issues ranging from the need for retaliation to preemptive raids designed to weaken a threatening enemy. Among the Illinois, elder male relatives of anyone killed by an outsider could call a council and demand revenge. “If my strength and my courage equalled yours, I believe that I would go to avenge a relative as brave and as good as he was,” one Illinois elder said in a late-seventeenth-century war council. “But being as feeble as I am, I cannot do better than address myself to you,” the young warriors. He persuaded them to fight by appealing to their sense of collective revenge and individual masculine honor.29 Louis Hennepin witnessed similar negotiations among the eastern Sioux, seeing for himself the beginning of a process that had led to his own enslavement. The warrior or family initiating the raid sent invitations around the village, and sometimes to neighboring villages, to join in a war feast. Accepting this invitation meant accepting the call to war. Those who wished to join the war party gathered at the home of the one who invited them, singing their warrior songs as they arrived. In these songs, Sioux men recounted their deeds of bravery and recalled the captives they had taken, vowing similar success on this raid: “I am going to War, I will revenge the Death of such a Kinsman, I will slay, I will burn, I will bring away Slaves, I will eat Men.” A feast and dance followed, called by the Sioux šunkahlowanpi—literally “ceremonial song of the

28  “Chien, tte beste domestique et par mepris esclave”; “genisse, chienne, femelle de beste”: Gravier and Largillier, “Dictionnaire illinois-français,” 10. Anishinaabemowin also occasionally employed the term “alemwa” for dog, suggesting a proto-Algonquian origin. See, for example, André, “Preceptes, phrases, et mots de la langue algonquine outaouaise,” s.v. “chien.” For “Achaᴕe nitaïa,” see Gravier and Largillier, “Dictionnaire illinois-français,” 1. Carl Masthay, Kaskaskia Illinois-to-French Dictionary (Saint Louis, Mo., 2002), 47, translates the phrase as “my dog, animal, servant runs here and there,” inexplicably substituting the English “servant” for the original French “esclave,” or “slave.” For nitaïa, see Gravier and Largillier, “Dictionnaire illinois-français,” 14. 29  Pierre Deliette, “Memoir of De Gannes [Deliette] concerning the Illinois Country,” in Theodore Calvin Pease and Raymond C. Werner, ed. and trans., The French Foundations, 1680–1693, Collections of the Illinois State Historical Library, XXIII (Springfield, Ill., 1934), 377 (hereafter cited as Deliette, “Memoir”).

1092

Rushforth

dog”—described by later Sioux informants as “a parade with singing made by those who are on the point of going to war.”30 Eating men and taking slaves expressed the central theme of the šunkahlowanpi and other war feasts in the Pays d’en Haut. As Pierre-FrançoisXavier de Charlevoix explained, warriors “say also in direct Words, that they are going to eat a Nation; to signify, that they will make a cruel War against it; and it seldom happens otherwise.” The metaphorical equation of eating men and taking slaves found its physical expression in the ritual consumption of dog meat, which was the centerpiece of the warriors’ feast that preceded the raid. “This feast is one of dog’s flesh,” explained Nicolas Perrot of the rite among the Anishinaabes, “which [among them] is ranked as the principal and most esteemed of all viands…. Feasts of this sort are usually made only on the occasion of a war, or of other enterprises in which they engage when on expeditions against their enemies.” The practice was so entrenched in the Pays d’en Haut that one colonial official concluded, “The feast of dogs is the true war feast among all the savages.” Dog feasts sometimes continued into the journey toward an enemy’s territory, and evidence suggests that war parties sometimes killed and ate enemy dogs as a sign of power over them.31 Illinois warriors followed a similar logic in the ceremonial songs, dances, and feasts preceding slave raids. “When they wish to go to war,” Pierre Deliette explained, a chief “offers them a feast, usually of dog.” Then, “when the women see that they are preparing for this dance, they lead away all their dogs to a distance, for any of them that they find they kill and feast on.” The ritual consumption of dogs takes on new meaning in light of Illinois and other Algonquian terminology for slaves as dogs. Because enemies, once conquered and domesticated, assumed the social position of dogs, they were considered awakaan, aremᴕa, nitaïa, just as the animals were. The departing warriors thus performed their intention to both destroy and consume their enemies—the dogs, the slaves—in the coming raid.32 30  Hennepin, New Discovery, II, 72; Eugene Buechel and Paul Manhart, Lakota Dictionary: Lakota-English / English-Lakota (Lincoln, Nebr., 2002), 191, 291. 31  Charlevoix, Letters to the Dutchess of Lesdiguieres, 131 (“eat a nation”); Perrot, Memoir, in Blair, ed. and trans., Indian Tribes, 53–54 (“most esteemed”); Antoine Denis Raudot, “Memoir concerning the Different Indian Nations of North America,” in W. Vernon Kinietz, ed., The Indians of the Western Great Lakes, 1615–1760 (Ann Arbor, Mich., 1940), 403 (“true war feast”); Henri Joutel, The La Salle Expedition to Texas: The Journal of Henri Joutel, 1684–1687, ed. William C. Foster, trans. Johanna S. Warren (Austin, Tex., 1998), 119. 32  Deliette, “Memoir,” 376–377, 387. This tradition is also summarized in John A. Walthall, F. Terry Norris, and Barbara D. Stafford, “Woman Chief’s Village: An Illini Winter Hunting Camp,” in Walthall and Emerson, eds., Calumet and Fleur de Lys, 149. The authors found

I Make Him My Dog / My Slave

1093

When captured, slave raiders could find their own logic turned disastrously against them. An Ottawa war chief named Sinagos, who had a reputation as a brutal slaver, conducted a raid in Sioux territory in the early 1670s, “putting the men to flight and carrying away the women and children whom they found there.” Those who escaped the raid gathered reinforcements who pursued and captured Sinagos and the surviving captives. Recognizing his prominence, the Sioux decided to make an example of him rather than kill or enslave him. “They made him go to a repast,” wrote Nicolas Perrot, “and, cutting pieces of flesh from his thighs and all other parts of his body, broiled these and gave them to him to eat—informing Sinagos that, as he had eaten so much human flesh and shown himself so greedy for it, he might now satiate himself upon it by eating his own.”33 Because Indians imagined enslavement as the violent consumption of flesh, they compared freeing captives to vomiting: a violent release of the flesh they had eaten. When a Fox war chief entertained a French delegation in the 1680s, he offered his French guest some venison. When the Frenchman refused on the grounds that he was unhappy with Fox slave raids against French-allied Indians, the Fox man called for four captives, whom he released to his French guest. “Here is how reasonable the Fox can be … he vomits up the meat that he had intended to eat … even as it is between his teeth he spits it out, he asks you to return it to where he captured it.” Jacques Gravier recorded an expression for releasing captives among the Illinois: “nisicarintamaᴕa acᴕiᴕssemahi. je done la vie a cinq prisoniers,” I give life to five prisoners. The Algonquian verb means “to vomit.”34 Carrying mental images drawn from this rich verbal and ceremonial milieu, warriors began their journey to enemy territory seeking slaves to domesticate. In the indigenous war culture of the Pays d’en Haut, taking captives took precedence over killing enemies and especially over territorial conquest, which was extremely rare. “When a Savage returns to his own country laden with many scalps, he is received with great honor,” wrote Sébastien Rale, “but he is at the height of his glory when he takes prisoners and brings them home alive.” the charred remains of dogs in ceremonial fire pits, providing additional evidence of this practice. 33  Perrot, Memoir, in Blair, ed. and trans., Indian Tribes, 189, 190. 34  “Le Chef prenant la parole dit, voici en quoi l’Outagamis peut être raisonnable … il vomit la viande qu’il a eû dessein de manger … et l’ayant entre ses dents il la crache, il te prie de la remettre où il l’a prise.” Bacqueville de La Potherie, Histoire de l’Amérique Septentrionale, II, 214; Gravier and Largillier, “Dictionnaire illinois-français,” 527 (“to vomit”).

1094

Rushforth

Such feats of bravery, Rale explained, allowed a warrior to be considered “truly a man.” “They are so eager for this glory that we see them undertake journeys of four hundred leagues through the midst of forests in order to capture a slave.”35 Once warriors carried captives a safe distance away from a raided village, they bound them tightly by the hands and neck with a halter. Keeping the captives’ legs free except when they slept, the captors “immediately tie their hands and compel them to run on before at full speed, fearing that they may be pursued … by the companions of those whom they are taking away.” Louis Hennepin explained the danger of slowing the captors down. “When they have taken a slave, they garrote him and make him run,” he wrote shortly after his own release from slavery among the Sioux. “If he cannot keep up they strike him on the head … and scalp him.” This was a second wave of sorting strong captives from those who might become an immediate liability or a long-term drain on local resources. The old and infirm were rarely taken from the village, and the necessity of running, bound, for long stretches with little food ensured another level of fitness for surviving slaves.36 When the retreating party stopped for the night, captors staked their prisoners to the ground. Pierre Boucher, who traded extensively with Great Lakes Algonquians and spoke several of their dialects, described the restraints in 1664: When [the Indians of New France] capture prisoners … they bind them by the arms and by the legs with cords; except when they are marching, they leave the legs free. In the evening, when they camp, they lay the prisoners with their backs against the ground, and they plant some small stakes in the earth next to the feet, the hands, the neck, and the head; then they bind the prisoner to these stakes so tightly that he cannot move, which is more painful than one can imagine.

35  Raudot, “Memoir,” in Kinietz, ed., Indians of the Western Great Lakes, 355–356. For the Rale quote, see Jesuit Relations, LXVII, 171–173. 36  “Quand ils ont pris un esclave, ils le garotttent [sic] et le font courir; s’il ne peut les suivre, ils luy donnent un coup de hache à la teste et le laissent après luy avoir enlevé la peruque ou cheveleure”: Hennepin, Déscription de la Louisiane, 63; Jesuit Relations, LXVI, 275 (“tie their hands”).

I Make Him My Dog / My Slave

1095

figure 5 Native Captives in the Mississippi Valley. Vincenzo Coronelli, Partie occidentale du Canada ou de la Nouvelle France. Paris, 1688. Courtesy Library and Archives of Canada, Ottawa.

Binding served the obvious practical purpose of preventing escape, but it also symbolized the victim’s powerlessness and inferiority. These were tamed creatures, captured in the wild and brought home to domesticate.37 In one illustration, drawn on a 1688 map drafted from several travel accounts and personal interviews, a captive is shown kneeling, tethered to a stake by the neck like a dog, being beaten from behind with a stick. Another captive is kneeling with her hands bound in front, attached to a tumpline in her captor’s hands. In the emerging sign language of the eastern Plains, which included Sioux, Missouri, and even some Illinois territory, the word “slave” was signed by wrapping the thumb and middle finger of one hand around the wrist of the other and pulling the arm away from the body, implying physical control of the body and domination of the will. Supporting the interpretation of captive restraints as leashes, the Anishinaabes described releasing captives as “cut[ting] the cords on the dogs.”38 37  [Pierre Boucher], Histoire veritable et naturelle des moeurs et productions du pays de la Nouvelle France, vulgairement dite le Canada (Paris, 1664), 123–124. For Boucher’s familiarity with Native cultures, see Raymond Douville, “Boucher, Pierre,” DCB, II. See also Jesuit Relations, XLVIII, 101–103. For the Illinois, see Jesuit Relations, LXVI, 273–275; Deliette, “Memoir,” 381. For the Sioux, Ojibwas, and Ottawas, see Hennepin, Description de la Louisiane, 64. 38  Vincenzo Coronelli, “Partie occidentale du Canada ou de la Nouvelle France” (Paris, 1688), NMC-6411, LAC; Claude Charles Le Roy, Bacqueville de La Potherie, History of the Savage Peoples Who Are Allies of France, in Blair, ed. and trans., Indian Tribes, II, 39–41 (“cut the cords”). For sign language, see Garrick Mallery, A Collection of Gesture-Signs and Signals of the North American Indians, with Some Comparisons (Washington, D.C., 1880), 205–206, 236.

1096

Rushforth

Led along by a leash, captives faced physical and verbal abuse during their long march to the captors’ village. An Illinois warrior might refer to a slave he had captured as ninessacanta, “my slave, the one whom I bring,” a phrase drawn from the root word “to beat, batter, bludgeon” and, occasionally, “to beat to death.” Louis Hennepin found his repeated beatings more terrifying than debilitating, enhancing his captors’ arbitrary power by making him fear constantly for his life. Indeed, the march northward to the Sioux villages was so disorienting for Hennepin that he lost all sense of place and distance, sketching on his return to Europe a map of the Upper Mississippi Valley that stretched it hundreds of miles north of its headwaters. Copied by several subsequent cartographers, Hennepin’s bewilderment registered in European cartography long after his release from captivity.39 As the returning warriors neared their village, the war chief signaled their arrival with a series of high-pitched cries, one for each captive in the party. In some accounts, individual warriors also cried out once for each of the captives they had taken. “As soon as he arrives, all the people of the village meet together, and range themselves on both sides of the way where the prisoners must pass,” wrote Sébastien Rale. “This reception is very cruel; some tear out the prisoners’ nails, others cut off their fingers or ears; still others load them with blows from clubs.” Among the most degrading of the gauntlet’s many torments was the participation of women and children, whose taunts fell with special poignancy on captured male warriors. Like the ceremonies that initiated the slave raid, the logic of subordination required that captives’ incorporation into village society be a public affair involving all segments of Native society.40 39  Gravier and Largillier, “Dictionnaire illinois-français,” 28, 340. The first map influenced by Hennepin’s information was a 1681 Paris map titled Carte de la Nouvelle France, which could have been drawn only from Hennepin because of its placement of the “Issati,” or Sioux villages. Derek Hayes, America Discovered: A Historical Atlas of North American Exploration (Vancouver, 2004), map 92. Hennepin’s own map, first published in his 1683 Déscription de la Louisiane, was reprinted several times into the early eighteenth century, including in his Nouvelle découverte (1697 and subsequent editions). 40   Jesuit Relations, LXVII, 173. Rale speculated that the Illinois adopted these cruelties only after their similar treatment as captives of the Iroquois: “It was the Iroquois who invented this frightful manner of death, and it is only by the law of retaliation that the Illinois, in their turn, treat these Iroquois prisoners with an equal cruelty.” See Jesuit Relations, LXVII, 173–175. This statement should be assessed cautiously, however, as the French frequently minimized the violence of their allies and exaggerated that of the Iroquois. See, for example, a report from 1660 that describes French-allied Indians’ tearing out fingernails, cutting off fingers, and burning hands and feet at Quebec, dismissed by another Jesuit as “merely the game and diversion of children” ( Jesuit Relations, XLVI, 85–101, esp. 93).

I Make Him My Dog / My Slave

1097

Those disfigured by the gauntlet bore permanent marks of their status as a captive enemy, especially when such wounds occurred in conspicuous locations like the face or hands. Maiming the hands also served another purpose: preventing escape or rebellion. Describing a similar strategy used by the Iroquois, one Jesuit observed, “They began by cutting off a thumb of each [captive], to make them unable to unbind themselves.” According to one account, Algonquians adopted this practice to avenge those captured by their Iroquois enemies. The resulting scarring and disfiguration were considered “the marks of their captivity,” which remained with living captives long after the trauma of initiation had passed.41 Strategic slave marking registered in the Algonquian languages of the Pays d’en Haut. The Anishinaabes used a phrase that Louis André translated as “I cut a young slave to mark him.” The Illinois had a whole family of expressions dealing with personal marking of slave bodies, all derived from the root word iscᴕ, meaning “mark of imperfection / defect.” These included iscᴕchita, “someone who has a cropped ear”; iscᴕchipagᴕta, “bitten on the ear, ear removed with the teeth”; nitiscᴕicᴕrepᴕa, “I crop his nose with my teeth.” Other Indians of the Pays d’en Haut marked men in an especially painful way, using “red-hot javelins, with which they pierced the most sensitive parts of his body.” Even these physical markers of slavery narrated the act of enslavement as domestication, emphasizing the very personal power exercised over these enemies by capturing warriors. Like enslavement itself, biting the tip from a captive’s nose or ear was at once an alienating and terribly intimate act of dominion.42

“Tout le Village assiste à cette derniere Ceremonie” (Bacqueville de La Potherie, Histoire de l’Amérique Septentrionale, II, 26). For a description of the gauntlet among eastern Algonquians, see James Axtell, “The White Indians of Colonial America,” WMQ, 3d Ser., XXXII (1975), 70–71. For the Iroquois, see Daniel K. Richter, “War and Culture: The Iroquois Experience,” WMQ, 3d Ser., XL (1983), 557. 41   Jesuit Relations, L, 39 (“cutting off a thumb”), XLV, 257 (“marks”). For the best description of Iroquoian disfiguration, see Roland Viau, Enfants du néant et mangeurs d’âmes: guerre, culture, et société en Iroquoisie ancienne (Montreal, 1997), 172–186; William A. Starna and Ralph Watkins, “Northern Iroquoian Slavery,” Ethnohistory, XXXVIII (1991), 43–45. For additional examples from the Pays d’en Haut, see Jesuit Relations, XLVIII, 85–101, LXVIII, 171–175. For hand mutilation, see also Gravier and Largillier, “Dictionnaire illinois-français,” 176. 42  “Je coupe un jeu[n] esclave pour marquer”: André, “Preceptes, phrases, et mots de la langue algonquine outaouaise,” s.v. “marquer.” “Qui a l’oreille coupée”; “mordu a l’oreille, oreille emportee avec les dents”; “je luy coupe le nés avec les dents”: Gravier and Largillier, “Dictionnaire illinois-français,” 111, 176. Jesuit Relations, XLVIII, 99 (“red-hot javelins”).

1098

Rushforth

Nose cropping might have survived from a Mississippian rite of incorporation that turned dangerous outsiders into domesticated insiders, claiming warriors’ spiritual power for the capturing village. Ancient Cahokian iconography includes prominent examples of a figure known as the “longnosed god” that adorned amulets and earrings, many of which had bent or cropped noses. Archaeologists link this imagery to traditions in the Pays d’en Haut of potent spirit beings called “thunderers,” who were sometimes twins, one domesticated and the other wild and threatening. Cropping the nose of the wild thunderer tamed him and transferred his significant power to the dominating village. Consistent with the intent of enslavement generally, this rite expressed the prowess of the captor while appropriating the power of the captive. Perhaps this is why enemy warriors perceived as especially threatening were made to “suffer according to their Merits.” Those who cried out during torture were considered less potent and thus less worthy of an honorable death. “When a victim does not die like a brave man,” according to Charlevoix, “he receives his death’s wound from a woman or from children; he is unworthy, say they, to die by the hands of men.” Even in death male captives faced the prospect of emasculation from their enemies.43 After being beaten and marked, slaves were undressed and forced to sing (in some cases the singing began before entering the village). This was the final metaphorical act of stripping slaves’ former identities from them, preparing them for death or the forced integration that would follow. At least one Algonquian language made explicit the connection between the humiliation of stripping and slavery: Illinois-speakers used the phrase nikiᴕinakiha arena, which meant both “I lift up his loin cloth” and “I treat him like a slave.”44 Once the initial tortures subsided, another round of sorting divided captives marked for death from those who would stay alive as slaves. Among the Illinois, male heads of household “assemble and decide what they will do with the prisoner who has been given to them, and whether they wish to give him his life.” Hennepin came to understand that his fate had been decided in the same way among the Sioux. “When the warriors have entered their lodges, all the elders assemble to hear the account of all that has happened in the war, then they dispose of the slaves. If the father of an Indian woman was killed by 43  For the long-nosed god and thunderers, see James R. Duncan, “Of Masks and Myths,” Midcontinental Journal of Archaeology, XXV (2000), 1–26. Hennepin, New Discovery, I, 186 (“according to their Merits”). P. de Charlevoix, Journal of a Voyage to North-America (London, 1761), II, 107. 44  “Je luy oste son brayet, la traite en esclave”: Gravier and Largillier, “Dictionnaire illinoisfrançais,” 209.

I Make Him My Dog / My Slave

1099

figure 6 “ Woman Who Condemns to Death the Prisoner Given to Her” and “Woman Who Gives Life to the Prisoner Given to Her.” [Louis-Armand de Lom d’Arce], baron de Lahontan, Nouveaux voyages … dans l’Amérique Septentrionale. The Hague, 1703. Courtesy Special Collections Research Center, Swem Library, College of William and Mary.

their enemies, they give [her] a slave in his place, and the woman is free to give him life or have him killed.” The Anishinaabes did much the same, granting life to some and subjecting others to a slow and painful death. Although the particular reasons for sparing individual captives varied from family to family and village to village, captives could be kept alive to augment population growth, to replace a dead relative, or to facilitate alliances through trade. Once the captive had been granted life, he or she was washed, clothed, and given a new name, often that of the deceased he or she was intended to replace. One Illinois word describing the decision to grant life to a captive derives from the word meaning “to cure or heal.”45

45  Deliette, “Memoir,” 384 (“assemble and decide”). “Quand les guerriers sont entrés dans leurs cabannes, tous les anciens s’assemblent pour entendre la relation de tout ce qui s’est passé en guerre, ensuitte ils disposent des Esclaves. Si le pere d’une femme Sauvage a esté tué par leurs ennemis, ils luy donnent un Esclave à la place, et il est libre à cette femme de luy donner la vie ou de le faire mourir.” Hennepin, Description de la Louisiane, 65–66. For post-torture healing and naming, see Bacqueville de La Potherie, “History of the Savage Peoples Who Are Allies of New France,” in Blair, ed. and trans., Indian Tribes, II, 36–43; Daniel K. Richter, Ordeal of the Longhouse: The Peoples of the Iroquois League in the Era of European Colonization (Chapel Hill, N.C., 1992), 50–74; Viau, Enfants du néant, 137–160. For “nimpelakiihaa,” see Antoine-Robert Le Boullenger, “Dictionnaire français-illinois,” MS, ca. 1720s, John Carter Brown Library, Brown University, Providence, R.I.; Daryl Baldwin, personal communication.

1100

Rushforth

Captives marked to die were forced to sing what the French described as “chansons de mort,” or death songs, according to Pierre Deliette, “to afford entertainment to their executioners.” François de Montigny witnessed the spectacle when a Winnebago war party passed by an Illinois village “in triumph” with two Missouri slaves, “who were forced to sing their death songs, which is a custom among all the Indians.” Called kikitᴕinakiᴕa, meaning “slave songs,” by the Illinois, captives were forced to sing at the entrance of each household that had lost a family member to the captive’s people, allowing grieving kin a chance at violent (or at least verbal) catharsis. Condemned slaves were handed a special staff and forced to march from cabin to cabin as they sang. At ten to twelve feet long, the staff was wrapped in feathers to signify the captors’ otherworldly power over the slave. It must have also become a physical burden to captives who had to carry it around the village for hours. The staff was eventually planted in the ground to become a torture stake, where condemned slaves spent their final hours enduring a slow, smoldering death.46 Captors spared women and children more often than men. In addition to targeting the male warriors for revenge killings, this strategy maximized the demographic benefits of slavery, as increasing the number of adult males in a village would do little to change its reproductive capacity. During times of high mortality due to disease or warfare, female captives often represented the best hope for rapidly restoring lost population. Especially in the frequently polygynous societies of the Pays d’en Haut, female captives integrated smoothly into present social structures as second or third wives of prominent men. Children were especially prized because of the relative ease with which they assimilated into the capturing society, learning new languages and customs much more quickly than older captives. This selection process left a surplus of male captives, who were often traded outside the village. Welcomed communally, slaves were controlled individually. The warrior who captured each slave exercised mastery over the person as a private possession, and any family wishing to kill or adopt a slave would have to negotiate terms with the original master. This

46  Deliette, “Memoir,” 383 (“chansons de mort”); Jesuit Relations, XLV, 183 (“entertainment”). “En triomphe … ces pauvres prisonnieres qu’on obligeoit de chanter leur chanson de mort, qui est une maniere qu’ont tous les sauvages”: the Illinois demanded the slaves’ release because they were Missouris, “comme ayant toujours esté amis” (“as they had always been friends”): François de Montigny, “Lettre de M. de Montigny sur les missions du Mississippi,” Aug. 25, 1699, 1–2, ASQ, SME 12.1/001/041. For the torture staff, see Deliette, “Memoir,” 383–384; Gravier and Largillier, “Dictionnaire illinois-français,” 573.

I Make Him My Dog / My Slave

1101

process could lead to conflict among the captors as the ultimate future of each slave became a matter of group deliberation.47 Of the many possible fates facing a captive who survived the rituals of domestication, the most familiar to modern readers is a captive’s adoption into a household to take the place of the dead. If it was not the most common outcome, it was certainly the status that French observers recognized most readily. “When there is any dead man to be resuscitated, that is to say, if any one of their warriors has been killed, and they think it a duty to replace him in his cabin,” wrote Sébastien Rale, “they give to this cabin one of their prisoners, who takes the place of the deceased; and this is what they call ‘resuscitating the dead.’” In times of peace this role was played by other members of the same village, who took the place of prominent villagers who had died, thereby assuming their full identity and status. Nicolas Perrot insisted that among the Ottawas a dead person of high status was sometimes replaced by another resident of the village, “and they regard themselves as united to this family, as much as if they were actually kindred.” But because the adoptee “must be of the same rank” as the dead, captives were rarely chosen to replace influential men and women.48 What captive adoption meant in Native societies is elusive at best. French law and culture granted an adopted father an essentially proprietary authority over the adopted child, and this shaped what French observers meant to convey with the notion of captive “adoption,” never intended by French authors to indicate the creation of true kinship. Even captives themselves, like Louis Hennepin, found their relationship with adopted relatives hard to comprehend, not familial in any sense that they recognized yet still expressed as kinship. When it came to difficult choices, as it did with Sioux food supplies, Hennepin acknowledged that his adopted kin favored their actual relatives over the fictive kinship created by his ceremonial adoption. Transcending the limits of European observations and filling in their silences, indigenous languages provide a glimpse of how Indians themselves understood the category 47  Richter, Ordeal of the Longhouse, esp. 67–68; Gordon M. Sayre, Les Sauvages Américains: Representations of Native Americans in French and English Colonial Literature (Chapel Hill, N.C., 1997), 248–304. For Illinois social structure, see Susan Sleeper-Smith, Indian Women and French Men: Rethinking Cultural Encounter in the Western Great Lakes (Amherst, Mass., 2001), esp. 1–37, where she indicates the importance of women to integrating outsiders into Illinois kin structures. For a similar captive selection process among the Indians of the Southwest, see James F. Brooks, Captives and Cousins: Slavery, Kinship, and Community in the Southwest Borderlands (Chapel Hill, N.C., 2002), esp. 1–40. 48   Jesuit Relations, LXVII, 173 (“resuscitating”); Perrot, Memoir, in Blair, ed. and trans., Indian Tribes, I, 84 (“same rank”), 85 (“actually kindred”).

1102

Rushforth

of kinship created by captive adoption. Their own metaphors suggest meanings much more complex than colonists understood.49 When describing captive adoption, speakers of Miami-Illinois used the word nirapakerima, translated by Jacques Gravier as “I adopt him in place of the dead.” The key stem of this word is rapa, which means “in the place of another.” Forms of this word were used in many other settings that clarify the importance of place to its meaning. Rapahᴕteᴕi, for example, meant “attached/ tied to a place,” suggesting that the captive was more connected to the place of the dead than to his or her persona. The idea of the captive as a placeholder was implied by many other uses of the stem rapa. Stepping into a hole or putting a foot into stockings or boots, for example, were both expressed by a verb form of rapa to express the placement of an item into a gap, an idea that Jacques Gravier recorded as “entredeux,” or between two things. The death of a relative left a gap, an open space, that was occupied by the adopted captive, who served as a replacement.50 Understanding captive adoption as a link to place as much as to the people in it helps to explain the Illinois and Anishinaabe word for master: entachita. This word literally translates as “I stay at his place for a long time,” which is related, not to kinship, but to domain. What the French saw as incorporation into a family Algonquians expressed as incorporation into a household, tying captives to the place of the master. Other forms of this word, also used by the Anishinaabes, mean “where he dwells,” “at a certain place,” “at his house,” or “to stay in a certain place for a period of time.” Jacques Gravier, who first translated this term from Miami-Illinois, rendered it “he who has me as a slave,” or my master.51 Kinship terminology was conspicuously absent from Algonquian idioms describing adopted captives. Rather than using the common verb nintoohsimaa, 49  Kristin Elizabeth Gager, Blood Ties and Fictive Ties: Adoption and Family Life in Early Modern France (Princeton, N.J., 1996); Hennepin, Nouvelle découverte, 362. Compare this to Perrot’s report that Anishinaabe wives could turn to their extended kin when they needed protection or redress. Perrot, Memoir, in Blair, ed. and trans., Indian Tribes, I, 64–65. 50  Gravier and Largillier, “Dictionnaire illinois-français,” 493. 51  Ibid., 168. C.f. Richard A. Rhodes, Eastern Ojibwa-Chippewa-Ottawa Dictionary (New York, 1993), 94: “daad vai live in a certain place, endaad ‘where he dwells, at his house … daadaad vai live in a certain place for a period of time, Ot; pres ndaadaa; cc e-ndaadaad, endaadaad”; Nichols and Nyholm, Dictionary of Minnesota Ojibwe, 42: “dazhi-pvƷ in a certain place, of a certain place, there” and “dazhitaa vai spend time in a certain place, play in a certain place; 1sg indazhitaa; prt endazhitaad,” “endazhi- pvƷ in a certain place,” “endaad vai his/ her home.”

I Make Him My Dog / My Slave

1103

“I have him as a father,” for example, Illinois captives identified their master by the household where they stayed. Only two recorded kinship terms applied to adopted captives or slaves. The first was an expression used by families who wanted to kill a slave from a returning war party. They would say to the warrior who captured the slave, nitaᴕembima, a unique word form meaning “that is my relative,” which was said only “by the executioner to whoever brings a slave” to the village. Rather than indicating actual kinship, the term condemned the captive to death in memory of the dead relative. The Jesuit Thierry Beschefer recorded the presence of a second kinship term in 1683, specifically used to demean adoptees. Using a different form of the word son, captors signified “a submission of which They make use to command us, as They do the Slaves whom they have adopted.” According to a French trader, among the Anishinaabes adopted captives “never lie in their Masters Huts,” another mark of distinction from the household’s actual kin similar to the exclusion experienced by Hennepin among the Sioux. The master was, in the words of Pierre Chaumonot, a “feigned parent,” or fictive kin.52 Adopted slaves, then, were bound to a household of fictive kin, occupying the physical and metaphorical place of a child but constantly aware that they were not actual relatives. Slaves’ history and the terminology used to describe them equated them more with the family’s domestic animals than with their children. And because they had no actual kin but were attached to a household at the master’s pleasure, they were bound to the family at a single point rather than through the multiple lines created by kinship. Like dogs, their linguistic equivalents, adopted slaves were thus part of the household but never really part of the family. The presence of fictive kinship bonds created by slaves’ ritual adoption followed a pattern of linking family and slavery in a wide range of historical slave systems. The expression of mastery in familial terms—what sociologists of slavery call “quasi-filial” kinship—pervaded slaveries from ancient Rome to the antebellum southern United States. Indeed, the English word family derives from the Latin famulus, meaning household slave. Across vast cultural differences, masters have imagined themselves fathers, but they have always 52  “Cest mon parent, dit le boureau a qui conque amene un esclave”: Gravier and Largillier, “Dictionnaire illinois-français,” 36. For Illinois kinship terms, see David J. Costa, “The Kinship Terminology of the Miami-Illinois Language,” Anthropological Linguistics, XLI (1999), 28–53; Jesuit Relations, LXII, 213 (“to command us”). Pierre Deliette confirmed Beschefer’s sense that calling adoptees “son” was a mark of disrespect. See Deliette, “Memoir,” 363. For “never lie in their Masters Huts,” see Lahontan, New Voyages, II, 37. For “feigned parent,” see Jesuit Relations, XVIII, 29.

1104

Rushforth

understood the difference between their slaves and their biological family. This was especially evident in the particular language of kinship used by those adopting slaves in the Pays d’en Haut, where they made careful distinctions between real and fictive sons, between actual kin and those forced to take their place in the household.53 Slaves’ ritual subordination worked together with the language of fictive kinship in the context of diplomacy, as encapsulated in a 1674 French-Illinois encounter. When Jacques Marquette and Louis Jolliet completed their journey down the Mississippi River, they returned with a young male slave who had been given to them by a Peoria chief. As they arrived at the chief’s village, he offered them a warm welcome, expressing his hope that the French and his people could form a new alliance. The past twenty years had not been kind to his people. Threatened by the Iroquois from the east and the Foxes and Sioux from the northwest, the Peorias hoped that the French visit signaled better times to come. Since rumors and goods traveled faster than the French themselves, the chief already knew of these strange and powerful foreigners, and indeed Marquette noted the presence of European cloth in the village. A few Peoria men had traded with the French and their Ottawa allies over the past few years, and he eagerly offered them his friendship. Inviting the travelers to stop and rest, he assured them of his good intentions. “How beautiful the sun is, O Frenchman, when thou comest to visit us! All our village awaits thee, and thou shalt enter all our Cabins in peace.”54 Wasting no time, the chief presented his French guests to the “great Captain of all the Illinois,” who lived nearby. This chief offered the French three gifts as a pledge of alliance. “The Captain arose,” according to the French account, “and, resting His hand upon the head of a little Slave whom he wished to give us,” promised his friendship. “Here is my son,” the chief said, “whom I give thee to Show thee my heart.” After placing the slave in front of the French, the chief then offered “a second present, consisting of an altogether mysterious Calumet,” a pipe attached to a reed shaft adorned with colorful feathers. A feast 53  Orlando Patterson, Slavery and Social Death: A Comparative Study (Cambridge, 1982), 62–65; Oxford English Dictionary, s.v. “family.” In 1576, the French political philosopher Jean Bodin wrote, “For the very name of a Familie, came of Famulus and Famulatio, for that it had in it a great number of Slaves: and so of the greatest part of them that are in subjection in the Familie, men call all the whole household a Familie; or els for that there was no greater means to gather wealth than by slaves and servants, which the Latines call Famuli, the auntients not without cause have called this multitude of Slaves and servants a Familie.” Bodin, The Six Bookes of a Commonweale, ed. Kenneth Douglas McRae (Cambridge, Mass., 1962), 32. 54   Jesuit Relations, LIX, 117.

I Make Him My Dog / My Slave

1105

followed at which several Illinois Indians fed the Frenchmen, “as one would give food to a bird.” These three tokens of friendship—a slave, a feathered pipe, and a meal—converged in a symbolic expression of Illinois strength and generosity to convince the French of their good intentions.55 When Marquette’s new slave arrived in the village as a recent captive, he would have faced a terrible inversion of these symbolic offerings. Rather than giving him food, he and his people had become the food, consumed as dog’s flesh in the flames and by their enemies in acts of violent dominion. Rather than receiving a calumet, the feathered staff he carried would have been a slave stick as he and his kin were forced to face their potential executioners. Rather than receiving a slave, he received a master, who expressed his relationship to the boy in terms of dominion rather than mutual friendship. His defeat was now replayed for Marquette as evidence of Illinois power—his submission reinforced as his fictive father gave him to a stranger as a diplomatic object. This young slave embodied the dual imperatives of indigenous slavery in the Pays d’en Haut, balancing the impulse for war with the drive for diplomacy and trade.

Nitaïa—My Possession / My Slave I mark my slave to make him known [as mine]. Algonquian Phrase Book, ca. 1690s

If adoption emphasized slaves’ status as their captors’ fictive kin, Algonquian and Siouan sources make clear that even adopted slaves were owned and controlled by a master, dishonored by all within the village and living under constant threat of violence even long after their incorporation. In short, they became a form of property, described with the same words used to identify other personally owned objects like tools, food, clothing, or weapons. Language and ritual reduced slaves to the status of animals, and like dogs they became the personal possession of their masters. Understanding the status of dogs can clarify what Indians meant when they compared slaves to their only truly domesticated animal. Dogs played many important roles in the households and villages of the Pays d’en Haut, ranging from the sacred to the mundane. Indians first domesticated dogs from wolves several thousand years ago, creating the only domesticated animal in most of North America. Biologists consider wolves “preselected” for domestication because they are social animals 55   Jesuit Relations, LIX, 119, 121 (“a little Slave”), 123 (“food to a bird”).

1106

Rushforth

that recognize and respect a dominant alpha as their master. This social and submissive behavior is an essential characteristic for domesticated animals because it allows them to function as subordinates within human society. The domestication of dogs was a striking exception to the standard Native relationship with nonhuman animals as independent creatures demanding reciprocity rather than yielding to dominion and mastery. Dogs’ cousins, wolves and foxes, were seen as powerful and independent spirit beings who demanded respect. Dogs, on the other hand, could be owned, trained, abused, and harnessed for human labor. They served their masters faithfully, but Indians studiously avoided spoiling them with food and affection to keep them dependent and eager to please.56 Dogs also figured prominently in Native rituals, both those enacting success in war and others seeking spiritual power for hunting, fishing, or healing. The Ottawas used dogs in the feast of the dead, as an item of food and an object of sacrifice. “Before eating,” wrote Antoine Lamothe de Cadillac, “they set up two great poles and fasten a dog to the top of them, which they sacrifice to the sun and the moon, praying to them to have pity and to take care of the souls of their relations, to light them on their journeys, and to guide them to the dwelling place of their ancestors.” The Ottawas also drowned dogs as a sacrifice to a powerful manitou that lived in Lake Huron, hoping the spirit would reciprocate with a plentiful harvest of sturgeon. According to Antoine-Denis Raudot, “They threw dogs, with their jaws and legs tied, into the water as a gift for him.” Alexander Henry confirms the persistence of this ceremony in the Great Lakes a full century after European settlement. “On our passage” through the Pays d’en Haut, he wrote, “we encountered a gale of wind, and there were some appearances of danger. To avert it, a dog, of which the legs were previously tied together, was thrown into the lake; an offering designed to soothe the angry passions of some offended Manito.”57 The Sioux made dog sacrifices at a specific falls in the Mississippi, evoking language that linked the dog with slavery. “Thou art a Spirit,” they would say, 56  Virginia DeJohn Anderson, Creatures of Empire: How Domestic Animals Transformed Early America (New York, 2004), 7–8, 17–18; Marion Schwartz, A History of Dogs in the Early Americas (New Haven, Conn., 1997), 1–28. See also Gravier and Largillier, “Dictionnaire illinois-français,” 400, which suggests private ownership of domesticated animals, translating “ᴕissᴕerimᴕeᴕa, ᴕissᴕerimᴕeta” as “qui a plusieures chiens, bestes privees,” or “someone who has several dogs, privately owned animals.” For more on dogs and hunting, see Raudot, “Memoir,” in Kinietz, ed., Indians of the Western Great Lakes, 364. 57  Raudot, “Memoir,” in Kinietz, ed., Indians of the Western Great Lakes, 283–284 (“two great poles”), 286 (“jaws and legs tied”). For Alexander Henry, see Schwartz, History of Dogs, 99.

I Make Him My Dog / My Slave

1107

“grant that Those of my Nation may pass here without any Disaster; That we may meet with a great many wild Bulls; and that we may be so happy as to vanquish our Enemy, and take a great many Slaves.” Later, captured slaves were returned to the falls to acknowledge that success was tied to the sacrifice. Like slaves marked for sacrifice, dogs destined to die were singled out for special ritual attention58 But for most dogs, as for most slaves, daily life in the village was driven more by practical than ritual concerns. Dogs’ main function in Indian society was hunting. “The Indians always bring a large number of dogs on their hunts,” wrote Pierre-François-Xavier de Charlevoix. “These are the only domestic animals that they raise, and they only raise them for hunting … they are very faithful and very attached to their masters, who nevertheless feed them very little and never pet them.” When Henri Joutel traveled down the Mississippi with the La Salle expedition, he found Indians’ dogs so hungry that they chewed the leather straps on the Frenchmen’s horses. “The dogs were always hungry because the Indians did not feed them too well,” explained Joutel, “particularly in this season when they were in the height of corn production (unless they were hunting).” One Mississippi Valley people “tied up the mouths of the dogs and fastened a forepaw under the throat so they could not get to the stalks of corn” during harvest. Archaeological remains reveal that some Indians even filed down their dogs’ teeth with stones to make them less dangerous to their masters or other handlers59 In the Mississippi Valley and the Great Plains, dogs packed heavy loads of meat, fish, and other supplies “on a travois”: a sled or a woven tarp attached to the dog by two long poles or straps. The poles were twelve feet long and “the thickness of a man’s arm,” connected to a leather “saddle” tied around the dog’s torso. In 1724, Étienne de Bourgmont encountered a large group of Kansa Indians in the Missouri River valley who loaded their dogs with a pile of bison hides that he estimated to weigh three hundred pounds. For lighter loads, 58  Hennepin, New Discovery, 278 (“Thou art a Spirit”). For special treatment for sacrificial dogs, see Schwartz, History of Dogs, 85. 59  “Les Sauvages menent toujours à leurs chasses un grand nombre de Chiens; ce sont les seuls Animaux domestiques, qu’ils élevent; et ils ne les élevent, que pour la chasse … ils sont fort fidéles, et fort attaché à leurs Maîtres, qui les nourissent pourtant assez mal, et ne les caressent jamais”: Charlevoix, Journal d’un voyage, V, 176; Joutel, Journal, ed. Foster, trans. Warren, 236 (“always hungry”), 241 (“tied up the mouths”). For filing teeth, see Schwartz, History of Dogs, 53. See also Lahontan, New Voyages, II, 44. Summarizing the distinctive role of dogs in Native American cultures, Schwartz writes, “Our own concept of the pet dog must be discarded … [or] we will lose sight of the complexity and richness of the lives of those who inhabited a world far different from our own” (3).

1108

figure 7

Rushforth

“Dog Dragging a Fish Called Namé or Sturgeon.” Louis Nicolas, “Codex canadiensis.” MS, ca. 1670. Permission Gilcrease Museum, Tulsa, Ok.

especially when snow made the dragging easier, leather collars were placed around dogs’ necks and attached to sleds with far more flexible leather straps.60 In the 1680s a traveler highlighted the skill and discipline of Indians’ dogs by contrasting them to a French dog from La Rochelle who joined them on a bison hunt. Frenzied by the thunderous stampede, the French dog had no sense of how to behave, running wildly after the animals and chasing off the whole herd. Through intensive training and strategic starvation, Mississippi Valley dogs were taught to chase individual animals away from the pack, steering them toward the hunters. Male hunters controlled hunting dogs, but women controlled those used as pack animals, relying on them to share what was throughout the Pays d’en Haut the exclusively female burden of carrying loads on hunting and fishing expeditions. Other animals were also tamed and brought into the village to serve human purposes but never actually domesticated. La Salle commented on the similarity of the Illinois practice of 60  Étienne de Véniard, sieur de Bourgmont, “Journal of the Voyage … to the Padoucas,” in Frank Norall, ed., Bourgmont: Explorer of the Missouri, 1698–1725 (Lincoln, Nebr., 1988), 136–137.

I Make Him My Dog / My Slave

1109

capturing and rearing young bison with their treatment of slaves. “The buffalo calves are easy to tame and can be of great use,” he wrote in 1680, “as well as the slaves in which these people are accustomed to traffic and whom they compel to labor for them.”61 If the beginning and end of captive raids were collective affairs, the act of taking slaves was intensely individualistic, as was the right to their bodies, which was one of the most personal forms of property among the Indians of the Pays d’en Haut. Lahontan’s claim that Indians were “utter Strangers to distinctions of Property, for what belongs to one is equally anothers,” was an exaggeration that overlooked several categories of private property ownership and exclusive control, including domesticated animals and captives, both of which were controlled by individuals and families rather than villages or clans. Although the accumulation of private property did not translate into higher status, it did pay practical dividends to control both animals and people who could produce for the master’s household. Speakers of Anishinaabemowin put the possessive to the term for dog to represent slave ownership. Pierre du Jaunay recorded that the Ottawa term nit’aouakan meant both “my dog” and “my slave.” One family of Illinois terms derived from nitaïa, which was used to denote “my domestic animal, my dog, my cat, also my slave.” Literally, the word meant “my possession.” Gravier rendered another form of the word, nitataïma, as “I have him/her as a slave, as a domestic animal. That is my slave.” When slaves were distributed throughout the village, those receiving slaves would say nitintarehigᴕa, “he gave him to me, or he made him my property.” Whether a dog or a slave, to make someone the master of something or someone was nitintaretaᴕa. Another variant, nitaïagᴕa, meant both “I am his slave, he is my master,” and “he wagered me in a bet.” Although apparently used sparingly, the expression speaks to masters’ ability to dispose of slaves as they saw fit. Another Illinois word describing physical detention, kikiᴕnakiᴕi, meant “slave” in the seventeenth century, but in modern usage the root has softened into a meaning that suggests being detained, held up, or merely running late. As used in the seventeenth century, the word rendered the slave in an inanimate

61  For strategic starving, see Joutel, Journal, ed. Foster, trans. Warren, 115, 181. For carrying, see Bourgmont, “Journal of the Voyage to the Padoucas,” in Norall, ed., Bourgmont, 136–137; Lawrence J. Burpee, ed., Journals and Letters of Pierre Gaultier de Varenne de La Vérendrye and His Sons … (Toronto, 1927), 317–318. “Les petits beufs sauvages sont aisez a apprivoiser et peuvent etre d’un grand secours aussi bien que les esclaves dont ces gens ont coutume de faire commerce et qu’ils obligent de travailler”: “La Salle on the Illinois Country, 1680,” in Pease, ed., The French Foundations, 10.

1110

Rushforth

intransitive form, the only of its kind referring to a person. The slave was thus grammatically as well as socially dead.62 Other words expressed the presence of verbal or physical markers designed to set a particular captive apart as belonging to a particular master. According to a seventeenth-century Jesuit, the Illinois used the term nikikipenara, meaning “I mark my slave to make him known [as mine].” The Ottawas had a similar term that Louis André recorded with the meaning “I cut a young slave to mark him.” A French trader greeted an Ottawa war party returning from a raid in the 1690s. As their canoes approached, he heard “the songs of the slaves, who stood upright, each having a wand in his hand. There were special marks on each, to indicate those who had captured him.” Warriors also used specific marks to signify their own success in slaving. Some made notches on their war clubs to indicate the number of prisoners captured or killed, as emblems of masculine honor. Others tattooed themselves with simple lines or images of war clubs, battle scenes, or the totemic symbols of their defeated enemies. And both within and beyond the calumet ceremony, all Indian warriors in the Pays d’en Haut affirmed their masculinity by performing complex dance rituals, during which they recounted their killing and capturing of enemies. As a check against the inevitable temptation to exaggerate, those who falsely boasted were publicly shamed by having ashes rubbed on their heads by other warriors.63 62  Lahontan, New Voyages, II, 7 (“utter Strangers”). Yet the Algonquian and Siouan languages spoken by the peoples of the Pays d’en Haut are full of references to items controlled as private property. See, for example, Gravier and Largillier, “Dictionnaire illinois-français,” 168–169 (and “nit’aouakan”); Du Jaunay, “Dictionarium gallico-outaouakum,” s.v. “esclave.” For the literal translation, see Masthay, Kaskaskia Illinois-to-French Dictionary, 122 (for original, see Gravier and Largillier, “Dictionnaire illinois-français,” 14, 19). For “kikiᴕnakiᴕi,” see Pierre Pinet, “Dictionnaire” (commonly referred to as the “St-Jérôme Dictionary”), 210. The author thanks Michael McCafferty for bringing this word to his attention. Thanks, also, to Daryl Baldwin for the modern usage. For details about Pinet’s dictionary, see McCafferty, “The Latest Miami-Illinois Dictionary and Its Author,” Papers of the Thirtysixth Algonquian Conference (2005), 271–286, and David Costa, “The St-Jérôme Dictionary of Miami-Illinois,” 107–133; Appendix A. 63  For “nikikipenara,” see Gravier and Largillier, “Dictionnaire illinois-français,” 199; Masthay, Kaskaskia Illinois-to-French Dictionary, 138. “Je coupe un jeu[n] esclave pour marquer”: André, “Preceptes, phrases, et mots de la langue algonquine outaouaise,” s.v. “marquer.” For “songs of the slaves,” see Bacqueville de La Potherie, History, in Blair, ed. and trans., Indian Tribes, II, 37–38. For marking, see Raudot, “Memoir,” in Kinietz, ed., Indians of the Western Great Lakes, 351. For tattooing, see Arnaud Balvay, “Tattooing and Its Role in French— Native American Relations in the Eighteenth Century,” French Colonial History, IX (2008), 2–3. For the calumet dance and the shaming of liars, see Raudot, “Memoir,” 347–348; and Perrot, Memoir, in Blair, ed. and trans., Indian Tribes, I, 54–59.

I Make Him My Dog / My Slave

1111

Warriors’ claim on slaves’ bodies began as an exclusive form of ownership. Although village politics demanded that slaves be distributed upon their arrival, the captor was never compelled to release a slave he preferred to keep, and those wishing to claim slaves for themselves had to compensate him for their release. Although this was not a sale in the sense that the human body had been assigned a price, it was nevertheless a recognition of the captor’s right to possess the captive, and, as in all redistributive gift-giving, generosity brought additional prestige to the successful warrior.64 Subdued and controlled by male warriors, in the long run most indigenous slaves spent much of their time supervised by women, who in the cultures of the Pays d’en Haut generally managed the domestic resources and spaces on which slaves depended. Hennepin experienced this gendered control, constantly reminded by the women who supervised him that he was a slave and that he competed with their natural children for resources. Households, even the male-dominated cultures of the Pays d’en Haut, were run by women, making them what anthropologists call “matrifocal” (in contrast to “matriarchal,” which implies a much broader reach for women’s power in the society). Men were absent from the village for extended periods during hunting and raiding expeditions, leaving women as the primary guardians of slaves of all sexes and ages. Male agricultural laborers not only had to do women’s work; they also labored under the supervision of women, a doubly demeaning fate that further differentiated male slaves from their fictive male kin.65 Female slaves who became secondary wives also had to negotiate a complicated relationship with the woman who managed the household: the senior wife. Blending into the gendered hierarchies of a household headed by a man but practically controlled by another woman, captive women faced the usual challenges of authority and power inherent to polygamy complicated by their status as a degraded outsider. One term for a subordinate wife, nᴕkiᴕiᴕagana, suggests the low status of this role. Meaning “the other wife, as in the second wife,” it was a “term of contempt” that contrasted with kitcahiᴕeta, meaning “the most loved of his wives who is the mistress of all.” If the master was a successful hunter or, especially, a prominent chief, a female captive’s painful path into his household could result in significant social mobility, especially for their children. The Illinois used what Jacques Gravier considered a fairly common 64  Deliette, “Memoir,” 384. 65  For “matrifocal” households, see Richard R. Randolph, “The ‘Matrifocal Family’ as a Comparative Category,” American Anthropologist, LXVI (1964), 628–631; Raymond T. Smith, “The Matrifocal Family,” in Jack Goody, ed., The Character of Kinship (New York, 1973), 121–144; Sleeper-Smith, Indian Women and French Men, 30.

1112

Rushforth

phrase, kikiᴕnaᴕiba nᴕnghi ᴕirᴕnigᴕtchi anapemari tchirakiᴕeᴕa, roughly translatable as “slave woman who is now better accommodated [because] her husband is rich.”66 Demonstrating that slaves functioned as a kind of privately possessed property, the Illinois and Miamis offered slaves as both bride price and dowry. The suitor’s “father, if he has one, or his uncle in lieu of him, takes five or six kettles, two or three guns, some skins of stags, bucks, or beavers, some flat sides of buffalo, some cloth, and sometimes a slave, if he has one, in short something of all he has, according to his wealth and the esteem in which the girl is held.” The practice also went the other direction, with a woman’s giving a slave to her new husband, conceivably one who could serve as her minion, making the slave a gift that benefited both husband and wife. Paying a dowry with a slave was sufficiently standardized to develop into a common Algonquian phrase, ᴕchitᴕa kicᴕ[naᴕa] ᴕchiheᴕa atintaᴕaganari, niarinta aᴕira, awkwardly glossed by Jacques Gravier as “she carries something, brings a slave to her husband. [said] of a young bride to whom has been given a slave that she brings to her husband’s home.”67 Being given away as a gift must have reminded Native slaves of the tenuous protections they received from their adopted kin. This degradation was constantly reinscribed with threats of violence and insulting language. As a consequence of low status, adopted outsiders lived under the perpetual threat of violence and murder. Even after years of successfully participating in the capturing society, as one historian noted of the Iroquois, “a recalcitrant captive could expect a quick and unceremonious death.” Illinois and Ottawa slaves could face violence as a consequence of their master’s actions. In the 1690s, when an Illinois woman angered her brother, his rage fell upon “one of her slaves—whom he cruelly killed, out of revenge.” The Illinois phrase meaning “kill my wife when I am dead” most likely referred to enslaved, subordinate wives, rather than to those who could turn to their kin for protection.68 66  “L’autre femme, comme la seconde f. terme de mepris”: Gravier and Largillier, “Dictionnaire illinois-français,” 355, and “la plus aimée de ses femmes qui est maitresse de tout,” 224. For “better accommodated,” see 399. The translation is based on Masthay’s gloss of the Algonquian term in Masthay, Kaskaskia Illinois-to-French Dictionary, 237. 67  Deliette, “Memoir,” 331 (bride price). “Elle porte qq chose, mene une esclave a son mary. d’une jeune mariée a qui on a doné une esclave quon mene chez le mary”: Gravier and Largillier, “Dictionnaire illinois-français,” 377. 68  Richter, Ordeal of the Longhouse, 69 (“recalcitrant captive”); Gravier and Largillier, “Dictionnaire illinois-français,” 404. For the protection of kin, see Perrot, Memoir, in Blair, ed. and trans., Indian Tribes, I, 64–65.

I Make Him My Dog / My Slave



1113

Nitarᴕrieᴕo—He Makes Me Work Aquipaguetin also sent me to a neighboring island with his children and some women to work the land. Louis Hennepin, Sioux captive, 1683

Dishonored and brought to heel by their captors, slaves were also domesticated in another sense: they performed valuable tasks for the domus, or household, to which they were bound. Although Indians did not demand that slaves produce valuable export commodities, indigenous slaves did contribute to the overall productivity of the master’s family and village, and they had very little control over the nature or the fruits of their labors. All adoptees were expected to do the work of the person they replaced, thereby mitigating the social costs of that person’s death. As in any preindustrial economy, Indians spent most of their labor on subsistence activities, which they divided along fairly rigid gender lines. Whether male or female, slaves generally played the roles assigned to women and children, enhancing the burden of shame born by male slaves, who would have understood the emasculating symbolism of hoeing in the cornfields or carrying baggage on a hunting expedition. The reproductive labor of female slaves provided additional strength to the master’s household, countering population decline brought by the wars that originally prompted the slave raid.69 “The Women Slaves,” wrote Lahontan of the Pays d’en Haut, “are employed to Sow and Reap the Indian-Corn,” drawing on skills they would have learned from a very young age in their home village. Female slaves were also required to perform routine household chores, like drawing water and gathering wood. Nicolas Perrot noted that Anishinaabe women went for their own water only “if they have no servants in the house.” Many Indian peoples had terms for these “chore wives” that distinguished them from insiders who married under their own will and who had extended families to protect them against a husband’s violence or neglect.70 In the 1680s, Lahontan noted that, among the nations of the western Great Lakes, male captives assisted in the hunt by carrying their masters’ baggage, tending to sled dogs, and preparing animal skins by picking fleas from them and stretching the skins to dry. Especially for the region’s northern inhabitants, 69  Richter, “War and Culture,” WMQ, 3d Ser., XL (1983), 531. 70  Lahontan, New Voyages, II, 18 (“women slaves”); Perrot, Memoir, in Blair, ed. and trans., Indian Tribes, I, 75 (“no servants”). For “chore wives,” see Brooks, Captives and Cousins, 191; Pekka Hämäläinen, The Comanche Empire (New Haven, Conn., 2008), 257.

1114

Rushforth

hunting was the most demanding and time-consuming form of work, making slaves especially desirable for these tasks. Sometimes slaves hunted, but they were not trusted to do so alone. Lahontan also recorded that captives among the Sauks, Potawatomis, and Menominees served food at ceremonial feasts for visitors. Masters sick with disease were “carried about by their Slaves” on a litter, an unpleasant and perilous task. If the master did not survive, slaves sometimes carried his body to the burial place (where family presumably attended). Slaves could also serve as messengers of death, carrying the news throughout the village.71 Many of the tasks assigned to male captives violated expected gender roles, as when they served food, prepared skins, carried packs on hunting expeditions, or performed farm labor, exacerbating the already negative experience of captivity. Louis Hennepin experienced this sort of labor-as-humiliation among the Sioux when he was forced to work alongside women and children in the fields. According to Hennepin, his master, Aquipaguetin, “sent me into a Neighbouring Isle, with his Wives, Children, and Servants, where I was to hough and dig with a Pick-axe and Shovel.” Whether Hennepin even recognized this work as emasculating within Sioux culture is unclear. But his humiliation would have been obvious to everyone around him, and enslaved Indian men would have no trouble recognizing the symbolic degradation of doing women’s work.72 Some of slaves’ most important work occurred beyond the bounds of the village as they acted as both agents and objects of diplomacy. In a general sense, slaves often served as translators and intercultural brokers in the tense but always fluid relationships that produced their captivity in the first place. In rare cases, this work opened avenues of opportunity for slaves to rise in stature as important intermediaries between their people and their master’s. As in other North American regions, the violence that produced captives could also be a means of “cultural interpenetration,” as everything from languages to

71  Lahontan, New Voyages, I, 53, 58, 62, 105, II, 2, 18, 24, 37, 46 (“carried about”), 48, 50–53, 62; Jesuit Relations, LIV, 93–95, LXVI, 263. Given Lahontan’s intent to use Indians’ egalitarianism as a foil to criticize European hierarchy, his description of slavery among western Algonquians is especially telling. If anything, such an admission subverted his ideological agenda. For Lahontan’s motivations, see Paola Basile, “Lahontan et l’évolution moderne du mythe du ‘bon sauvage’” (master’s thesis, McGill University, 1997). 72  Hennepin, New Discovery, II, 171. The original French version reads, “Aquipaguetin me menoit encore dans une Isle voisinne, avec ses enfans et des femmes pour labourer la terre” (Hennepin, Description de la Louisiane, 246).

I Make Him My Dog / My Slave

1115

pottery styles meshed in a violent web of connections that stretched all across North America.73 Slaves could facilitate language acquisition and serve as translators, although their ability to speak the captor’s language could vary widely. For older captives, in particular, the inability to master a new language sometimes limited their effectiveness as diplomatic actors. In 1695, the Jesuit Gabriel Marest “resolved to learn the language of the Savages” in the Lower Mississippi Valley to facilitate relations with a new village. He tried to hire a tutor, but he struggled to learn much, because the man he hired “was a slave from another Tribe, who knew only imperfectly their Tongue.” Yet slaves’ linguistic knowledge was helpful in many contexts, making them important actors in rebuilding relationships among former enemies. At times, their third language was the most helpful. In 1669, one slave, who had been captured by the Illinois and given to the Ottawas as a token of friendship, finally became the slave of Jacques Marquette at Michilimackinac. Having learned to speak Illinois while enslaved among them, then Anishinaabemowin among the Ottawas, he provided Marquette with important geographic and cultural information about the Illinois. He also served as the first translator between the French and the Illinois. Acting as an agent of diplomacy did not free him from his enslavement—in fact, it led to his re-alienation to a new master—but it did give him a measure of independence as he shaped the understanding and influenced the actions of both his Illinois and French masters.74 All peoples in the Pays d’en Haut released enemy captives as a gesture of diplomatic goodwill when they hoped to bring an end to the cycles of revenge that had brought the slave into their village. Secure in their large numbers and never in need of slave labor for agriculture like many of their neighbors, the Sioux in particular were “often content with the glory of winning a victory, and sending back free and uninjured the prisoners taken by them in battle.” They might have adopted this strategy more often than others because they lived on the western margins of this diplomatic world and had no direct access to Great Lakes trade networks except through potentially dangerous territory. Over time, however, reciprocal violence persuaded them that their interests were best served by treating captives as they could expect their own people to be treated. Claude Allouez witnessed the convergence of Fox and Sioux war culture, noting that the Foxes were “fully resolved not to treat The Nadouessi (Sioux) more humanely than they are treated by them,” a reciprocity that

73  Brooks, Captives and Cousins, 31. 74   Jesuit Relations, LIV, 177, LXVI, 105 (“slave from another Tribe”).

1116

Rushforth

would facilitate violence, to be sure, but also one that established protocols of war that could restrain potential attackers.75 Acting as an agent of diplomacy could be perilous work. In one example from the 1660s, a slave among the Potawatomis was released and sent to the Miamis with an invitation to join them against the French. When the slave arrived, he “said many unkind things about the French; he said that the Pouteouatemis held them in the utmost contempt, and regarded them as dogs.” Unfortunately for the slave, there were French traders among the Miamis who heard his speech, and those “who had heard these abusive remarks, put him into a condition where he could say no more outrageous things.” Slaves often carried unwelcome messages back to their own people, as when a group of Anishinaabes sent slaves back to their own nation with a war club painted black to declare war against the slaves’ people.76 Full of ambiguities, captives’ ability to link former enemies could also be complicated by uncertainty about the captive’s intentions, or even identity. Captives taken while very young were hard to distinguish from the captors’ people, given that they learned the captors’ language as children and adopted their style of dress and tattooing. French soldiers witnessed this confusion when a Fox enslaved by the Illinois claimed that he was actually an Illinois. “This unhappy person maintained to his last breath that he was an Illinois, and had been taken when a child by the Outagamies, who had adopted him.” They did not believe him, but his assertion of Illinois identity implies both the possibility of such discoveries and the difficulty of assessing their accuracy. Had he provided a credible Illinois genealogy, he probably would have been freed; but, the younger he was when he had been captured, the less likely he would be to remember his Illinois family.77 75   Jesuit Relations, LV, 171 (“often content”). More than anyone else, the Sioux seem to have released captives as a strategy of ingratiating themselves with potential allies. According to Raudot, “They generally send back any prisoners they make, in hope of obtaining peace; and it is only after they have lost a great many of their men and are tired of sending back prisoners without obtaining the result hoped for, that they burn them. They never torture them” (“Memoir,” in Kinietz, Indians of the Western Great Lakes, 378). Jesuit Relations, LVIII, 67 (“fully resolved”). For the tendency of regional war cultures to develop through reciprocal violence, see Wayne E. Lee, “Peace Chiefs and Blood Revenge: Patterns of Restraint in Native American Warfare, 1500–1800,” Journal of Military History, LXXI (2007), 703. 76  “Perrot Visits the Wisconsin Tribes, and Induces Them to Become Allies of the French,” Wis. Hist. Coll., XVI (1902), 47 (“many unkind things … could say no more”); Lahontan, New Voyages, II, 82. 77  Charlevoix, Journal of a Voyage to North-America, II, 206.

I Make Him My Dog / My Slave

1117

Despite the possibilities available to slaves as agents of diplomacy, their most important diplomatic role came as they were made objects of trade, symbolic gifts that signified peace between peoples. Because of their symbolic power to mitigate the effects of warfare or murder, captives became an important medium of exchange in the gift-giving that characterized Indian diplomacy. Captives accompanied peace delegations as gifts ceremonially offered to allies or erstwhile enemies. “Usually, they are used to replace the dead,” wrote Antoine-Denis Raudot of captives in the western Great Lakes, “but often some of them are also given to other nations to oblige these nations to become their allies.” Employing the language of kinship, those offering slaves introduced them by saying, “Here is my flesh,” to represent the physical blending of familial interest between two previously disconnected groups. Throughout the Mississippi Valley, ceremonial giving of slaves was described alongside the calumet by members of La Salle’s expedition as one of the “ceremonies with which these tribes customarily confirm their alliances.”78 In one of these exchanges a Fox chief received two Iroquois captives from his Algonquian “neighbors [who] took them prisoners and made me a present of them.” Rather than keeping the slaves, he released them back to the Iroquois in an effort to initiate an alliance with them. This act highlights the perils that slaves could pose as a diplomatic gift. Recipients had the option of releasing their newly acquired slaves and thereby forging a new alliance. But such risks were rarely taken, especially when slaves were given as part of the calumet ceremony. A gift of captives, even more powerfully than wampum or the calumet, signified the opposite of warfare, the giving rather than the taking of life. As living witnesses to the power and ferocity of their captors, captives also offered a subtle warning of the dangers one could face as the captor’s enemy. In many cases this gift proved sufficient to erase long periods of violence between two peoples, satisfying the demands of customary justice and symbolizing the possibility of true friendship, as when the Potawatomis “made satisfaction of an old quarrel, by presents of slaves and pipes.” Especially when coupled with the ceremonies of the calumet, slaves symbolized new friendships by demeaning a common enemy.79 78  Raudot, “Memoir,” in Kinietz, ed., Indians of the Western Great Lakes, 341–410, 360 (“given to other nations”); Jesuit Relations, LIX, 121 (“here is my son”); Dubuisson, Official Report, 1712, Wis. Hist. Coll., XVI (1902), 282 (“here is my flesh”); Joutel, Journal, ed. Foster, trans. Warren, 128 (“ceremonies”). 79   Jesuit Relations, LIV, 227 (“made me a present”); Dubuissson, Official Report, Wis. Hist. Coll., XVI (1902), 284 (“presents of slaves and pipes”). For a discussion of “covering the grave” see White, Middle Ground, 75–82.

1118

Rushforth

At other times, a gift of captives could persuade an ally to action against a third party. In 1665, for example, while Nicolas Perrot negotiated an alliance with the tribes around Green Bay, he noted that the Potawatomis offered a captive to the Miamis to persuade them not to enter into an alliance with the French. When attacked by a Sioux war party in 1672, Perrot observed, the Ottawa chief Sinagos fell into captivity. After torturing Sinagos, the Sioux discovered a “Panys who belonged to [this] great chief,” then released the captive “back to his own country that he might faithfully report what he had seen and the justice that had been administered.” The Sioux chief hoped that, by freeing a captive of another western nation, he could persuade the captive’s people to join him against Sinagos.80 Both practically valuable and symbolically potent, captives often passed from village to village through overlapping systems of captive exchange, journeying hundreds or even thousands of miles from their birthplace. The Illinois took captives from the central and southern Plains and traded them into the Lake Superior region. The Ottawas joined Upper Mississippi Valley allies to raid deep into the Southwest, then traded the captives as far east as Lake Nipissing. In 1669 Sulpician missionary François Dollier de Casson described meeting a Nipissing chief who “had a slave the Ottawas had presented to him in the preceding year, from a very remote nation in the southwest.” The next year, Dollier received from the Senecas a gift of two captives, one taken from the Ottawas near Michilimackinac and one from the Shawnees. They were to serve as guides and translators as Dollier and La Salle traveled through the Ohio River valley. Although the Ohio expedition never materialized, the captives remained in Dollier’s possession as slaves. These diplomatic exchanges placed all enslaved captives at risk, reinforcing their own dependence on their hosts’ goodwill for whatever social position they attained. The acts of violence that structured these raid-and-trade networks, producing the captives who passed through them, created barriers more often than they forged connections. Slaves’ potential as intercultural brokers was also inhibited by the memory of their enslavement. Indigenous slaves thus became a symbol of peace only through a double act of violence: their initial capture and their forced subordination to an allied people.81 80  Perrot, Mémoire, in Wis. Hist. Coll. (1902), XVI, 30–31 (“faithfully report”), 46. 81   Jesuit Relations, LIX, 127, LXVII, 171; “The Journey of Dollier and Galinée, 1669–1670,” in Louise Phelps Kellogg, ed., Early Narratives of the Northwest, 1634–1699 (New York, 1917), 167 (“had a slave”), 181–182, 190. Kellogg uses the word “tribe” instead of “nation” in her translation, but I have retained “nation” from the French original (“d’une nation fort esloignée du Sud-Ouest”). See “Récit de ce qui s’est passé de plus remarquable dans le

I Make Him My Dog / My Slave

1119

Slave labor was obviously not central to the economic or social organization of the Pays d’en Haut. Once slaves had been dominated and domesticated, they had fulfilled their most important purpose and thereafter filled roles more similar to the rest of the village. Policing the boundaries of slave status was thus less important than it would be in Atlantic societies that demanded slaves primarily for their productive and perpetual labor. Despite the persistent stigma attached to slaves and their constant vulnerability, the journey from degradation to integration was built into a system designed to produce the most fully assimilated and domesticated slaves possible. This goal ensured that indigenous slaves generally did not pass their status to the next generation. Not only was social rank rarely inheritable in most cultures of the Pays d’en Haut, but enslaving children did not meet the cultural demands that made slavery desirable. Neither the honor brought to warriors through the subordination of enemies nor the domestication of their power in the interest of the village was accomplished in this way. “Those who are allowed to live are like Slaves and valets among them,” Louis Hennepin remembered. “But in time they lose their Slave status and are as if they were from their [captors’] Nation.” Hennepin clearly overstated the universality of slaves’ assimilation and honor, as his and many others’ experiences demonstrate, but other French observers generally agreed. As Lahontan wrote in the late-seventeenth century: Upon the Death of a Savage his Slaves marry the other Women Slaves, and live by themselves in a distinct Hut, as being then free, or such as have no Master to serve. The Children that spring from this sort of Marriages, are adopted and reputed the Children of the Nation, by reason of their being born in the Village and in the Country. There’s no reason, say they, that such Children should bear the Misfortunes of their Parents, or come into the World in Slavery, since they contributed nothing towards their Creation.82 If Lahontan’s tone suggests that his comment was as much a critique of hereditary status in Europe as a description of Indians’ logic of ownership, his observation accurately explained one of the central differences between indigenous voyage de MM. Dolleir et Galinée (1669–1670),” in Pierre Margry, comp., Découvertes et établissements dans l’ouest et dans le sud … (Paris, 1876–1886), I, 112. 82  “Ceux à qui on donne la vie, sont parmy eux comme des Esclaves et des valets, mais à la longueur du temps ils perdent leurs Esclavages, et sont comme s’ils estoient de leur Nation.” Hennepin, Description de la Louisiane, 68; Lahontan, New Voyages, II, 53 (block quote). See also William Christie MacLeod, “Debtor and Chattel Slavery in Aboriginal North America,” American Anthropologist, XXVII (1925), 370–380.

1120

Rushforth

slavery and the labor-based chattel slavery developing simultaneously in the Atlantic world. Rather than a closed slave system designed to move slaves “up and out”—excluding slaves and their descendants from full participation in their masters’ society, even when freed—indigenous slavery moved captives “up and in” toward full, if forced, assimilation. Slavery therefore maximized the productive and reproductive capacities of the capturing village as it robbed enemies of the same sources of strength.83 Emphasizing this aspect of captivity to the exclusion of all others, however, both flattens and romanticizes the experience of indigenous slaves. Not only does it understate the persistent stigma and vulnerability associated with being a defeated outsider, but it also ignores the very real possibility that any captive could be traded to a third party and reenslaved in another unfamiliar environment. In a culture so conscious of kinship, the natal alienation of enslavement was its greatest burden, and the stain of slaves’ identity never quite left them. Indeed, even years after a slave was “adopted and married,” wrote Alexander Henry of the Ojibwas, “I never knew her to lose the degrading appellation of wa’kan’, a slave.” Indicative of slaves’ persistent marginalization, Algonquianspeakers often drew upon slaves’ degradation in their insults. According to Louis André, when “Missipissi,” a powerful manitou who controlled the waters, failed to deliver sturgeon, they said that he “was worth nothing; that he was a slave.” When Claude Allouez searched for a term to describe the devil to the Ottawas, he chose to call him a “vile slave” to show that he was powerless before God and “has no control over men’s lives.” The Jesuit’s understanding of awakaan as an odious term was confirmed by a group of Anishinaabes of Green Bay, who feared that branding the devil a slave risked reprisals from the insulted manitou: “Thou hast no sense,” they chided the Jesuits; “thou angerest the Devil too much.” Jacques Gravier modified the metaphor, making Satan a master and threatening his Illinois listeners that they “will be the slaves of the devil, if they die without baptism.” Gravier’s description stuck, and a generation later Illinois Christians still called nonbelievers taregᴕa matchimanetᴕare, slaves of the evil spirit, slaves of a slave.84

83  For “up and in” versus “up and out” slave systems, see James L. Watson, “Slavery as an Institution, Open and Closed Systems,” in Watson, ed., Asian and African Systems of Slavery (Berkeley, Calif., 1980), 10. 84  Alexander Henry, Travels and Journals, 307 (“degrading appellation”); Jesuit Relations, LVII, 289 (“worth nothing”), 281 (“vile slave” and “no control”), 269 (“no sense,” “angerest”), 279, 283, 289, LXIV, 191 (“slaves to the devil”), 219. For eighteenth-century uses of Gravier’s metaphor, see Le Boullenger, “Dictionnaire français-illinois,” s.v. “esclave.”

I Make Him My Dog / My Slave

1121

Simply calling someone a slave in any of the Algonquian or Siouan languages was considered a great offense, a universal “term of contempt,” according to Gravier. The Illinois insult matchirinisse was a vocative diminutive meaning, roughly, “[you] pathetic little bad man,” said to and about slaves. To avoid any ambiguity, Gravier wrote beside another form of the term, “It is an insult.” One could abuse an Illinois man by calling him a woman: kimitemᴕsserimere, “I consider you a woman. You are a coward.” For a much more serious insult, an antagonist could call him kikiᴕnaᴕarakiagana, or “slave woman’s vile / cheap vagina.” Although difficult to translate, Jesuits rendered the term in Latin (to protect the innocent) as “vile membrum servae, viris sensu metaphor,” literally “slave woman’s worthless penis, said metaphorically about men.”85 This cross-gendered and highly sexualized insult underscores the low status of slaves in the villages of the Pays d’en Haut. It also evokes the centrality of enslaved women’s sexuality to their identity and purpose as slaves. The value of enslaved women as kin-makers for capturing households, while affording opportunities not available in closed slave systems, also made their bodies, and especially their reproductive capacities, a primary object of enslavement. The question of slaves’ assimilation was therefore not a simple matter of softer slavery allowing assimilation and harsher slavery excluding outsiders permanently. Structuring slavery around forced assimilation gave a hard edge to the experience of enslaved women, whose loyalty and sexuality were policed with special vigor and violence.86 Because they had no actual kin within the village, slaves were extremely vulnerable to insult and injury. The Algonquian term kichecamᴕtehaganecᴕe meant “orphaned woman or girl without kin” and was described by a French 85   “Gueux, esclave. cest une injure. it[em] mechant homme.” Gravier and Largillier, “Dictionnaire illinois-français,” 259. The author thanks Michael McCafferty for pointing out the meaning of the vocative diminutive in this case. “Je te regarde comme une femme. tu es un lache”: Gravier and Largillier, “Dictionnaire illinois-français,” 199 (“kikiᴕnaᴕarakiagana”), 298. 86  There is a vast literature on the gendered implications of slavery, but, for an introduction among Native American women, see James F. Brooks, “‘This Evil Extends Especially … to the Feminine Sex’: Negotiating Captivity in the New Mexico Borderlands,” Feminist Studies, XXII (1996), 279–309; Brooks, Captives and Cousins, esp. 1–40. For gendered slavery in other contexts, see Gwyn Campbell, Suzanne Miers, and Joseph C. Miller, eds., Women and Slavery, I, Africa, the Indian Ocean World, and the Medieval North Atlantic, and II, The Modern Atlantic (Athens, Ohio, 2007); Jennifer Morgan, Laboring Women: Reproduction and Gender in New World Slavery (Philadelphia, 2004); Kathleen M. Brown, Good Wives, Nasty Wenches, and Anxious Patriarchs: Gender, Race, and Power in Colonial Virginia (Chapel Hill, N.C., 1996).

1122

Rushforth

writer as a “great insult.” Were enslaved women to offend their master, their husband, or another of their husband’s wives, they would have no extended kinship networks to rely upon for protection. One common word for a severe thrashing—sᴕpikipacaminta—meant “badly beaten like a slave.” Serious mistreatment or abuse was described by the phrase kikiᴕnanga nintepinaricᴕo: “He treats me like his slave.”87 This familial isolation may, in part, explain the gendered violence reported by French men in the villages of the Pays d’en Haut. Because the Illinois, like most Algonquian and Siouan peoples, spared the lives of captured women more often than of men, there was a surplus of females living in their households. In a polygynous society, this posed little problem, providing additional wives to prestigious men who could benefit from both their production and their reproduction. But French observers recoiled at the way some husbands treated their wives. According to Jacques Marquette, “They have several wives, of whom they are Extremely jealous; they watch them very closely, and Cut off Their noses or ears when they misbehave. I saw several women who bore the marks of their misconduct.” As mutilation (including nose cropping) carried specific symbolic meaning for slaves, and as an enslaved woman would have no father, uncles, or brothers to protect her from her husband’s wrath, enslavement may help to explain how such violent acts could be tolerated by the Illinois. Rather than a pervasive feature of all Illinois gender relations, then, this violence may be what they called tchekikicᴕ kikiᴕnaᴕiki epinatonghi: “that which one does to a slave.”88 An enslaved wife’s infidelity was the ultimate rejection of her master’s authority and the most basic threat to his control over her reproductive labor. Infidelity also challenged the authority of senior wives who could find their coercive power over an enslaved sister wife challenged by her lover and his kin. It was punished brutally. Pierre Deliette described in gruesome detail the ritual gang rape of adulterous Illinois wives, who the evidence suggests were enslaved outsiders rather than Illinois women. Surrounded by her husband’s 87  Gravier and Largillier, “Dictionnaire illinois-français,” 188 (“fille orpheline sans parents. grosse injure”), 533 (“badly beaten”). “Il me traite comme son esclave”: Le Boullenger, “Dictionnaire français-illinois,” s.v. “esclave.” 88   Jesuit Relations, LIX, 127 (“cut off their noses”). For nose clipping as a mark of captivity, see James R. Duncan and Carol Diaz-Granados, “Of Masks and Myths,” Midcontinental Journal of Archaeology, XXV (2000), 1–26. For the subordination of captive wives, see Judith A. Habicht-Mauche, “Captive Wives? The Role and Status of Nonlocal Women on the Protohistoric Southern High Plains,” in Cameron, ed., Invisible Citizens, 181–204. “Ce quon fait a ce qui est esclave”: Le Boullenger, “Dictionnaire français-illinois,” s.v. “esclave.”

I Make Him My Dog / My Slave

1123

male relatives, the adulteress was told that, because she was so hungry for men’s flesh, she would get her fill of them now. They then raped her in turns, berating her for the act of disloyalty. Reminiscent of the language comparing slavery to consuming human flesh and mirroring the tortures enacted on slaving warriors who were forced to eat their own flesh, the ritual language suggests the victim’s status as a slave. It also seems unlikely that this level of violence would be tolerated against an Illinois woman, whose kin could have come to her aid. Following a double standard, adultery brought lesser punishments to Indian men, but revenge occasionally drew slaves into the dynamic. Slaves could be given as a gift to mend internal disputes or violence within a village. When a man was killed by a jealous husband for seducing one of his wives, the murder could be covered by a generous gift, including a gift of slaves. Although an unlikely scenario, these customs would mean that, if an enslaved woman and her master/husband were both unfaithful, the slave could be raped for her punishment and then forced into another man’s bed as a gift to cover her master’s/ husband’s infidelity89 Such violence was rare because most of the enslaved understood the benefits of assimilation and sought to integrate themselves fully into their new households. Having seen a similar process in their own villages as they came of age and probably having participated in ritual violence as captives entered their home village, indigenous captives understood how successful slaves behaved. For those willing to adapt, there were relatively few constraints on moving up and into the captors’ society. Those who resisted assimilation, if they survived at all, found themselves vulnerable to alienation, violence, and trade. The French did not find in the Pays d’en Haut a benign system of captivity that they would transform into slavery. The Siouan and Algonquian peoples there had an elaborate and often brutal war culture centered on a form of slavery, built on different assumptions and employed for different reasons than the plantation slavery developing in the contemporary Atlantic. Focused on the act of enslavement rather than the production of commodities, indigenous slavery was at its heart a system of symbolic dominion, appropriating the power and productivity of enemies and facilitating the creation of friendships built on shared animosity toward the captive’s people. The intensely personal violence experienced during the first few months could be brutal and often deadly, but those who survived and weathered the storm of insults that followed found themselves in a system with many well-worn pathways out of slavery. These paths were often difficult to take, they were not available to everyone, and the prevalence of diplomatic slave trading served as a constant reminder of slaves’ 89  Deliette, “Memoir,” 334–337.

1124

Rushforth

marginal and precarious position. And it goes without saying that, given the choice, no one would have sought enslavement, no matter what the outcome. But, if only in the next generation, slaves could at least hope to rise to full acceptance by the society that enslaved them, becoming identified with a people they once considered less than fully human. Slavery in the Pays d’en Haut simultaneously disrupted and facilitated the broader political economy of trade and intermarriage that linked peoples and places throughout the region and beyond. Among allies, the sharing of enslaved enemies cemented alliances and created the bonds of fictive kinship that linked the region’s peoples to one another across ethnic and linguistic lines. Among enemies, enslavement provided outlets for violent expressions of enmity that stopped short of total destruction and provided mechanisms of repopulation and enhanced productivity. As in all historical contexts, enslaved individuals in the Pays d’en Haut were agents as well as objects, responding to the trauma and alienation of slavery with creative adaptation to their new surroundings. French colonizers would bring their own evolving notions of slavery to the colonial encounter, which began a century-long conversation that would transform the vocabularies and structures of slavery in both the Pays d’en Haut and the French Atlantic world.

The Process of Enslavement and the Slave Trade John K. Thornton

Warfare and Slavery

We have established so far that Africans were not under any direct commercial or economic pressure to deal in slaves. Furthermore, we have seen not only that Africans accepted the institution of slavery in their own societies, but that the special place of slaves as private productive property made slavery widespread. At the beginning, at least, Europeans were only tapping existing slave markets. Nevertheless, one need not accept that these factors alone can explain the slave trade. There are scholars who contend that although Europeans did not invade the continent and take slaves themselves, they did nevertheless promote the slave trade through indirect military pressure created by European control of important military technology, such as horses and guns. In this scenario—the “gun–slave cycle” or “horse–slave cycle”—Africans were compelled to trade in slaves, because without this commerce they could not obtain the necessary military technology (guns or horses) to defend themselves from any enemy. Furthermore, possession of the technology made them more capable of obtaining slaves, because successful war guaranteed large supplies of slaves.1 Hence, through the operation of their control over the “means of destruction,” to use Jack Goody’s descriptive term,2 Europeans were able to influence Africans indirectly. They could direct commerce in ways that helped them and also compel Africans to wage wars that might otherwise not have been waged. This would cause Africans to seek more slaves than they needed for their own political and economic ends and depopulate the country against their wishes. The quantitative increase would exceed Africans’ own judgment of a proper level of exports. In the end, this not only might increase economic dependence Source: Thornton, John K., “The Process of Enslavement and the Slave Trade,” in John K. Thornton, Africa and Africans in the Making of the Atlantic World, 1400–1800, New York: Cambridge University Press, 1998, pp. 98–126. © Cambridge University Press, reproduced with permission. 1  For an extreme statement, but carefully argued, see J. E. Inikori, “Introduction,” in idem, Forced Migration, pp. 45–51, and his introduction to his article “The Import of Firearms into West Africa, 1750 to 1830: A Quantitative Analysis,” ibid., pp. 126–53 (as well as the article itself). For a more cautions analysis that tends to the same conclusions, see Lovejoy, Transformations in Slavery, pp. 66–8, 78–87, 103–7. 2  Goody, Tradition, Technology and the State.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_036

1126

Thornton

but could result in large-scale destruction of goods, tools, and ultimately development potential. Hence, in the end, Africans would be helpless, exploited junior partners in a commerce directed by Europe. However, this argument will ultimately not be any more sustainable than the earlier commercial and economic ones. Certainly in the period before 1680, European technology was not essential for warfare, even if Africans did accept some of it. Likewise, it is much easier to assert than to demonstrate that Africans went to war against their will or solely to service the slave trade. Indeed the more we know about African warfare and resulting enslavement, the less clear and direct the connections between war and the export slave trade become. The contemporary evidence strongly supports the idea that there was a direct connection between wars and slavery, both for domestic work and for export. This did not mean that there was no nonmilitary enslavement, of course. Judicial enslavement was one common way of obtaining slaves, and judges, moreover, were not above distorting the law to provide more captives or enslaving distant relatives of guilty parties. Jesuit observers believed that this was common in Ndongo as early as 1600,3 and missionary travelers often commented on it in the seventeenth-century Upper Guinea region.4 But however scandalous this may have been, it is unlikely that judicial enslavement accounted for more than a few percent of the total exports from Africa. Thus the fact that military enslavement was by far the most significant method is important, for it means that rulers were not, for the most part, selling their own subjects but people whom they, at least, regarded as aliens. The fact that many exported slaves were recent captives means that they were drawn from those captured in the course of warfare who had not yet been given an alternative employment within Africa. In these cases, rulers were deciding to forgo the potential future use of these slaves. Some of the exports were slaves whom local masters wished to dispose of for one reason or another and those who had been captured locally by brigands or judicially enslaved. This is exactly the situation described by da Mosto in his account of Jolof in 1455. After a description of the use of the slaves in the domestic economy, da Mosto noted that most slaves were captured in wars with neighboring countries and the civil wars. Many of these captives were integrated into the domestic economy, but the rest were sold to the “Moors” for horses (i.e., they entered the Saharan trade), although “Christians” had recently entered the trade on the

3  Pierre du Jarric, Histoire des choses les plus mémorables…, 3 vols. (Bordeaux, 1608–14), 2:80–1. 4  See the documents cited and analyzed in Rodney, Upper Guinea Coast, pp. 106–9.

The Process Of Enslavement And The Slave Trade

1127

coast.5 This account focuses on two aspects of African societies that predisposed them to participate in the slave trade. The first is the regular use of slaves in the domestic economy and particularly as revenue for centralizing states, and the second is the role of warfare. The causes and motivations behind these wars are crucial for understanding the slave trade. Philip Curtin has examined the Senegambian slave trade of the eighteenth century and has proposed a schema for viewing African warfare that resulted in slave captures that can be fruitfully applied to the earlier period as well. He proposes that wars be classified as tending toward either an economic or a political model. In the economic model the wars were fought for the express purpose of acquiring slaves and perhaps to meet demands from European merchants; in the political model wars were fought for mostly political reasons, and slaves were simply a by-product that might yield a profit. Both models are seen as “ideal types,” and individual wars might contain a mixture of motives, of course. On the whole, however, Curtin believes that the eighteenth-century Senegambian data support a political, rather than the economic, model.6 Actually, discerning between an economic and a political model is not easy in practice. Consider the case of Portuguese Angola, a state seemingly founded on the premise of exporting slaves. Angola’s wars ought to fit the economic model if any state’s would. Yet many of Angola’s wars, and the majority of the most lucrative ones in terms of acquiring slaves, had more or less clear-cut political motives. Portugal’s early wars in Angola, for example, were as much for establishing a foothold in the area as for capturing slaves. In 1579, after all, the Portuguese were nearly driven out by the forces of Ndongo, and the wars between then and about 1595 were defensive as much as offensive.7 When the Portuguese at last went on the offensive against Ndongo, their series of wars in the early seventeenth century also resulted in gains in territory, and after 1624 they became embroiled in a long series of wars that might be called the War of the Ndongo Succession, in which Portuguese officials hoped to place a pliant king on Ndongo’s throne and met with the resistance of Queen Njinga (1624–63).8 This war, almost continuous from 1624 until 1655, 5  Da Mosto, “Mondo Novo” (ed. Gasparrini-Leporace), p. 42. 6  Curtin, Economic Change, pp. 153–68. 7  See the summary in Pero Rodrigues, “Relação,” 1594, MMA 4:565–77, and other documentation in MMA 3, passim. 8  Heintze, “Ende,” pp. 224–66, on the first ten years of the War of the Ndongo Succession. Later stages are chronicled in Cadornega, História, vols. 1 and 2. A concise summary can be found in Birmingham, Trade and Conflict in Angola.

1128

Thornton

can account for most of the activities of the Portuguese army in the period, and the expansion of the war to the east can account for the appearance of many eastern Angolan slaves in the New World during this rime.9 Of course, this does not mean that all of Angola’s wars fit the political model, but only that we should keep in mind that in Africa, as elsewhere, wars could always be multivalent, even defensive or strategic ones. It was precisely in this context that Cadornega, the chronicler of the seventeenth-century Angolan wars, rejected charges made in Portugal that the Angolan wars were simply “guerras de negros” intended solely to capture slaves.10 The Jolof and Angolan models we have just examined suggest that the solution to the problem of the nature of African wars will not be easy to assess. In theory, wars with objectives that Europeans might see as political, such as annexing territory (or defending territory) or acquiring and strengthening political rights, should be classed as political, whereas raids conducted solely to acquire loot or trade goods should be classed as economic. In practice, however, as we have already seen from our previous examination of Jolof politics, capturing slaves made political sense to Jolof rulers as well as economic sense. Slaves were sources of wealth, and even in a hit-and-run operation that did not envision political conquest (although Jolof certainly did make conquests as well), the slaves could be made to produce wealth in the same way that a conquest would. Likewise, Jolof’s rulers might employ the slaves to generate private revenue for them or to act as personal servants, soldiers, and administrators and thus raise them up against their rivals for power within Jolof. Hence, even a war that simply resembles a raid with no political objectives would have major political consequences if slaves were taken—and the fact that some were sold to outside parties should not lead us to the conclusion that the war had no political motives. These considerations thus make it difficult to evaluate testimony about the motivations of warfare by European observers, who thought of wars that netted simply booty, even if this meant slaves, as being economic ventures (i.e., ventures aimed at securing slaves for export). Surely some of the ventures conducted by African rulers seem to have had no other purpose. Edward Fenton believed that his request for slaves in 1580 in Sierra Leone led the ruler to conduct a war merely to fill his requests—a war he thought would net 3,000–4,000

9  Bowser, African Slave, pp. 41–3. See also the amazing variety of central African ethnonyms in Venezuelan “Empadrono” documents in Ermila Trochis de Veracoechea, Documentos para el estudio de los esclavos negros en Venezuela (Caracas, 1969), pp. 194–200. 10  Hilton, Kingdom of Kongo, pp. 104–6.

The Process Of Enslavement And The Slave Trade

1129

slaves.11 De Almada, likewise, recalled that once the ruler of Kayor waged a war in 1576 solely to obtain slaves to pay a debt he owed the Cape Verdian merchant.12 Similarly, several contemporary observers believed that many of the wars waged by the Angolan governor Mendes de Vasconcellos in the early seventeenth century had the acquisition of slaves as their only motive.13 Such wars, however, may well have been waged solely in order to acquire slaves even without the demands of Atlantic traders. One example of this comes from Fernandes’s late fifteenth-century informants. According to them, the Sanhaja of the desert made war against the people south of them “more for pillage than for power.”14 Likewise, Jannequin de Rochefort argued on the experience of his observations in 1639 that the wars were not for conquest but to raid for people and cattle.15 This issue goes to the heart of the unusual nature of African politics and one of the matters that makes it different from Eurasian politics. Just as slavery took the place of landed property in Africa, so slave raids were equivalent to wars of conquest. For this reason, one must apply a different logic to African wars than the equations of political motives equals war of conquest and economic motives equals slave raid. This analysis changes our understanding of the objectives of war and must ultimately change our assessment of African warfare. Lovejoy, for example, has proposed that warfare was endemic in Africa as a result of political fragmentation. In other words, the very fact that Africa had few large-scale political units meant that wars would be more frequent, and thus enslavement increased. As fragmentation increased (a situation that he believes took place during the period of the slave trade), war naturally increased. Underlying this is the assumption that a political situation of small states would naturally lead to a movement to consolidate them into larger, Eurasian-style polities. Thus, although African politics actually determined the course of warfare, the intrinsic structure of those politics created more wars.16 Furthermore, one need not consider most wars as being explained by the economic model but by the political model, in which wars were an attempt to

11  Fenton, Troublesome Voyage, p. 109. 12  De Almada, “Tratado breve,” MMA2 3:243. 13  Heintze, “Ende,” pp. 231–9; see also documentation in MMA 6, passim, which regularly refers to “guerras de negros.” 14  Fernandes, “Descriçā”, fol. 69v. 15  Jannequin de Rochefort, Voyage, pp. 86–7. 16  Lovejoy, Transformations in Slavery, pp. 66–87.

1130

Thornton

remedy the fragmentation by consolidating power. The failure to consolidate was thus the fuel that fired the slave trade. Lovejoy’s solution would be more helpful if it were true that there is a correlation between political centralization and peace, but unfortunately this does not seem to have been the case. This emerges clearly from an examination of the policies of the empire of Songhay, which controlled a huge area in the sixteenth century—larger than any other state at the time and on a scale that rivals most European states. Songhay was an expanding empire and thus waged wars of territorial conquest, capturing slaves along the way. According to the Tarikhs, local chronicle sources, for example, Sonni Ali, who ruled from 1468 to 1492 (about the same time as the Atlantic trade developed), conducted some sort of war or campaign in every year of his reign.17 The campaigns varied in length, size, and complexity, and not all were conducted by the king himself. The chroniclers who described these wars commented often on the motivations of their ruler (which were usually to extend territory, punish insults, retaliate for attacks on his territory, and the like) but never specifically mentioned the capture of slaves as one of the goals, nor did they take the trouble to enumerate or boast of the slaves captured, clearly implying that the Songhay expansion was politically rather than economically motivated.18 The exploits of his successor, Askia Muhammed, are less clear, because a different chronicler discussed them, but he too conducted many large expeditions, although perhaps somewhat less frequently than his predecessor. As in the case of Sonni Ali, the acquisition of slaves was never mentioned or the number of captives discussed, and moreover, Askia Muhammed’s motivations were similar to those of his predecessor.19 But the warfare of an expanding Songhay was perhaps the exception in Africa. This is because most of Africa was, as Lovejoy argues, fragmented. However, this fragmentation was not simply the result of a failure of politics, nor did it increase appreciably during the time in question. Instead, it appears 17  Adam Koraré Ba, Sonni Ali Ber, pp. 65–75, worked out from the complicated and not unambiguous notes in the Tarikhs. 18  Cf. Tarikh al-Fettash, p. 91. 19  According to the notes in the Tarikh al-Fettash, pp. 133–45, he had annual campaigns in the periods 1498–1502 and 1505–6 and two campaigns between 1507–10 and 1511–14, after which the chronicle becomes vague and ceases an annual account of events. During the period 1492–7, Askia Muhammed was away on a pilgrimage to Mecca and the chronicle covers his life there, though campaigns probably were waged in the Sudan in his absence. Likewise, there were probably campaigns conducted by Askia Muhammed or his generals in the 1514–29 period not recorded in the Tarikh.

The Process Of Enslavement And The Slave Trade

1131

as a constant feature of African society, characteristic of the entire precolonial period. As a measure of African fragmentation, consider the normal size of African states, based on the boundaries of these states as drawn in Maps 1–3. There was no African state as large as the larger Asian or Euro-American empires of the period. Late imperial China occupied between 3 and 4 million square kilometers of land, the Ottoman Empire some 4 million square kilometers at its height, and the Russian Empire in the seventeenth century covered some 2.5 million square kilometers and was expanding rapidly. Mogul India controlled as much as 2 million square kilometers of land. In its American possessions, Spain had effective control over as much as 4.5 million square kilometers. In Africa, by contrast, the largest states (what historians call empires, like Songhay) controlled areas in the range of 500,000–1 million square kilometers. This was on the scale of Safavid Iran (1.5 million square kilometers) and the larger European states, such as France (with 550,000 square kilometers) and Spain (with its Italian and Burgundian but not Austrian or American possessions—at just under 1 million square kilometers). But states on the scale of Songhay occupied only one part of Africa, principally the rich river valleys of the western and central Sudan, and outside that area there were no states as large (the Lunda Empire, the next-largest state outside the western Sudan region, controlled some 300,000 square kilometers at its height in the nineteenth century, but it was only about half that size in 1680). Medium-sized states, that is, states with a surface area in the range of 50,000–150,000 square kilometers, or the size of European states like England (150,000 square kilometers) and Portugal (90,000 square kilometers) or the larger of the Italian city-states and German principalities (in the 50,000– square-kilometer range), were found in more areas. By 1680, the Oyo Empire (in Nigeria) may have exceeded 150,000 square kilometers, though not by much. The states of the lower Niger valley, such as Nupe, Igala, and Benin, would have reached close to the middle of this range, and the Hausa states of northern Nigeria (Kano and Katsina at about 60,000 square kilometers, the rest smaller) were in the smaller range. In central Africa, Kongo was by far the largest of the states, with as much as 130,000 square kilometers; other states, like Ndongo, were somewhat smaller but still in this range. In all, only perhaps 30 percent of Atlantic Africa’s area was occupied by states with surface areas larger than 50,000 square kilometers, and at least half of that area was occupied by states in the medium-sized (50,000–150,000 square kilometers) range. The rest of Atlantic Africa was occupied by small, even tiny, states. Of this group, a few states in the southeastern part of modern Ghana and Benin-Togo—including Allada, the core of the later kingdom of

1132

Thornton

Dahomey, and the larger Akan states (Akim, Denkyira, Akwamu, and the core of the later Asante kingdom)—each controlled perhaps 5,000 square kilometers in the late seventeenth century.20 But they too occupied a relatively small part of the total, and certainly more than half the area of Atlantic Africa was ruled by ministates whose surface area ranged from 500 to 1,000 square kilometers. If this were not dramatic enough, one should consider that if these statistics were broken down by population, a portion considerably greater than half of all of the people in Atlantic Africa lived in the ministates, because these states were found in the most densely populated parts of the region. Thus, one can say with confidence that political fragmentation was the norm in Atlantic Africa. By this account, the “typical” Atlantic African probably lived in a state that had absolute sovereignty but controlled a territory not exceeding 1,500 square kilometers (smaller than many American counties, perhaps the area occupied by a larger city). Populations might vary considerably; in the sparsely inhabited areas of central Africa, such a state might have 3,000–5,000 inhabitants, but on the densely inhabited Slave and Gold coasts it could control as many as 20,000–30,000 people. Virtually all the land from the Gambia River along the coast to the Niger delta was in states of this size, and much of the land stretching into the interior. In areas like Angola, ministates like these occupied the mountainous land between Kongo and Ndongo and the area of the Kwanza River between Ndongo and the larger states of the central highlands. In short, enlargement of scale does not seem to have been a priority for leaders. Historians, anxious to assert that Africans did build large states, have to some extent focused too much attention on the empires and the mediumsized states, and thus the point is often overlooked. But the reasons for Africa’s small states were probably not the result of some sort of backwardness that prevented them from seeing the advantages of larger units. One reason for the smallness of scale (not necessarily the only one) may derive from the legal system, which did not make land private property, and may also explain why the Americas, the other world area without landed property, was also the home of small and even tiny states (outside its own few dramatic empires). In Eurasia, control over large areas of land was essential, because it was through grants of land that one rewarded followers, and this land was normally worked by tenants of one kind or another. Eurasians were relatively less interested in controlling people, for without land, the people’s labor could 20  The Asante kingdom, which would eventually form a medium-sized state with perhaps 100,000 square kilometers, and Dahomey, which eventually controlled a somewhat smaller area, were only beginning to form in 1680.

The Process Of Enslavement And The Slave Trade

1133

not be assigned or its reward collected by landowners. African states were not concerned with land—for as long as there was no population pressure on the land, more people could always be accommodated. Hence, African wars that aimed at acquiring slaves were in fact the exact equivalent of Eurasian wars aimed at acquiring land. The state and its citizens could increase their wealth by acquiring slaves and did not need to acquire land, unless they were short of land at home (which was not the case, as far as we can tell). The acquisition of slaves instead of land in wars had other advantages. Whereas conquest of land necessarily required administration of larger areas and expansion of military resources, the acquisition of slaves only required a short campaign that need not create any new administrative conditions. Moreover, conquest of land and its subsequent government usually required sharing the proceeds of land with existing landlords, state officials, and other wealthy members of the defeated state, who might be defeated but usually still had to be co-opted. Slaves, on the other hand, were unable to bargain as wealthy landlords might have and could be integrated individually or in small groups into existing structures. We can see these processes in operation in the case of Sierra Leone. The Sapes, as the early Sierra Leone inhabitants were called in documents of the time, were not creating empires or even larger states. They seem to have exported many slaves, however, for although the Sapes did not apparently enter into the trade immediately, by 1500 they accounted for a large proportion of the slaves imported into Europe. If the frequency of the ethnonym in bills of sale is any guide to relative volume, early sixteenth-century port records have them as the third most common group, behind Jolofs (which probably also includes exports from Songhay and some from Mali) and Mandinkas (Mali exports).21 The Sape slaves, according to Fernandes’s informant Alvaro Velho, were the result of “constant wars” of the region.22 These wars do not appear to have been waged for territorial expansion; although we lack the chronicle sources of the Sudanese region to confirm this, 21  P. E. H. Hair, “Black African Slaves at Valencia, 1482–1516: An Onomastic Inquiry,” History in Africa 7 (1980): 132 (1498); Alfonso Franco Silva, Registro documental sobre la esclavitud sevillana, 1453–1513 (Seville, n.d. [1979]), no page number: entry for 1502. 22  Fernandes, “Descriçā,” fols. 125–125V; see also the legend of the Cantino Atlas (1502), in Cortesão and Teixeira da Mota, Monumenta 1:10, pl. 5. Note that this evidence tends to discount the idea advanced by Rodney (Upper Guinea Coast, pp. 60–70) and based on early seventeenth-century Jesuit “complaint” literature that the Sapes were peaceful until the Mane invasion of ca. 1545, whose ambitions for expansion increased warfare and slave raiding.

1134

Thornton

certainly there was no consolidation in Sierra Leone as a result of warfare.23 But as Velho also testified, slaves were used in the domestic economy to increase the ruler’s personal income,24 and perhaps this in itself can explain the propensity for wars that did not increase wealth by the annexation of territory but by the annexation and transport of people. This feature can also explain the existence, already in Velho’s time (1499), of small raids being conducted in the “Rivers” region, composed, according to Fernandes, of Falup raiding parties in canoes penetrating all the rivers of the region.25 In the sixteenth century the Falup were joined by the Bissagos Islanders, who were soon renowned for their naval attacks on the mainland.26 This type of war was very common in that region in the sixteenth and seventeenth centuries, where a host of visitors describe canoe-based parties that would move silently and strike suddenly (sometimes at night) and carry off people.27 In these instances, the slaves could well have been used either by the rulers of the small states of the area to increase their personal dependents and thus strengthen their power base or by private citizens, merchants, or aristocrats to increase their wealth or to increase their power vis à vis the rulers. Although some of these raids may also have been undertaken to supply European demand, this demand was in addition to the greater African demand for slaves to be used domestically as well as for export. Many Africans retained females from the raids and sold off males, because the Atlantic trade often demanded more males than females. The Bissagos Islanders held many female slaves, and observers believed that virtually all the productive labor was done by women.28 Lemos Coelho, a Cape Verdian merchant, believed that many societies held large reserves of slaves who could be sold but who would work for their owners in the meantime.29 Naturally enough, the Portuguese in Angola fell into the same pattern, retaining many 23  The area may have undergone some consolidation in the late sixteenth century as a result of the Mane invasion, which is mentioned in numerous contemporary accounts. See the dated account of Rodney, Upper Guinea Coast, pp. 39–70, and a partial, but better, account of the resulting kingdom in Jones, “Kquoja Kingdom.” 24  Fernandes, “Descriçā,” fol. 129. 25  Ibid., fol. 117V. 26  De Almada, “Tratado breve,” MMA2 3:315–20. 27  Ibid., pp. 288, 315–20; Francisco de Lemos Coelho, Duas descriçōes seiscentistas da Guiné de Francisco de Lemos Coelho, edited by Damião Peres (Lisbon, 1935), pp. 42–5. 28  Thornton, “Sexual Demography,” pp. 39–48. 29  Pierre Culrru, ed., Premier voyage de Sieur de la Courbe, fait à la coste d’Afrique en 1685 (Paris, 1913), p. 252.

The Process Of Enslavement And The Slave Trade

1135

slaves on their plantations along the major rivers, and still selling off many, especially males.30 Increasing wealth through warfare and enslavement was of course a cheap way of increasing power. Slaves could be captured in wars and in raids and carried back to the home territory by the victors and put to work, without the attacking armies having to conquer and occupy territory. For small states with small armies, this was a logical way to become richer. But of course, in the medium-sized states and empires, territorial expansion also took place. We have already discussed the wars of expansion in Songhay and Jolof and noted that in addition to increasing territory, their wars also resulted in capturing slaves. For the expanding empire, enslavement of the conquered population allowed the rulers of the expanding state to increase their personal wealth and also to build armies and administrative corps of direct dependents, just as the revenues from the conquered territories provided continuous new income. Thus, external expansion could also increase wealth, and the slaves that were a byproduct of the wars of expansion could increase centralization at home. All these factors resulted in an enormous slave population in Africa at the time of the arrival of the first Europeans and during the whole era of the slave trade. They meant that the necessary legal institutions and material resources were available to support a large slave market, one that anyone could participate in, including Europeans and other foreigners. Those who held slaves and did not intend to use them immediately could also sell them, and indeed, this is why the number of African merchants who dealt in slaves was large. Central African data corroborate this process very well. Although there are few useful records for the pre-1483 period, it is clear that Kongo was expanding territorially during the initial period of Portuguese contact, because it was regularly cited as fighting frequent wars.31 However, we have also already seen that one of the most important aspects of Kongo’s centralization was the development of a large urban center with numerous slaves, giving the ruler an advantage over other members of the coalition that began the kingdom. Political motives such as increasing territory, revenues, and a loyal power base played a role in Kongo as in Jolof. Thanks to Portuguese participation in some of Kongo’s wars, we can find out how slaves were used. As in the other areas, the ostensible motive for wars was quite strictly political. For example, Kongo made war against islands in 30  Thornton, “Slave Trade in Eighteenth Century Angola,” pp. 417–27. 31  Zorzi, “Informatiō,” fol. 131 (ca. 1515); de Enciso, Suma, p. 110 (1519); Saintogne, Voyages advantureux, fols. 54, 55 (1536); Saintogne, Cosmographie (ed. Musset), pp. 339–40; Letter of Jacome Dias, 1 August 1548, MMA 2:180.

1136

Thornton

the mouth of the Zaire River in 1491 to bring them back to obedience.32 The instructions of the king of Portugal to Gonçalo Rodrigues in 1509 strongly support the idea that slaves were captured in such wars. Rodrigues was given orders concerning how to sell whatever slaves the king of Kongo might choose to grant him for participation in the campaign.33 A still better description concerns a campaign conducted in 1513 or 1514 against Munza, an enemy south of Kongo. This war was apparently a defensive war, for Munza was said to have attacked Afonso’s son, the Mweni Mbamba, and the war was to relieve him and punish Munza. King Afonso and the Portuguese in his service sent at least 600 slaves back to the capital during the war (and when the army returned, they brought at least 190 more), of which 510 were diverted to the Atlantic trade. Of all these slaves at least 90 remained in Kongo, and Afonso complained that the Portuguese whom he had entrusted with disposal of the slaves in the war had done so improperly, leaving too few in Kongo and, moreover, among them only those who were “old and thin.”34 Afonso was clearly concerned with both domestic use and foreign exports and, at least in this case, believed that his interests were not served by the export of too many, but he was also clearly willing to allow a substantial number to be sold outside the country. Perhaps one of the reasons that the central African region was a rich source of slaves was that there were several states like Kongo for whom slaves were both a by-product of wars of expansion and useful in themselves for increasing centralization and loyalty. Beginning around 1520, Kongo ran into the growing power of Ndongo, which like Kongo was expanding and using slaves to support centralization. That such wars as developed in the mid-1520S worked in this way is suggested by several letters of complaint written by Afonso in 1526; in one of his bitterest he deplores the (unofficial) help that the Portuguese gave the ruler of Ndongo,35 which resulted in the capture and sale of Afonso’s subjects, even the nobility.36 A similar struggle seems to have been waged earlier against the Nziko kingdom, which continued exporting slaves itself, becoming a major exporter by the 1530s.37 Even the countries that were not leading the 32  Rui de Pina, untitled MS, 1492, fols, 97vb–98va. 33  Instructions for Gonçalo Rodrigues, MMA 4:61. 34  Afonso to Manuel I, 5 October 1514, MMA 1:314–15. 35  Not named in the letter but presumably the “vassal” whom Afonso complains of; see John Thornton, “Early Kongo—Portuguese Relations,” pp. 192–4; cf. Afonso to João III, 6 July 1526, MMA 1:470. 36  Pacheco Pereira, Esmeraldo, bk. 3, chap. 2 (ed. Silva Dias), p. 134. 37  João III to Afonso I (ca. 1529), MMA 1:526 (noting that the northeastern region was the main exporter of slaves).

The Process Of Enslavement And The Slave Trade

1137

expansion might export slaves as a result of this warfare; the inhabitants of the islands in the mouth of the Zaire River, called “Pamzelungos” in sixteenthcentury sources, also exported slaves, perhaps taken in unsuccessful attempts by Kongo to suppress their revolts.38 On the basis of the available evidence it is possible to make a very strong case for a simple political explanation of slaveproducing wars, even when these wars did not have expansion as a goal and in spite of the Portuguese involvement and the Africans’ own strong interests in exporting slaves.

Domestic Use versus the Export of Slaves

This interpretation of African politics has reemphasized the importance of domestic slavery in Africa. Obviously, slaves were sufficiently important that one could find many in African societies. Likewise, as we have suggested, exports tended to be drawn from those slaves who were recently captured and had not yet found a place in the society of their enslavers. This aspect of slavery obviously places emphasis on the African decisions concerning which slaves to sell to Europeans and when. These decisions were in turn a product of the specific situation in each country, including price and availability of slaves. In large measure, the decision to participate in the Atlantic trade required that specific conditions be met, and countries often entered and left the trade. The factors that caused an African region to desist from selling slaves to European buyers are dramatically shown in the case of Kongo and Benin. Both countries were early exporters of slaves, both for São Tomé and for the European trade, where a few people from both regions appear in port records.39 Yet, Ryder has shown that beginning perhaps as early as 1520 the state of Benin began to restrict the slave trade, finally cutting it off by perhaps 1550.40 In any case, bills of sale and inventories from the New World do not list Benin slaves in the late sixteenth and seventeenth cunturies.41 Visitors to Benin in the late sixteenth and seventeenth centuries speak of the area as producing cloth and pepper but do not speak of slave exports.42 Although it must be considered 38  Afonso to António Carneiro, 5 March 1516, MMA 1:359. On the location of the “Pamzelungos,” see François Bontinck, “Les ‘Pamzelungos’ ancêtres des Solongo,” Annales Aequatoria 1 (1980): 59–86. 39  Hair, “Black African Slaves,” p. 131; Franco, Sevillana. 40  Ryder, Benin and the Europeans, pp. 56–90. 41  Bowser, African Slave, pp. 40–43. 42  Ryder, Benin and the Europeans, pp. 82–5.

1138

Thornton

speculative, it is possible that the demands for labor in these industries conflicted with European demands, and local interests put pressure on the state (or the state itself decided) to restrict exports. In the case of Kongo, the export of slaves continued longer, for it was only in the late sixteenth and early seventeenth centuries that Kongo began to stop exporting slaves—though most early seventeenth-century sources speak of Kongo as a source for a few slaves.43 Kongo slaves do not disappear from inventories in part because, unlike Benin, it was still a shipping point for slaves (traded from the interior) and in part because Kongo was still being attacked by its neighbors, either the Portuguese (as in their major war of 1622),44 Matamba (after Njinga moved her capital there about 1635), or its northern neighbors. In both cases, the decision to stop exporting slaves may have come as a result of a change in political direction as well as from economic considerations. Both countries seem to have stopped expanding at about the time they stopped their exports, Benin in the 1550s,45 and Kongo at the end of the century.46 Less frequent wars may have meant fewer enslaved people and a subsequent increase (at least domestically) in prices for slaves. Both states had also, perhaps, met their domestic demand for labor and both states had developed suitable exports to pay for any imports they desired; these exports, moreover, were items that were labor-intensive and this may have driven slave prices up. In Benin, the goods for export were cloth and pepper, and in Kongo, cloth—the conquest of the cloth-producing eastern regions seems to have been instrumental in Kongo’s change.47 Neither state seems to have been impelled by pressure, either commercial or military, to continue trading in slaves. It was only in the late seventeenth and early eighteenth centuries, under the pressure of civil wars, that slave trading resumed in both places.48 We can apply a similar logic when we examine those regions that did not participate in the slave trade immediately. For example, the Gold Coast region exported few slaves during its first century and a half of contact with Europe, in spite of the substantial trade taking place elsewhere. Moreover, the Gold Coast was divided into many states, had many wars,49 and was characterized 43  Thornton, Kingdom of Kongo, p. xv. 44  Described in Mateus Cardoso, “Relação da morte,” fol. 175–75V (MMA 15). 45  Ryder, Benin and Europeans, pp. 15–18, 73–5. 46  Thornton, Kingdom of Kongo, p. xv. 47  See the observations made by Hilton, Kingdom of Kongo, pp. 104–19. 48  Ryder, Benin and the Europeans, pp. 167–9; Thornton, Kingdom of Kongo, pp. 95–6. 49  Wars are mentioned in 1502 (ANTT, Cartas Missivas, 2/180), 1510 (ibid., CC I/9/60, Factor and Officials of Mina to King, 2 September 1510), 1519 (Order of Fernão Correia to Factor,

The Process Of Enslavement And The Slave Trade

1139

by social and political structures similar to those of Sierra Leone, which did participate much more fully in the trade. The Gold Coast states surely used slaves in their domestic economies, and probably to support centralization as well, but they were perhaps less willing to export slaves. Indeed, the Gold Coast states imported slaves from their neighbors and from the Portuguese to meet their needs for labor and political dependents and paid for them in gold. Slightly different economic situations thus led to substantially different strategies for accomplishing the same things. Some similar sort of logic must also have applied in the northern part of the Gabon coast, north of Loango, where, as on the Gold Coast, there were small states and many wars. Here, too, the decision not to sell slaves in any numbers to Europeans was matched by a decision to import slaves in exchange for supply services and exports of cloth.50 It was perhaps the fact that cloth was exported that meant that the long stretch of coast between the Volta and Benin rivers, later to be called the Slave Coast, did not immediately export slaves, although two slaves called “Lucumies” (Yoruba) did appear in a 1547 inventory from Española; perhaps they were captives of the wars that disrupted Oyo and its neighbors in the first half of the sixteenth century.51 Only in the later sixteenth century, as we shall see, when a shadowy series of events led to the rise of states such as Allada and Popo (and perhaps also the Oyo Empire), did this region begin to export slaves. Thus, the evidence strongly suggests that it was the decisions of African states that determined participation in this particular branch of trade and not so much European pressure. These decisions were the product of processes that our sources can reveal only dimly to us—they probably concerned the relative price of slaves versus the prices of other commodities, competing demands for labor, or the relative price of European imports versus exports other than slaves. Of course, Europeans always had a good market for slaves 26 September 1519, MMA 1:427), 1520 (ANTT, CC II/91/92, Order of Pacheco Pereira to Factor of Mina, 12 November 1520), a lengthy one in 1548 (ibid., I/80/74, Gonçalo Francisco de Almeyda to King, 14 April 1548), and another protracted war that ended in 1557 (Afonso Gonçalves de Botafogo to Queen, 18 April 1557, fol. 3V, published with English translation in Teixeira da Mota and Hair, East of Mina). 50  See David Patterson, The Northern Gabon Coast to 1875 (London, 1975), pp. 1–22. 51  “Escriptura que enbio el liçençiado çerrato sobre la hazienda de gorjon,” 21 September 1547, in J. Marino Incháustegui Cabrai, ed., Reales cédulas y correspondencias de govemadores de Santo Domingo, 5 vols. (Madrid, 1958), 1:237–8. Both men were forty years old in 1547, strongly suggesting enslavement after 1525. On the wars in Oyo, see Robin Law, The Oyo Empire, c. 1600–c. 1836: A West African Imperialism in the Era of the Atlantic Slave Trade (Oxford, 1977), pp. 37–9.

1140

Thornton

and naturally they were a preferred commodity, but the Europeans would not abandon trade and relations with a country simply because it would not or could not sell slaves. As long as some exchange was possible, trade occurred. At the same time, however, they were perfectly willing to buy slaves whenever an African country decided to sell them, and they always hoped they could get more.

African Warfare and European Military Technology

Although I have shown that African wars led to enslavement on a large scale and that African politics can explain even slave raids that seem to have no political motive, the hypothesis that Europeans influenced African behavior through control over military resources must still be addressed. Given the significance of warfare for expansion of wealth in Africa, the military case must be carefully examined. Certainly, Europeans did participate, wherever possible, in African politics, often as “military experts” or advisors, occasionally as armed mercenaries. They did this both officially through government-sponsored assistance programs, such as the aid that Portugal gave to Kongo in 1491, 1509, 1512, and 1570, or unofficially and without authorization, as in the support for Ndongo in the 1520s, the help that gunners gave to the Mane in the 1550s, and perhaps the assistance to Benin in the 1510s and 1520s.52 Other foreigners of European origin also provided assistance—Hawkins’s help in Sierra Leone and Ulsheimer’s in Benin are two more sixteenth- and early seventeenth-century examples. Acceptance of this assistance might simply be seen as the desire of centralizers to make use of foreign, rather than local, officials and dependents as a means of keeping local political debts to a minimum and of creating a dependent bureaucracy. But it is also clear that Europeans provided new military techniques and technology as well, perhaps at the price of demanding more vigorous participation in the slave trade than their patrons wished. However, the kind of military assistance that Europeans in the sixteenth and seventeenth centuries could render in Africa was not as decisive as much of the writing on the “gun–slave” and “horse–slave” cycles implies. For example, Elbl has examined records of Portuguese horse imports into Senegambia and found that they can scarcely be considered numerous enough to be crucial to the military survival or even success of Jolof cavalries.53 Moreover, as Law has 52  See documentation cited in Chapter 3. 53  Elbl, “Portuguese Trade.”

The Process Of Enslavement And The Slave Trade

1141

pointed out in a detailed study of the horse problem, much of the Sudanese region was capable of breeding fairly large numbers of horses by the fourteenth century.54 Thus, the Atlantic trade coincided with a period when demands for horses were declining, and perhaps the trade must be seen in much the same light as the trade in other commodities—as supplementing or even complementing an existing trade and production. Firearms and other personal weapons (as in the gun–slave cycle) are even more problematic. European firearms and crossbows, the missile weapons that differed most from those in use in Africa, were designed to counteract armored cavalry or for naval warfare in Europe. Although they had great range and penetrating power (capabilities that developed out of a long-standing projectileversus-armor contest), they had a very slow rate of fire.55 For Africans, who generally eschewed armor, the advantages of range (penetrating power being relatively unimportant) were more than offset by the disadvantages of the slow rate of fire, except in special circumstances. Such circumstances were found in naval warfare, for example. As we have seen, in the initial encounters between da Mosto’s ships and Mandinga craft in the Gambia, the crossbowmen in the tops were able to fire to good effect against the attacking forces, protected as they were by the high sides of their ships. This particular feature of the ship may be one of the reasons Kongo favored Portuguese assistance in its wars with the Zaire islands, where African craft, Portuguese ships, and the long-range weapons of the Europeans could be used to good effect. Likewise, artillery would be useful for attacking fortified locations. European artillery was used in Sierra Leone in the 1500s56 and in Benin, probably in 1514, when the king of Benin seized a Portuguese bombard,57 but certainly in 1601, when Ulsheimer joined Dutch sailors who used a gun to blast down the gate of

54  Robin Law, “Horses, Firearms and Political Power in Pre-Colonial West Africa,” Past and Present 72 (1976): 112–32. Law also makes a case for a horse–slave cyde operating within the West African region between northern horse-breeding areas and southern horse-using areas, especially in regard to the famed cavalry of Oyo. Whatever validity there may be for this case, it is not applicable in any clear way to the Atlantic horse trade. 55  Some of this technology is examined in H. A. Gemery and J. S. Hogendorn, “Technological Change, Slavery and the Slave Trade,” in C. J. Dewey and A. G. Hopkins, eds., The Imperial Impact: Studies in the Economic History of Africa and India (London, 1978), pp. 246–52. They distinguish an earlier (pre-1650) period with its matchlock technology from a later period when the flintlock musket was in use. 56  De Almada, “Tratado breve,” MMA2 3:375. 57  Mina Officials to Contadores of Portugal, 29 April 1514, ANTT, CC II 46/165.

1142

Thornton

a rebel town.58 In these cases, as in those of the naval engagements, however, the new weapons were hardly of such overwhelming decisiveness that they tipped the scales of warfare strongly. European ships could be employed only as a supplementary force, for we have already seen that unsupported European ships were helpless close in to shore. Only slight changes in fortifications could greatly reduce the effectiveness of artillery. In the Benin area, artillery was not effective, because most of the fortifications were largely earthworks. Indeed, it is only because they could not defend their gate that the defenders of the town that Ulsheimer helped attack were defeated. The stockades of Sierra Leone were perhaps more vulnerable, but earthworks could render them much safer. The common use of earthworks and hedges of living trees in fortifications probably explains why the cannon had such little value as a siege engine in the Angolan wars of the sixteenth and seventeenth centuries.59 In this instance the Portuguese operations in Angola after 1575 are especially informative, for here the Portuguese attempted direct conquest with their own weapons and, at least in some instances, with their own soldiers. If European military technology and techniques were of special merit, surely this would be demonstrated in Angola. Portuguese operations might then be of the simple slave-raiding model that many scholars prefer, in which an all-powerful state conducts systematic wars on its weaker and ill-organized neighbors to gather slaves, relying on the strength of its weapons and the organization and size of its armies to ensure victory and minimize losses. If European weaponry or military organization were indeed superior, given their strong motivations to acquire slaves for export, one would surely expect this model to describe the Portuguese attacks in Angola. Certainly, the early exports of slaves from Angola were clearly linked to the operations of the Portuguese army. This can been seen in customs data from Luanda covering the period 1579–85, when great surges of exports in 1579–80 and lesser ones in 1581 and 1583 are correlated with wars (described in great detail by contemporary Jesuit observers), and the periods of relative peace in 1580 and 1584–5 show almost no exports.60 Similarly, Beatrix Heintze has estimated that the wars promoted by Mendes de Vasconcellos, Portuguese 58  Ulsheimer, “Raysen,” fols. 32a–32b. 59  Thornton, “The Art of War in Angola, 1570–1680,” Comparative Studies in Society and History 30 (1988): 360–78. 60  “Papéis vários pertencentes as Conquistas de América e India,” Biblioteca do Palácio de Ajuda (Lisbon), MS. 52-VIII-58, fols. 152–7V. See a survey of wars in Pero Rodrigues, “História,” in MMA 4:565–77.

The Process Of Enslavement And The Slave Trade

1143

governor from 1617 to 1622, resulted in the export of over 50,000 slaves in just a few years—though this ferocious rate of export was not kept up.61 But the documentation cannot support the idea that the Portuguese wars in Angola were simply raids of a militarily dominant European power against its weaker neighbors. Cadornega, the chronicler of the Angolan wars of the seventeenth century, was quick to point this fact out. His campaign and battle descriptions are lengthy and show a soldier’s eye for military detail. He often records fairly small-scale Portuguese operations conducted, one might easily say, simply for obtaining slaves, that ended in failure and disaster. After recounting one particularly difficult campaign, he asserts that this was far too difficult to be simply a “guerra de negros” (war for slaves). Although one can discount this comment in part as simply a reply to critics who believed that all the Angolan wars were just slave raids, with no political or diplomatic gains in mind (and perhaps running counter to such gains), the detail of his documentation does confirm that whoever conducted war to capture slaves was in no way guaranteed success and might well be killed by his quarry. Cadornega reports an apt saying in this context: “He who would singe another man’s whiskers had better look out for his own.”62 Portugal’s African enemies often possessed skilled and well-equipped armies and very often constructed strong fortifications. Cavazzi described in detail the complicated operations needed to attack one of these fortified locations during a campaign that he accompanied in 1659.63 This particular campaign, moreover, continued after taking the town, only to lose badly in another battle, with the result that virtually all its Portuguese members were killed.64 Likewise, although the Portuguese played the role of a heavily armored infantry in many of the campaigns, their presence was not decisive, and in most respects their tactics were identical to those of their enemies. Portuguese soldiers could not win unsupported by Africans and were regularly massacred when they tried to do so.65 If Angola was a major participant in the Atlantic slave trade and the source of export for many thousands of people, it was not through the superiority of European arms.

61  Heintze, “Ende,” pp. 204–9. 62  Cadornega, História 1:81, 204, 390. 63  MSS Araldi (Papers of the Araldi Family, Modena), Giovanni Antonio Cavazzi da Montecuccolo, “Missione evangelica al regno de Congo” (MS of 1665–8), vol. B, pp. 508–30. 64  Cadornega, História 2:157–63. 65  Thornton, “Art of War.”

1144

Thornton

In summary, we can say that although European arms may have assisted African rulers in war in some cases, they were not decisive. It is unlikely that any European technology or assistance increased the Africans’ chances of waging successful war (as the Portuguese in Angola could surely have attested) or that it made the attackers suffer fewer losses. Therefore, Europeans did not bring about some sort of military revolution that forced participation in the Atlantic trade as a price for survival.

The Rapid Growth of Slave Exports and Innovations in Military Technology and Warfare

It is possible to conclude that European influence over the slave trade may not have been significant in the first century and a half of the trade simply by acknowledging that Africans had slaves and a slave trade already, and that early forms of European military technology and organization were not critical to the success of African armies. But it might still be possible to argue that ultimately Europeans forced Africans to exceed their capacity to deliver slaves at a later period when high demands for slaves and improved military technology played a more important role.66 One potential piece of evidence is the dramatic increase in slave exports after 1650. This increase is roughly correlated both with the explosion of growth of plantation economies under northern European control in the Caribbean and with the large-scale arrival of northern Europeans on the African coast. These newcomers brought with them improved weapons technology and a generally greater industrial capacity than Portugal had. Could these events have signaled the arrival of a new and potentially more disruptive group of merchants and resulted in Africans being forced to expand the existing trade against their will? My research suggests, however, that the changes were more of quantity than quality, and that although the increased demand (and subsequent rise in

66  Gemery and Hogendorn (“Technological Change,” pp. 248–51) attempt to assess this hypothesis but by and large do so by showing that the musket was a “good weapon” or at least was a useful weapon, without really comparing it with its closest competitor as a missile weapon—the bow and arrow. Moreover, they also do not examine exactly how muskets were used in actual fighting or deal with how the musket would fare in action with hand-to-hand, rather than missile, tactics.

The Process Of Enslavement And The Slave Trade

1145

prices) may have persuaded more Africans to part with their slaves, it did not force them to do so against their will. African exports of slaves expanded dramatically beginning in the mid-seventeenth century, to the point where the number of exported slaves grew from being a relatively small number relative to the total population of the African regions from which they were taken to having a major demographic impact. Virtually all the work on the volume of the slave trade shows that the total number of slaves exported increased relative to the total areas or to the (estimated) African populations involved. The negative demographic impacts, although somewhat apparent in the beginning of the period in some areas (such as central Africa), intensified and spread to virtually the whole of Atlantic Africa. In the late eighteenth century much of Africa reached demographic exhaustion. It is possible to trace the growth in slave exports in considerable detail for most of the seventeenth century. Even allowing for a fair margin of error, it is obvious that exports did indeed grow significantly. Thanks to the detailed data on shipping available in the archives of Seville and the union of the Spanish and Portuguese crowns in 1580, which brought much of Portuguese trade in Africa under Spanish supervision,67 we have a detailed picture of trade in the last years of the sixteenth century to 1640, when the ending of the union once again clouds the picture. Table 4.1 gives some idea of the distribution of slave exports in time and space. Several points are worthy of note. First of all, it is clear that there was dramatic growth, increasing from a rate of 0.6 percent per annum in the sixteenth century to well over 1.5 percent per annum by the second half of the seventeenth century, with exports nearly doubling between 1650 and 1700. However, this growth was uneven, for some regions exported more slaves, while others maintained more or less the same level.

67  Excluded is the Angola–Brazil and Angola–São Tomé trades, for which we have some spotty customs records, most from the 1620s in Portuguese and Brazilian material, analyzed by Mauro, Portugal et l’Atlantique, p. 180.

1146 Table 4.1

Thornton Estimated annual exports of slaves from Africa, 1500–1700 1500

1550

1600

1650

1700

Coastal region Western coasta Gulf of Guineab West centralc

2,000 1,000 2,000

2,000 2,000 4,000

2,500 2,500 4,500

 2,500  3,300  8,000

 5,700 19,400 11,000

Total

5,000

8,000

9,500

13,800

36,100

a From the mouth of the Senegal River, but including Arguim, to Sierra Leone. b From the “Malaguetta” and “Kwakwa” coasts to Cameroon. c From Cameroon to the Cape of Good Hope. Sources: For the period around 1500, see documentation in Chapter 3. For the period around 1550 I have relied on Magalhães-Godinho, Economia 2:205–7. For the period 1600–50, basic data on shipping are compiled in Chaunu and Chaunu, Séville et l’Atlantique, 5:70–3, 138–41, 156–7, 188–9, 208–9, 218–19; summarized in Curtin, Atlantic Slave Trade, pp. 106–7. Various estimates by Chaunu and Chaunu and by Curtin are based on multiplying tons of shipping by presumed numbers of slaves per ton. Their estimates rely on a ratio of 1.8 slaves per ton; mine rely on an estimate of 4 per ton, hence the much higher figure. This higher estimate is based on my accepting the argument advanced in Bowser, African Slave, pp. 34–44, 363, n. 24, and Duncan, Atlantic Islands, pp. 200–1, that the slavers of the seventeenth century, unlike those of the eighteenth, tried to overload ships with slaves (both cite documentaion showing that this was a problem), rather than use the right to trade in slaves as a subterfuge for smuggling in other commodites. I have added data on the Angola and São Tomé trades according to Mauro, and include figures on the Dutch trade, based on shipping data, compiled by Johannes Postma, “The Origin of African Slaves: The Dutch Activities on the Guinea Coast,” in Stanley Engerman and Eugene Genovese, eds., Race and Slavery in the Western Hemisphere (Princeton, 1975), p. 49. This procedure has produced slightly different estimates from those of Lovejoy, Transformations in Slavery, pp. 53, 54, 56, 58–9. For the 1700 figure, I have used an average for the decade 1700–9 found in David Richardson, “Slave Exports from West and West Central Africa, 1700–1810: New Estimates of Volume and Distribution,” Journal of African History 30 (1989): 17 (Table 7).

For example, the trade of the western regions, such as Senegambia, hardly grew at all throughout the period. Angolan trade grew more, but still modestly, during the same period. Angola’s total exports moved from something on the order of 2,000–3,000 slaves in the early sixteenth century to 4,500 by century’s end, continuing to 8,000 by 1650 and eventually 11,000 by century’s end. The

The Process Of Enslavement And The Slave Trade

1147

growth increased the Angolan share of the trade from approximately half in 1500 to better than 65 percent by 1650. But Angolan growth was eventually eclipsed by the dramatic rise of exports from the Gulf of Guinea. In 1500 most slaves from this area came from Benin, but by the end of the sixteenth century Benin had ceased selling slaves in any appreciable number. This loss was more than compensated for by the rapid growth of the slave trade of Allada in the last half of the sixteenth century and throughout the seventeenth century.68 Slaves from this area first appear in American inventories about 1550: a “Lucumi” (Yoruba) slave first appears on Hispaniola in 1547,69 and “Ardras” (Allada) first appear in Peru in the 1560S.70 All such groups become increasingly numerous in the seventeenth century, both in absolute terms and relative to slaves from other areas in American data. Continued growth from the region of Allada and its immediate neighbors eventually earned the area the title of “Slave Coast” in the late seventeenth century. From virtually no exports in 1500, this region was exporting over 19,000 slaves per year, more than half the entire African total, by 1700. But the seventeenth-century growth of slavery in Lower Guinea (Gold Coast to Cameroon) was also enhanced by the entry of the Gold Coast states into the slave trade after 1630, and especially after 1650. This growth is particularly dramatic because the Gold Coast began the seventeenth century as a net importer of slaves and exporter of gold and ended it as a net exporter of slaves and was even importing gold.71 Thus, in order to understand why the numbers of slaves increased so dramatically in the seventeenth century, we really need to focus on Angola and the Lower Guinea region. For Angola we must seek the causes of the continued growth of slave trading, and for Lower Guinea the reasons for its people’s decision to participate in the trade and expand their exports toward the end of the century. 68  The proportion of the Gulf of Guinea slaves in the trade is exaggerated, in my view, in Philip Curtin’s discussion in The Atlantic Slave Trade: A Census (Madison, 1969), pp. 103–4, because he has assumed all ships declaring “Guinea” as their destination went to the Gulf of Guinea. Yet in seventeenth-century nomenclature, Guinea could be any African destination, from Senegal to Angola, and these slaves cannot be said to come certainly from the Gulf of Guinea only. 69  “Escriptura” of Licenciado Cerrato, in Incháustegui Cabrai, Reales cédulas 1:237–8. 70  Bowser, African Slave, p. 40. 71  Rodney, “Gold and Slaves on the Gold Coast” Transactions of the Historical Society of Ghana 10 (1969): 13–28.

1148

Thornton

There are several possible explanations for the growth of the slave trade in these areas. Both Curtin and Lovejoy have suggested that increases in the price of slaves, which can be documented for the period, might have enticed more slaves from their owners. It may have encouraged more “economic model” wars, and it may have persuaded owners that it would be better to forgo domestic use in exchange for the higher price available from the Atlantic trade. Also, owners of slaves living far from the coast might be willing to bear the transport costs of moving slaves to the coast if a higher price were offered.72 This explanation does give European merchants a role in the growth of enslavement in Africa, but it clearly places the economic decisions in the hands of Africans. Other explanations focus simply on the increase in wars caused by African political dynamics, discounting the role of trade. The connection between African trade, control over the trade, and politics is a complex and controversial one, but for our purposes, such an explanation still rules out European coercion. Finally, of course, there is the idea that European coercion, either direct or indirect, is responsible for the increase in warfare, which resulted in more slaves for the Atlantic. In the late seventeenth century the musket was developed into a more effective weapon. Moreover, very large numbers of such weapons were produced as European armies re-armed into bodies in which every infantryman carried a musket.73 Naturally enough, larger quantities of the improved weapons were also available to ship to Africa, where, it is argued, they may have revolutionized warfare. Thus, by directing weapons selectively to those willing to supply slaves, European merchants may have been able to effect the gun–slave cycle.74 A detailed examination of both Angola and the Gulf of Guinea can shed some light on probable causes for the transformation of slave exports. In both cases, however, it seems clear that economic motives and political motives not directly connected to the slave trade were far more important than European coercion or influence. In Angola, the growth can be explained in large measure by the fact that the same areas continued supplying slaves, and slaves whose capture took place farther and farther east joined the exodus from central Africa. For example, 72  Lovejoy, Transformations in Slavery, pp. 49–52; Curtin, Economic Change, pp. 156–68. 73  The flintlock musket was a much improved weapon over the earlier matchlocks, especially in that it had roughly twice the rate of fire and nearly half the number of misfires (see David Chandler, The Art of War in the Age of Marlbourough [New York, 1976], pp. 76–9). 74  This is the underlying argument of Inikori, “Import of Firearms.”

The Process Of Enslavement And The Slave Trade

1149

the war of the Ndongo succession really only ended in 1672, though hostilities between Matamba, Angola, and Kasanje (in various combinations, not always involving Portuguese participation) continued sporadically, as, for example, in 1679–85.75 If supplies of slaves captured in the wars of the Ndongo succession were lost with the ending of the war, the Kongo civil wars (1665–1718) surely contributed more than Kongo had earlier in the century, for all the central African slaves in the Remire plantation in Cayenne (French Guiana) acquired from Dutch traders between 1685 and 1690 were baptized Christians from Kongo.76 But, as we have already seen, the big wars were not the only source of Angolan slaves, for smaller campaigns (which some thought were no more than slave raids) also continued in the same areas as before. Cadornega’s chronicle, which provides detailed documentation on the operations of the Portuguese army up to 1681, mentions several such wars directed against the usual enemies: the Ndembu and Mbwila region to the north; Kisama, Benguela, and the central highlands to the south; and the lands bordering Matamba on the east. The effects of these eastern wars and similar operations by Matamba’s army were noted by several late seventeenth-century travelers. The absence of comprehensive chronicles, like Cadornega’s, for the period after 1681 obscures the exact direction and nature of these wars, but they surely seem to have continued.77 If improved musketry was somehow a factor in the conduct of any of these wars, great or small, it is not visible in these records. Portuguese military success seems no better in 1680 or even 1700 than it was a century earlier, whatever rearmament or reorganization may have taken place as a result of the entry of muskets. 75  Cadornega, História 2:403. 76  Thornton, Kingdom of Kongo, pp. 68–113; MS Montbret 125, Bibliothèque Municipale de Rouen (hereafter BM Rouen), fols. 86–90. 77  Cadornega, História 2:380–3, 347–9; also Manoel Ribejro to Antonio de Souza, 15 January 1674, MMA 13:254–65 passim; Bernardo da Firenze to Propaganda Fide, 22 June 1705, Archivio “De Propaganda Fide,” Scritture originali nelle congregazione generale, vol. 552, fols. 64V–65V; Filippo Bernandi da Firenze, “Ragguaglio del Congo” (1714), Archivio Provindale dei PP. Capuccini, Provincia da Toscana, fol. 685. The evidence cited here convinces me that the earlier pattern of warfare and enslavement was unchanged throughout the seventeenth century, which would invalidate the idea advanced by Joseph Miller in “The Paradoxes of Impoverishment in the Atlantic Zone,” in Birmingham and Martin, eds., History of Central Africa, 1:141–3; 145–50, that the expansion of warfare and enslavement to the east was the product of the exhaustion of those areas near Luanda and Ndongo that had formerly supplied the Atlantic and an increasing search eastward for fresh supplies of slaves.

1150

Thornton

To these sources, which supplied much of the earlier slave trade, came others from far in the interior. It seems reasonable to suggest that the motive for the capture and transport of these slaves, who often came from hundreds of kilometers east of the positions of the Portuguese, may well have been the higher prices paid for slaves in the late seventeenth century. Cadornega mentions contacts between Kasanje and the emerging Lunda state of the far interior that took place before 1680 but does not mention slaves as among Lunda’s exports. Nevertheless, Lunda did begin exporting slaves soon afterward, soon contributing a large supply, and capturing many during its wars of expansion and consolidation.78 It is interesting to note in this regard that Lunda’s armies were not reliant on guns from the Atlantic, for as late as the mid-eighteenth century they still eschewed muskets as cowards’ weapons.79 We have much less information about the causes of the sudden surge of slave exports from Lower Guinea. The region around Allada, first of the states of the Slave Coast to begin large-scale exports, is very poorly documented by sixteenth- and early seventeenth-century sources, and it is not really until well into the seventeenth century that this situation is remedied. Oral traditions collected in the nineteenth century suggest that this period was characterized by the rise of the powerful Oyo Empire, which is perhaps in some way correlated with the surge of exports.80 That many of Allada’s exports were slaves captured during Oyo expansion is suggested by the fact that Capuchin visitors of the 1660s believed that many of the slaves Allada exported came from the interior and were purchased at markets.81 Allada and its subject states and neighbors fought numerous wars during the later seventeenth century, even as the kingdom of Dahomey came to dominate the interior and then the coast.82 The local African politics of the Gold Coast, the second region to enter the slave trade from Lower Guinea, involved a complicated series of wars between 78  For this period, see John Thornton, “The Chronology and Causes of Lunda Expansion to the West, c. 1700–1852,” Zambia Journal of History 1 (1981): 6–9. 79  Manuel Correia Leitão, “Viagem que eu, sargento-mór dos moradores do distrito do Dande, fiz ás remotas partes de Cassange e Olos, no ano de 1755 até o seguinte de 1756,” in Gastão Sousa Dias, ed., “Uma viagem a Cassange nos meados do seculo XVIII,” Boletim da Sociedade de Geografia de Lisboa 56 (1938): 20–21. 80  Studied in detail by Law, Oyo Empire, pp. 33–53. On our documentary knowledge of the Slave Coast, see idem, “Problems of Plagiarism, Harmonization and Misunderstanding,” pp. 337–58. 81  Biblioteca Provincial de Toledo, Coleción de MSS Borbón-Lorenzana, MS. 244, de Zamora, “Cosmografia,” fol. 61. 82  The period after 1680 is well documented in I. A. Akingogbin, Dahomey and Its Neighbors, 1700–1828 (London, 1966).

The Process Of Enslavement And The Slave Trade

1151

the local states, whose motives are not clear to us and were equally unclear to the Europeans, although the more perceptive observers, such as Willem Bosman, the Dutch factor, provided detailed historical background for some. In total, however, the complexities of local politics and the steady rise of the interior kingdoms of Denkyira, Akwamu, and ultimately Asante overwhelmed the petty politics of the area—the result of deep-seated social changes in the interior kingdoms that owed little to coastal influences, at least as it has been analyzed by Kea.83 The military reorganization of Lower Guinea that led to the rise of the great interior powers, such as Asante and Dahomey (and perhaps even the late seventeenth-century expansion of Oyo), has often been blamed on imports of European firearms. But Kea, whose study of this period for the Gold Coast takes such military factors seriously, notes that although the methods of warfare were revolutionized by the interior kingdoms, it was mainly by their use of mass-recruited armies, and firearms had very little effect. The development of these mass armies was the product of social changes and was not determined by the availability of new military technology. Indeed, he argues that the early expansion of Asante was accomplished by mass armies armed with missile weapons, but these were bows and arrows in the crucial early phases, and only later were the troops re-armed with muskets.84 It is worth noting that the creation of mass armies and their subsequent rearming with firearms may have done a great deal to increase the numbers of people enslaved. If earlier wars involved relatively small professional armies, and the majority of the slaves were taken from the military captives, then obviously the group of people vulnerable to enslavement would be fairly small. But with the rise of mass armies, battles were likely to involve more soldiers, thus increasing the number of potential slaves accruing to the victor. However, the simple correlation between imports of firearms and exports of slaves is not a causal relationship. It is more likely that African demands for guns increased simply because they were creating larger armies, which itself had complicated internal, social causes. The availability of European weapons did not provoke an increase in warfare. As African armies re-armed and became accustomed to the tactics of musket warfare it became harder to go back to some other “art of war,” thus ensuring continued demand. This certainly helped European business in general, but it did not deliver to any European power the capacity to engage in weapons blackmail against states that might wish to refuse to sell slaves. This is because 83  Kea, Settlements, pp. 158–64. 84  Ibid., pp. 97–117.

1152

Thornton

no European country or group of merchants came close to having control over the supply of arms to any African state, at least in any but the very shortest run. Europeans could therefore not freely decide to supply arms or not to supply them to force Africans into any decisions. In any case, the only real form of influence available to merchants would be withholding the very means to make war, and such a strategy would be more likely to inhibit than encourage warfare. As historians learn more about warfare in Africa in this period, and as they probe more deeply into the political and social structures of African states, they realize that warfare needs to be explained in terms of the internal dynamics of the state or state system. As such dynamics are understood, the role of Europeans in causing war (as opposed to benefiting from it, either as a vehicle to sell arms or to buy slaves) begins to diminish. Thus, for example, the study of the Kongo civil wars of the late seventeenth century yields explanations for the wars that lie in the politics of the country and not in Portuguese machinations, as was previously believed.85 The same conclusions can be drawn from the study of the Slave Coast and Gold Coast, where the explosion of slave exports and growth of arms imports are the clearest. The numerous surviving letters of Dutch and English factors on the Gold Coast from about 1680 onward certainly tell of a willingness to buy slaves, at least “if the price is right,” as one factor wrote in 1683,86 but there is nothing to suggest that they could or did exercise actual pressure to get the local people to sell them.87 They did certainly encourage and occasionally bribe local rulers in the multistate system of the coast to fight, including supplying them with arms and even soldiers, but it was normally to get military help in driving other European rivals from their posts and not simply to get slaves.88 One of the best examples of such an event was the Komenda war of the 1690s, detailed in several contemporary sources. Its origins lay in the 85  Thornton, Kingdom of Kongo; Hilton, Kingdom of Kongo. Though these two works differ substantially in their approach and conclusions on the causes of the civil wars, both assign essentially domestic causes and see European influence as peripheral at best. 86  Ralph Hassel to Elmina, 5 May 1683, Bodleian Library, Oxford, MSS Rawlinson, C 745, fol. 191V. 87  See especially the reports of English factors in ibid., passim, and in the Dutch records preserved in the Algemeen Rijkisarchiv, Nederlands Bezittung ter Kust Guinea. See the sample of such correspondence published in English translation by Albert van Dantzig, The Dutch on the Guinea Coast: Select Documents (Accra, 1978). 88  Hassel to Cape Coast, 5 May 1683, Bodleian Library, Oxford, MSS Rawlinson, C 745, fol. 191; Thomas Pearson to Cape Coast, 22 May 1683, ibid., fol. 201v–2 (but retelling events back to 1667); Mark Bedford Whiting to Cape Coast, 18 August 1683, ibid., fol. 281.

The Process Of Enslavement And The Slave Trade

1153

complicated politics of African trade, or, as the Dutch factor Bosman said, in “bad government and absurd customs.” All sides obtained mercenaries from the Europeans.89 The relatively small size of African states and the prevalence of professional armies along the coast (it was only in the interior that the mass armies were forming) made small bodies of mercenaries potentially effective, and consequently Europeans frequently hired themselves out in this way. The practice of acting as mercenaries was not restricted to Europeans; several states on both the Gold Coast and the Slave Coast routinely supplied mercenaries in the wars of the period.90 Robin Law has recently proposed that the kingdom of Dahomey served first as a sort of mercenary state.91 Their role as suppliers of mercenaries did not give the Europeans much power on the coast; rather, one gets a much stronger impression of European weakness and helplessness in the face of local African politics.92 African rulers continued to engage in wars, not unlike those of previous centuries, and naturally, as the new weapons figured more prominently in warfare, acquiring supplies of the weapons became important. Thus, in the late seventeenth century and into the eighteenth century, civil wars troubled the Senegambian states, and often pretenders sought and acquired weapons in order to make their claims. But it would be incorrect to say that somehow Europeans had persuaded the potential candidates to seek power in order to get slaves, even if they did delight in the prospect of increased slaves as a resuit.93 Senegalese state leaders built up substantial armies of slave soldiers, and often these armies engaged in local raiding (frequently without royal permission), which proved quite disruptive,94 but neither the origin of these 89  The political background is outlined in Bosman, Description, pp. 175–83 (quotation, p. 164). 90  On mercenary armies of African origin, see Barbot, “Voyage,” pp. 319–22, 351; Bosman, Description, pp. 362a, 395–6. 91  “Dahomey and the Slave Trade: Reflections on the Historiography of the Rise of Dahomey,” Journal of African History 27 (1986): 237–67. 92  David Harper to Cape Coast Castle, 20 April 1683, Bodleian Library, Oxford, MSS Rawlinson, C 754, fol. 189; David Harper to Cape Coast Castle, 5 May 1683, ibid., fol. 192; Hugh Shears to Cape Coast Castle, 21 May 1683, ibid., fol. 199V and passim. 93  See the detailed study, focusing on the early eighteenth century, in Charles Becker and Victor Martin, “Kayor and Baol: Senegalese Kingdoms and the Slave Trade in the Eighteenth Century,” (trans. Linda Zuck) in Inikori, Forced Migration, pp. 100–25. 94  Ibid., pp. 120–2 (dealing with the eighteenth century); for an analysis of internal dynamics, see James Searing, “Aristocrats, Slaves and Peasants: Power and Dependency in the Wolof States, 1700–1850,” International Journal of African Historical Studies 21 (1988): 475–504.

1154

Thornton

armed forces nor their kings’ lack of ability or desire to control them was the result of European policies or pressures. In conclusion, then, we must accept that African participation in the slave trade was voluntary and under the control of African decision makers. This was not just at the surface level of daily exchange but even at deeper levels. Europeans possessed no means, either economic or military, to compel African leaders to sell slaves. The willingness of Africa’s commercial and political elite to supply slaves should be sought in their own internal dynamics and history. Institutional factors predisposed African societies to hold slaves, and the development of Africa’s domestic economy encouraged large-scale trading and possession of slaves long before Europeans visited African shores. The increase in warfare and political instability in some regions may well have contributed to the growth of the slave trade from those regions, but one cannot easily assign the demand for slaves as the cause of the instability, especially as our knowledge of African politics provides many more internal causes. Given the commercial interests of African states and the existing slave market in private hands in Africa, it is not surprising that Africans were able to respond to European demands for slaves, as long as the prices attracted them.

Europeans and the Rise and Fall of African Slavery in the Americas: An Interpretation David Eltis Of the many post-neolithic slave societies, those of the European-dominated Americas appear to have had the most obvious economic foundations. In the words of a recent, widely read survey, “The Slaves of the New World were economic property, and the main motive for slaveholding was economic exploitation.”1 Historians are careful to distance themselves from purely economic theories of human behavior, but on this issue the distance is usually rather short. The often bitter debates on the nature and meaning of New World slavery have produced few since Adam Smith who questioned the basic motivation of the early plantation owners. Simply put, people from one continent forced those from a second continent to produce a narrow range of consumer goods in a third—having first found the third’s native population inadequate to their purpose. Even those who have wrestled with the relationship between racism and slavery have seen the racial basis of American slavery primarily as an economic phenomenon. Eric Williams, Oscar Handlin, Carl Degler, Winthrop Jordan, William McKee Evans, and others may disagree on the origins of racism but not on the origins of racial slavery in the Americas. Slaves from Africa were used to grow sugar and other plantation crops, it has been argued, because they comprised the least-cost option. Even George Frederickson, the historian who has perhaps looked furthest beyond class and economics for the sources of racism, has written that if white slavery had appeared profitable, Source: Eltis, David, “Europeans and the Rise and Fall of African Slavery in the Americas: An Interpretation,” American Historical Review, 98(5) (1993): 1399–1423. By permission of Oxford University Press. * I would like to thank Hilary McD. Beckles, Robin Blackburn, Timothy Joel Coates, Carl Degler, Seymour Drescher, Pieter C. Emmer, Stanley L. Engerman, Shelley Price Jones, Robert W. Malcolmson, James Pritchard, Mary Turner, and anonymous referees for extensive and helpful comments on earlier versions given at the Anglo-American Conference of Historians at the Institute of Historical Research, University of London, in July 1992, and the Department of History, Boston College, in April 1993. I have also benefited from exchanges with David and Nancy Northrup. Funding was provided by the Social Science and Humanities Research Council of Canada. 1  Robin Blackburn, The Overthrow of Colonial Slavery: 1776–1848 (London, 1988), 7.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_037

1156

Eltis

it would have been introduced.2 And, indeed, everything we know about early modern European commercial elites and planters would support such an interpretation. Whatever our definition of capitalism, we would expect such elites to have used the cheapest option possible within the limits of mercantilist policies. But did this in fact happen? It has become clear in recent years that economic paradigms have limited usefulness in explaining the ending of slavery. The number of slaveholders who voluntarily converted their slaves into wage-earning laborers in order to increase profits was not large, nor were the numbers of non-slaveholders who benefited directly from abolition. The major studies of abolition now draw rather heavily on the cultural and ideological spheres of human activity. Yet the origins of the system are in need of a similar reassessment. What follows is an attempt at a fresh examination of slavery in the Americas that suggests economics, narrowly defined, is no more capable of explaining its origins than its abolition. Put differently, the question is not why slavery per se but rather which groups are considered eligible for enslavement and why this eligibility changes over time. Slavery until recently was universal in two senses. Most settled societies incorporated the institution into their social structures, and few peoples in the world have not constituted a major source of slaves at one time or another. The African component may dominate interpretations of slavery in the Atlantic world and, from somewhat earlier, in the Islamic world, but the more fundamental question from a longer and wider view is what separates outsiders— those who are eligible for enslavement—from insiders, who are not. Thus Nathan Huggins has answered the often-asked question of how Africans could enslave other Africans and sell them into the slave trade with the astute response that the enslavers did not see themselves or their victims as Africans. Richard Hellie has made the same point differently in writing about the efforts of slaveowners in sixteenth-century Russia to claim spurious foreign origins for themselves so that the enslaved could be held at a distance.3 The question of why certain groups are deemed more appropriate than others for enslavement and the degree to which their status can be changed are of interest to anthropologists studying non-Western societies more often than historians 2  George M. Frederickson, The Arrogance of Race: Historical Perspectives on Slavery, Racism, and Social Inequality (Middletown, Conn., 1988), 194. 3  Nathan Irvin Huggins, Black Odyssey: The Afro-American Ordeal in Slavery (New York, 1977), 20; Richard Hellie, Slavery in Russia, 1450–1725 (Chicago, 1982), 393–94. I would like to thank Stanley L. Engerman for drawing my attention to these references.

Europeans And The Rise And Fall Of African Slavery

1157

and economists.4 However, posing such questions in the European rather than the Asian, African, and indigenous American contexts promises some new insights. If almost all societies in Europe, Africa, and the Americas, and, indeed, in the rest of the world, accepted slavery at the time of the Columbian contact, they had very different definitions of “outsider” status. In Africa and the Americas, such status might include anyone who was not part of the immediate lineage, but, more frequently, it was restricted to those not belonging to the nation or tribe—implying a somewhat wider definition of “insider.” This was roughly the situation in Europe during and immediately after Roman times; by the fifteenth century, in Western Europe at least, the concept of insider had come to include all natives of the subcontinent, among whom there were some who were non-white, although very few who were non-Christians.5 Slavery in Europe was not extensive, but where it existed it was confined to non-Christians or natives of Africa and their immediate descendants. A similar pattern was established in the Islamic world, substituting of course “Islamic” for “Christian” in this assessment. The line dividing insider from outsider in the European case had some flexibility even in the short term. The Spanish, though not the Portuguese or the English, had banned the enslavement of American Indians by the 1540s. Despite this move to include non-Europeans, it was religious rather than ethnic barriers against outsiders that fell first. Jews and Muslims (or at least North African and European Muslims) were accorded insider status well before Africans. But the line was never drawn strictly in terms of skin color or race, however defined. The Spanish in America may have reserved full chattel slavery for sub-Saharan Africans, but slaves in Sicily as late as 1812 were Arab or at least North African Muslims. Among insiders, the Portuguese-speaking community

4  See James L. Watson, “Slavery as an Institution, Open and Closed Systems,” James L. Watson, ed., Asian and African Systems of Slavery (Berkeley, Calif., 1980), 1–15, and other essays in this volume for an introduction to the literature. 5  A full evaluation of the amalgam of the cultural norms, behavior patterns, physical appearance, circumstances of birth, class, and economic status that determined insider-outsider status would require much more space than is available here, and precision is not, in any event, essential to the argument. Winthrop D. Jordan’s White over Black: American Attitudes toward the Negro, 1550–1812 (Chapel Hill, N.C., 1968) focuses on the English and Africans and puts more emphasis on the ethnic component in the insider-outsider divide than the present essay, but the early chapters still provide useful insights into European perceptions of nonEuropeans in the early modern era.

1158

Eltis

that ran the slave traffic to Bahia had an African component, and one of the slave ship captains of the Royal African Company in its heyday was black.6 Moreover, the widening process was neither inevitable nor irreversible. Apart from the Russian case and well-known “second serfdom” in Eastern Europe, there was the reinstitution of serfdom in seventeenth-century Scottish coal mining.7 These may have stopped short of or avoided slavery, but, as recently as 1802, the French reinstated full chattel slavery in their colonies a decade after it had been abolished. The key issue is thus not why slavery died out in Western Europe by the early modern period, or even why it had become confined to groups from other continents by 1500, but rather why it was not reintroduced on its old (and less exclusionary) basis when the potential of American mines and plantations became apparent. The reasons for changes in perceptions of which groups were suitable for enslavement are not entirely clear, although a tentative explanation for the European case is offered below. The main thrust of the present argument is, first, that a study of such change should be central for anyone wishing to understand slavery in the Americas and, second, that economic motivation should be assigned a subsidiary role in the rise and fall of the exclusively African-based bondage that Europeans carried across the Atlantic. I begin with a cursory assessment of the existing literature before exploring one crucial area that this literature has tended to by-pass. Comparisons between Europe and the rest of the Atlantic world and between one European country and another form the basis of an alternative view of the issues presented in the final section. In the last decade or so, the short description of the dividing line between insiders and outsiders has been “power relationships.” Analyses along class and interest lines have dominated literatures on both the rise and fall of American slavery. Elites in Europe and Africa cooperated to ensure a supply of labor for the American plantations, and the shifting perceptions of a European elite— driven by protest from below in either the dominant or the slave society, according to some—brought the system to an end. Myths about slaves, other

6  Great Britain, Public Record Office, Royal African Company (RAC) to Benjamin Alford, March 10, 1702, Series T70, piece 58 (henceforth, T70/58), fols. 16–17; Pierre Verger, Flux et reflux de la traite des nègres entre le Golfe de Bénin et Bahia de Todos os Santos du XVIIe et XIXe siècle (Paris, 1968), 460–75. For slavery in Sicily, see Charles Verlinden, The Beginnings of Modern Colonization: Eleven Essays with an Introduction (Ithaca, N.Y., 1970), 27, 40. 7  Twentieth-century Europe provides many further examples of “inward” shifts in perceptions of insider status.

Europeans And The Rise And Fall Of African Slavery

1159

underclasses, and foreigners emanated from associations that at root and at first had a large economic component.8 This approach fits rather well with the more narrowly focused literature on the transition to slavery in the Atlantic world. It is frequently argued that the substitution of African slaves for the various forms of coerced native labor and European indentured servants in the plantation regions of the Americas was driven by relative costs and was thus in essence an economic decision. Debate over the mechanics of this process continues. For British North America, a combination of an elastic supply of slave labor, falling slave transportation costs, and rising servant prices governed the transition, as laid out by David Galenson, Russell Menard, and others.9 For Brazil, Stuart Schwartz has shown, with less formal analysis, how more expensive and more productive African slaves came to replace their Brazilian Indian counterparts in the first seventy years of sugar production in the Recôncavo. Similar switches from indigenous to African labor occurred in the sixteenth-century Caribbean as well as later in South Carolina. In 1708, nearly one in four of the slave population in South Carolina was Indian, and the colony exported captives to the West Indies. Farther north, French Canadians could not afford African slaves, and their slaves remained overwhelmingly Amerindian.10 By contrast, the southern 8  William McKee Evans, “From the Land of Canaan to the Land of Guinea: The Strange Odyssey of the ‘Sons of Ham,’” AHR, 85 (February 1980): 15–43; and for the more recent literature, Barbara Jeanne Fields, “Slavery, Race and Ideology in the United States of America,” New Left Review, 181 (May–June 1990): 95–118. For a historiographical survey and a dissenting view, see William A. Green, “Race and Slavery: Considerations on the Williams’ Thesis,” Barbara L. Solow and Stanley L. Engerman, eds., British Capitalism and Caribbean Slavery: The Legacy of Eric Williams (Cambridge, 1987), 25–49. 9  Russell R. Menard, “From Servants to Slaves: The Transformation of the Chesapeake Labor System,” Southern Studies, 16 (Winter 1977): 355–90; David W. Galenson, White Servitude in Colonial America: An Economic Analysis (Cambridge, 1981). For a recent discussion of trends in contract lengths (an indicator of servant prices), which seem not to have increased in the 1654–1775 period, see Farley Grubb, “The Long-Run Trend in the Value of European Immigrant Servants, 1654–1831: New Measurements and Interpretations,” Research in Economic History, 14 (1992): 167–240. 10  Stuart B. Schwartz, Sugar Plantations in the Formation of Brazilian Society, Bahia, 1550– 1835 (Cambridge, 1985), 15–27, 51–72; Peter H. Wood, “Indian Servitude in the Southeast,” Wilcomb E. Washburn, ed., Handbook of North American Indians, Vol. 4: History of IndianWhite Relations (Washington, D.C., 1988), 407–09; Theda Perdue, Slavery and the Evolution of Cherokee Society, 1540–1866 (Knoxville, Tenn., 1979), 19–35; Robin W. Winks, The Blacks in Canada: A History (New Haven, Conn., 1971), 1–19. Sanford Winston commented that “the relationship between the number of Indian and Negro slaves was … inverse,” in “Indian Slavery in the Carolina Region,” Journal of Negro History, 19 (October 1934): 436.

1160

Eltis

temperate zones, closer to Africa and yielding exportable agricultural commodities of coffee and hides, had an entirely African slave labor force. Much remains unexplored, but future research will not undermine the role played by economic factors in these transitions. From a broad perspective, the largescale traffic in slaves for plantation use focused first on northeastern Brazil, the region closest to Africa, moved next to the easternmost of the Antilles, Barbados, and had an impact on the Americas, farthest away from Africa, only later. If we leave aside the issue of alternative forms of labor, this process is consistent with trends in slave prices and transatlantic shipping costs. But were costs as central as the extensive literature on the labor transition assumes? If coerced native, indentured or free waged European, and African slave labor were the only options open to planters, then the answer must be yes. But why should these combinations of ethnic groups and labor regimes be the only possibilities for the early plantation Americas? The extent to which other alternatives were attempted, or even considered, is ignored in the literature, despite the fact that widening the question provides insights into more than just substitution of one form of labor for another. As Eric Williams realized half a century ago, what is at stake here is not just the economics and morality of early European expansion but the foundation of relations between European and African peoples in the Americas. The current literature cannot deal easily with the questions “why no European slaves?” or “why no African indentured servants?” even though Winthrop Jordan posed the former a quarter-century ago.11 Scholars are now less likely to counterpose slavery and freedom, or coercion and consent, especially in studying the beginning and ending of slavery in the Americas.12 There can be no doubt that major groups subjected to slavery, such as Canaanites, Slavs, and Africans, were quickly assumed fit subjects for such status even if they were not so regarded initially. But the creation of later 11  Jordan, White over Black, 66. 12  Rebecca Scott, “Comparing Emancipations: A Review Essay,” Journal of Social History, 20 (Spring 1987): 565. Most recently, see the special issue of Slavery and Abolition, 14 (April 1993), Michael Twaddle, ed., titled The Wages of Slavery: From Chattel Slavery to Wage Labour in Africa, the Caribbean, and England (London, 1993). For a similar orientation, this time at the inception of the system and centered on the study of white labor in Barbados, see Hilary McD. Beckles, White Servitude and Black Slavery in Barbados, 1627– 1715 (Knoxville, Tenn., 1989), esp. 5–10. Orlando Patterson’s distinction between slave and non-slave, centered on the absolute power of the slave’s master and the origin of enslavement as an alternative to death, is a useful corrective to this tendency to lump together slave and non-slave status (Slavery and Social Death: A Comparative Study [Cambridge, Mass., 1982], 21–27).

Europeans And The Rise And Fall Of African Slavery

1161

stereotypes does not explain why Europeans went thousands of miles to Africa for slaves in the first place. Nor does it explain why at some point between the eighteenth and nineteenth centuries, in the face of three centuries of stereotyping, the majority of Europeans and their descendants decided that Africans were no longer suitable subjects for enslavement. If we wish to understand the origins of African slavery in the New World, or indeed in the pre-Columbian Old World, we must first explore the labor options of early modern Europeans—both those that were tried and those that were not. Second, we need to assess how close Europeans came to imposing slavery or slave-like conditions on other Europeans, and, finally, what for them set slavery apart as a status for others. These steps will help clarify the cultural and ideological parameters that at once shaped the evolution of African New World slavery and kept Europeans as non-slaves. Various forms of European forced labor were in fact tried. A comparison of these options across national boundaries reveals differences in what major European colonial powers considered feasible as labor regimes. Yet no West European power after the late Middle Ages crossed the basic divide separating European workers from full chattel slavery. And while serfdom fell and rose in different parts of early modern Europe and shared characteristics with slavery, serfs were not outsiders either before or after enserfment. The phrase “long-distance serf trade” is an oxymoron. Even in the twentieth century, totalitarian states have used slave labor primarily as a punitive strategy against enemies of the state and have never instituted full chattel slavery as an economic device. Although there is no evidence that Europeans ever considered instituting full chattel slavery of Europeans in overseas settlements, the striking paradox is that no sound economic reasons spoke against it. By the seventeenth century, the most cursory examination of relative costs suggests that European slaves should have been preferred to either European indentured labor or African slaves. And while American Indians were cheap to enslave, their life expectancy and productivity in post-Columbian plantation conditions hardly compared with that of pre-industrial or, indeed, post-industrial Europeans. Before pursuing the issue of costs, it is worth noting that social devices used in African and Indian societies to deprive people of their liberty were fully incorporated into European societies. Judicial process sent English, French, Portuguese, and, to a lesser extent, Spanish men and women to the plantations. It also condemned Frenchmen, Spaniards, Germans, Italians, and Poles to the galleys. England, the Netherlands, and France used houses of correction to exact labor from the poor, beginning in the sixteenth century, and, in the late seventeenth century, from criminals as well. Impressment into service was also

1162

Eltis

prevalent in Europe. The early literature on indentured servitude and engagés stressed the role of force in the acquisition of servants. In Scotland, an act of 1672 empowered coal mine owners to kidnap vagrants and set them to work.13 If in the Americas and Africa, informal raids or more organized military actions were major ways to obtain slaves, then in Europe wars appear to have been frequent enough both within and between nations—setting aside the incentive to acquire and sell slaves—to ensure a healthy supply of captives. The holding of captives and the infliction of death and torture in both war and as punishment were common to European, African, and Native American societies at the time of the founding of transatlantic slave colonies. By the seventeenth century, Europeans might not enslave other Europeans, but enslavement does not appear brutal or unlikely in view of the hanging, torturing, mutilation, and burning that they inflicted on each other. European prisoners, political as well as criminal, were frequently sent to the plantations instead of to execution, and the substitution of slavery for death, explicitly recognized by John Locke,

13  The Scottish act allowed “coal-masters, salt-masters, and others, who have manufactories in this kingdom, to seize upon any vagabonds or beggars … and put them to work in the coal-heughs or other manufactories, who are to have the same power of correcting them and the benefit of their work as the masters of correction houses.” Cited in James Barrowman, “Slavery in the Coal-Mines of Scotland,” Transactions of the Mining Institute of Scotland, 19 (1897–98), part 2: 119. For a fuller discussion of this and English analogies, see David Eltis, “Labour and Coercion in the English Atlantic World from the Seventeenth to the Early Twentieth Centuries,” Slavery and Abolition, 14 (April 1993): 207–26. For the “spiriting” of servants and vagrants from England to Barbados, see Beckles, White Servitude and Black Slavery, 52–58. For French treatment of engagés, see Gabriel Debien, “Les engagés pour les Antilles (1634–1715),” Revue d’histoire des colonies, 38 (1951): 5–274; and for vagrants, see Andre Zysberg, “Galley and Hard Labor Convicts in France (1550–1850): From the Galleys to Hard Labor Camps; Essay on a Long Lasting Penal Institution,” Centrum voor Maatschappij Geschiedenis, 12 (1984): 78–110, and Zysberg, “La société des galeriens au milieu de XVIIIe siècle,” Annales: Economies, Sociétés, Civilisations, 30 (January–February 1975): 43–65, for which references I thank James Pritchard. For Spanish vagrants, see I. A. A. Thompson, “A Map of Crime in Sixteenth Century Spain,” Economic History Review, 21 (August 1968): 244–50; and Ruth Pike, Penal Servitude in Early Modern Spain (Madison, Wis., 1983), 49–54, 67–69, 90–92. For Portuguese transportation, see Timothy Joel Coates, “Exiles and Orphans: Forced and State-Sponsored Colonizers in the Portuguese Empire, 1550–1720” (Ph.D. dissertation, University of Minnesota, 1993), which fills a large gap in the historiography of penology. There are brief references in Charles R. Boxer, The Golden Age of Brazil, 1695–1750 (Berkeley, Calif., 1962), 140, and Stuart B. Schwartz, “The Formation of a Colonial Identity in Brazil,” Nicholas P. Canny and Anthony Pagden, eds., Colonial Identity in the Atlantic World, 1500–1800 (Princeton, N.J., 1987), 22–23.

Europeans And The Rise And Fall Of African Slavery

1163

has legitimated slavery throughout history. The similarity of the mechanisms for depriving people of their liberty on all four continents is striking.14 From the strictly economic standpoint, one might imagine there were strong incentives for using European rather than African slave labor. The crux of the matter is shipping costs, which comprised by far the greater part of the price of any form of imported bonded labor in the Americas. First, although the trade winds of the Atlantic reduced the differential somewhat, it was normally quicker to sail directly to the Americas from Europe than to sail via Africa. In addition, mortality and morbidity among both crews and passengers (or slaves) were lower in the North Atlantic than in the South. If we take into account the time spent collecting a slave cargo on the African coast as well, then the case for sailing directly from Europe with a cargo of Europeans appears stronger again. In the 1680s, the Royal African Company (RAC) would often hire ships to carry slaves on its behalf. The hire rate was typically between £5 and £6 per slave landed alive in the Americas—a price that was understood to cover the full cost of sailing to Africa, acquiring slaves, and carrying them to the RAC’s agents in the West Indies. The return cargo from the Americas was a separate speculation.15 The cost of shipping convicts to Barbados and the Leeward Islands at this time was similar.16 As noted below, however, ships carrying convicts, indentured servants, and fare-paying passengers always carried far fewer people per ton than did slave ships. There is little doubt that if ships carrying Europeans had been as closely packed as those carrying Africans, costs per person would have been much lower for Europeans than for Africans. A further reason for using European rather than African slave labor derives from relative prices of African slaves and convict English labor—the nearest the English came to using Europeans as chattel slaves. (Although the discussion that follows centers on the eighteenth century, for which the data are relatively good, scattered evidence from the seventeenth century is also consistent 14  For discussion of these in an African context, see Suzanne Miers and Igor Kopytoff, “African Slavery as an Institution in Marginality,” in Miers and Kopytoff, eds., Slavery in Africa: Historical and Anthropological Perspectives (Madison, Wis., 1977). For the same phenomena in one important North American context, see William A. Starna and Ralph Watkins, “Northern Iroquoian Slavery,” Ethnohistory, 38 (January 1991): 34–57. 15  Kenneth G. Davies, The Royal African Company (London, 1957), 198. For sample rates, see PRO, T70/943, fols. 12 and 13. 16  John C. Jeaffreson, ed., A Young Squire of the Seventeenth Century: From the Papers (A.D. 1676–1686) of Christopher Jeaffreson, 2 vols. (London, 1878), 1: 159; 2: 206–07. For information on costs of freighting indentured servants, compare Hilary McD. Beckles and Andrew Downes, “The Economics of Transition to the Black Labor System in Barbados, 1630–1680,” Journal of Interdisciplinary History, 18 (Autumn 1987): 235.

1164

Eltis

with the conclusions.) Unskilled male convicts from England and Ireland sold for £16 each in Maryland in the years 1767–1775 at a time when newly arrived African male slaves in the prime age group were selling for about triple this amount in Virginia and Maryland.17 The British males worked for ten years or less, the Africans for life. If convicts and their descendants had been sold into a lifetime of service, it is reasonable to suppose that planters would have been ready to pay a higher price for them. At this higher price, the British government and merchants might have found ways to provide more convicts. Shipping costs alone would not have interfered with the process. From this standpoint, convicts could have been sold into lifelong servitude for a price little more than that for seven or ten years of labor. Thus in the absence of an improbably rapid decline in slave prices as buyers switched from Africans to Europeans, we might suppose that there were no shipping-cost barriers to European slaves forming the basis of the plantation labor forces of the Americas. The other major cost categories in shipping Africans to the New World were enslavement, factoring (the costs of assembling a cargo prior to shipment), and distribution (the costs of selling a slave in the Americas).18 Together, they made up about half the cost of a new slave in the Americas. According to Philip Curtin, enslavement costs in Africa were trivial. Most of the price of a slave sold to a factor on the African coast consisted of transportation costs.19 Could this low cost at the point of enslavement explain the apparent preference for Africans? Given the opportunities in Europe for enslavement discussed above, it seems unlikely. From an economic standpoint, adapting these opportunities to produce slaves instead of merely prisoners would have been neither difficult nor costly. Indeed, enslavement in Europe might have been less costly than its African counterpart. First, transportation costs, which loomed so large within Africa, were bound to be lower in a subcontinent where major population centers were located near navigable waters. Established routes for “chains” existed for French convicts heading to Marseilles, Toulon, and later Brest, for Spanish 17  Calculated from Richard Nelson Bean, The British Trans-Atlantic Slave Trade, 1650–1775 (1971; New York, 1975), 206–08; A. Roger Ekirch, Bound for America: The Transportation of British Convicts to the Colonies, 1718–1775 (Oxford, 1987), 71, 124–25. 18  These categories are taken from David Turnbull’s work. Turnbull had extensive experience of the nineteenth-century slave trade. See his Travels in the West: Cuba, with Notices of Porto Rico, and the Slave Trade (London, 1840), 403–06. 19  Philip D. Curtin, Economic Change in Precolonial Africa: Senegambia in the Era of the Slave Trade, 2 vols. (Madison, Wis., 1975), 1: 156–57, 168–69, 173–77; 2: 45–53; and “The Abolition of the Slave Trade from Senegambia,” in David Eltis and James Walvin, eds., The Abolition of the Atlantic Slave Trade (Madison, 1981), 89–91.

Europeans And The Rise And Fall Of African Slavery

1165

convicts to Cartagena, and for Portuguese to Lisbon, some incorporating travel on canals. In England, the barges carrying parish children from London to northern textile mills could have just as easily carried people to major ports. Second, population growth in Western Europe in general, and England in particular, was considerable during the era of the slave trade. Despite a net migration of 2.7 million from the mid-sixteenth to mid-nineteenth century, England’s population rose sevenfold. Scholars debate the impact on the African population of the loss of 12 million people to the Americas.20 It is certain that this number of additional emigrants from a more heavily populated Europe over the same period would have had a negligible effect. Finally, the argument that Africans could survive the epidemiological environment of the Caribbean longer than could Europeans is irrelevant here. Whatever the European-African mortality differentials, European labor always traveled to the Caribbean sugar sector as free labor or servants, never as slaves. Medical evidence would be pertinent only if European slavery had been tried and found wanting or perhaps if European labor had not been tried at all. European demographic barriers to a supply of slaves were thus no greater than those for their African counterparts. In fact, nearly 1,000 convicts a year left Britain in the half-century after 1718. This may not seem like many compared to an African slave trade drawing 25,000 a year from Africa in the last third of the seventeenth century and rising to an average of 50,000 a year in the half-century after 1700. Yet consider that the population of England was only 7 percent of that of Europe in 1680, and if the rest of Europe had followed the English practice in proportion, 14,000 convicts would have been available. A traffic in degredados from Portuguese possessions to Brazil and Angola existed from the sixteenth to the nineteenth century, and the massive fortifications at Havana and San Juan, as well as Spanish outposts in North Africa, were built in part with Iberian convicts. Germanic states that lacked maritime facilities sold convicts to Italian city-states for galley service. In France, about 1,000 convicts a year arrived at Marseilles in the later seventeenth and early eighteenth centuries, but these were all male and mostly between twenty and thirty-five years of age. The potential for a large, more demographically representative traffic in French convicts is clear. Possible sources other than convicts were numerous. Prisoners from wars, as in Africa, could have provided many additional plantation laborers. Indeed, when in rebellion, the Irish and Scots alone could have filled the labor needs of the English colonies. Nor does the above speculation fully incorporate vagrants 20  For different assessments, compare Patrick Manning, Slavery and African Life: Occidental, Oriental, and African Slave Trades (Cambridge, 1990), 38–85; and David Eltis, Economic Growth and the Ending of the Transatlantic Slave Trade (New York, 1987), 64–71.

1166

Eltis

and the poor. A properly exploited system drawing on convicts, prisoners, and vagrants from all the countries of Europe could easily have provided 50,000 forced migrants a year without serious disruption to either international peace or the existing social institutions that generated and supervised these potential European victims.21 If such an outflow had been directed to the plantation colonies, it is also unlikely that mercantilist statesmen would have questioned either the scale or the direction of the flow. More specifically, British convicts could have replaced African slaves in the Chesapeake, the destination of most transport ships. The first recorded shipment of convicts dates to 1615. Until 1770, the number of convict arrivals was at least two-thirds that of slaves in total, although in the 1730s and 1740s slave arrivals were between two and four times larger. If all whites sent against their will to the colonies, and of course their progeny, too, had been accorded the slave status of African immigrants, there is no reason why the number of white slaves in Maryland and Virginia would not have been at least as large as the actual black slave population at the end of the colonial period.22 And to touch 21  Gregory King estimated 600,000 adults receiving alms and 30,000 vagrants in late seventeenth-century England. “Natural and Political Observations and Conclusions upon the State and Condition of England, 1696,” in Peter Laslett, comp., The Earliest Classics: John Graunt and Gregory King (Farnborough, 1973), 48, 57. For convicts leaving England, see Ekirch, Bound for America, 27. The most recent estimate of slaves leaving Africa at this period is David Richardson, “Slave Exports from West and West-Central Africa, 1700–1810: New Estimates of Volume and Distribution,” Journal of African History, 30 (January 1989): 1–22. 22  Peter Wilson Coldham, Emigrants in Chains: A Social History of Forced Emigration to the Americas, 1607–1776 (Baltimore, Md., 1992), 43–44; Abbot Emerson Smith, “The Transportation of Convicts to the American Colonies in the Seventeenth Century,” AHR, 39 (1933–34): 233–36; J. M. Beattie, Crime and the Courts in England, 1660–1800 (Princeton, N.J., 1986), 470–83. Perhaps 100,000 Africans arrived in the Chesapeake by 1770; Allan Kulikoff, Tobacco and Slaves: The Development of Southern Cultures in the Chesapeake, 1680–1800 (Chapel Hill, N.C., 1986), 65–67; and Galenson, White Servitude in Colonial America, 216–17. No firm estimates of convict arrivals before 1718 exist, but the number was less than the 50,000 that came to the Americas in the fifty-seven years after this date. Coldham has found records of 7,500 convicts transported before 1718 from England alone (counted from The Complete Book of Emigrants in Bondage, 1614–1775 [Baltimore, 1988]; compare Smith, “Transportation of Convicts,” 238, who counted a “minimum” of 4,431, 1655–99); Scotland, Wales, and particularly Ireland supplied many more. The estimate of rough equivalency between the descendants of convicts and the black slave population takes into account the earlier arrival of convicts and the fact that rates of natural population increase were a little higher for the white population than for the black. For those who would argue that such a system would have foundered on the difficulty of control,

Europeans And The Rise And Fall Of African Slavery

1167

briefly on a related area of interest, if one purpose of the tidewater planters in bringing in Africans was to control the poor, white, non-propertied classes, enslavement of at least part of that class would seem a much more direct way of achieving the goal.23 A cursory examination of potential factoring and distribution costs indicates a similar pattern of benefits in Europe relative to Africa. A factoring network of sorts actually existed. French and Spanish merchants formed the link between courts and galleys. Between 1718 and 1775, the British government contracted out the collection as well as the shipping of convicts to the Americas. Companies in London came to specialize in assembling cargoes of convicts including transfers from regional jails prior to embarkation.24 A network of other merchants in the Caribbean and mainland North America placed convicts, indentured servants, and often slaves, too, with masters. The system could surely have functioned just as well if the greater part of this labor had been enslaved Europeans rather than enslaved Africans. It was, of course, inconceivable that any of the labor pools mentioned above—convicts, prisoners of war, or vagrants—could have been converted into chattel slaves. The barrier to European slaves in the Americas lay not only beyond shipping and enslavement costs but beyond any strictly economic sphere. The English Vagrancy Act of 1547, which prescribed slavery for “vagabonds,” limited the period of servitude to two years. The act was itself repealed after only two years as unenforceable on account of its severity. Across the English Channel in the sixteenth and seventeenth centuries, slavery was considered sufficiently alien that slaves brought privately to France were freed on arrival (unless headed for the galleys), prior to a 1716 law that specifically exempted them from enfranchisement.25 In Spain, Portugal, and all Mediterranean countries by the sixteenth century, Moors and Africans could be slaves; Christians (which meant in practice Europeans, because non-Europeans who became Christian remained slaves) could not.26 Even Jews were less likely to be enslaved in Spain as of the later Middle Ages. In 1492, they were expelled, not enslaved. Some who fled to Portugal saw their children taken from we should note that this issue never seemed to have threatened the existing systems of long-distance penal servitude, many involving private traders and employers. 23  Edmund S. Morgan, American Slavery, American Freedom: The Ordeal of Colonial Virginia (New York, 1975). 24  Ekirch, Bound for America, 17–18; Coldham, Emigrants in Chains, 59–87. 25  C. S. L. Davies, “Slavery and Protector Somerset: The Vagrancy Act of 1547,” Economic History Review, 19 (December 1966): 533–49; William B. Cohen, The French Encounter with Africans: White Response to Blacks, 1530–1880 (Bloomington, Ind., 1980), 5, 44–46. 26  David Brion Davis, The Problem of Slavery in Western Culture (Ithaca, N.Y., 1966), 221–47.

1168

Eltis

them and sent to São Tomé but not as chattel slaves. Between the late fifteenth century and 1808, the Spanish Inquisition burned at the stake 32,000 conversos (Jewish converts to Christianity) for false conversion. Enslavement for this group was not an issue.27 To historians of Eastern Europe in the fifteenth and sixteenth centuries, a key question is the second serfdom, not the second enslavement. But before addressing the aversion to slavery in the face of clear economic imperatives, we need to assess how close to chattel slavery (and to the slave trade that supported it) Europeans were prepared to come within Europe and what differences existed between European nations. As already noted, similar devices for depriving people of liberty existed in Africa, the Americas, and Europe. Yet, if on the African and American continents, these might result in enslavement—and often ownership by Europeans on both sides of the Atlantic28—in Europe, the same procedures fell short of creating such a status, at least for Europeans. Throughout Europe, states could take the lives of individuals, but enslavement was no longer an alternative to death; rather, it had become a fate worse than death and, as such, was reserved for non-Europeans. Europeans would accept the making of lawbreakers and prisoners into slaves only if they were not fellow Europeans. Conceptions of the insider had expanded to include the European subcontinent, whereas for Africans and American Indians, a less than continent-wide definition of insider still pertained. In a profound but scarcely novel sense, chattel slavery for Africans and Indians in the Americas was thus a function of the non-slave status that Europeans considered appropriate for themselves—a situation with historical parallels in many slave societies. Put differently, if Africans or Indians instead of Europeans had initiated the plantation system, and had possessed the means to begin a slave trade with Europe, a trade in Europeans would not have been extensive. Slavery might have been just as prevalent in the Western world but not likely confined to people from another continent. Logically, therefore, those European states least likely to countenance the coercion of their own citizens may well have been among the more likely to develop a system of chattel slavery overseas. 27  David Brion Davis, Slavery and Human Progress (New York, 1984), 95; Haim Beinart, “The Conversos and Their Fate,” Elie Kedourie, ed., Spain and the Jews: The Sephardi Experience; 1492 and After (London, 1992), 92–122; Coates, “Exiles and Orphans,” 98. 28  For the use of West Africans and American Indian slaves in the French galleys, see Clarence J. Munford, The Black Ordeal of Slavery and Slave Trading in the French West Indies, 1625–1715, 3 vols. (Lewiston, N.Y., 1991), 1: 169; Paul W. Bamford, Fighting Ships and Prisons: The Mediterranean Galleys of France in the Age of Louis XIV (Minneapolis, Minn., 1973), 156, 165–66, 310–11.

Europeans And The Rise And Fall Of African Slavery

1169

While the rejection of European slavery was, like the acceptance of its African counterpart, an unthinking decision, some states came closer to imposing chattel slavery on their citizens than did others. Convict labor, especially that sent beyond the domestic borders of the country, would seem to be the closest that Europeans came to enslaving other Europeans. The merging or increasing cooperation between state and church courts and the movement away from strictly physical chastisement of criminals in early modern Western Europe triggered experiments with transportation and penal servitude. The French, Portuguese, Spanish, and English manned many of their early exploration voyages with convicts, and the French and Portuguese opened up their galley corps to convicted felons at about the time that the Portuguese began experimenting with transporting the same group to São Tomé, Goa, and Brazil. Most West European countries, including England, contemplated or experimented with both galleys and transportation for convicts. In the end, the Spanish and Italian states sent most convicts to galleys, the English favored transportation to distant colonies, the French and Portuguese promoted both, and the Dutch, except for fleeting experiments, did neither. Eventually, the French accepted felons for use in galleys and naval bases from other European states such as Savoy, Poland, and some German principalities. In Spain, France, and Portugal, convicts worked alongside slaves from the Mediterranean littoral, Eastern Europe, Africa, and the Americas.29 At the lowest common denominator, two critical features kept convict labor separate from chattel slavery. First, an individual usually had to commit a crime in order to become a convict. Set laws and judicial procedure ensured that potential convicts faced less arbitrary treatment than those who fell into slavery. The power of the state over the convict and the master over convict labor was more circumscribed than that of the slaveowner over the slave. Second, criminal status was not heritable. In both French and Spanish galleys, convicts had great difficulty winning release from the galleys whatever the term of their sentence, their property was sold upon conviction, and they had no standing in law. But Europeans, except for some adherents to the Greek Orthodox faith, were not technically slaves. Unlike Turks, Muslims, Russians, and Africans who 29  Coates, “Exiles and Orphans,” 32–67; Emile Campion, Etude sur la colonisation par les transportes anglais, russes et français (Rennes, 1901), 44–45; Beattie, Crime and the Courts in England, 450–519, esp. 470–74; John H. Langbein, Torture and the Law of Proof; Europe and England in the Ancien Régime (Chicago, 1977), 27–44; Zysberg, “Galley and Hard Labor Convicts in France,” 78–110; Pike, Penal Servitude in Early Modern Spain, 3–26; Pieter Spierenburg, The Prison Experience: Disciplinary Institutions and Their Inmates in Early Modern Europe (New Brunswick, N.J., 1991), 259–62.

1170

Eltis

had been specially purchased to supplement the galley corps, they could not be resold, and their status was not transmitted to their progeny. It is striking that the more strenuous work—specifically, the farthest inboard position on the oars in galleys—was reserved for slaves, not convicts. And, when the useful life of oarsmen came to an end and the state wished to reduce expenses, old slaves were either returned to their state of origin on the basis of treaties, sold at European slave marts, or, if African, sometimes taken to the French West Indies to be sold. Convicts who managed to obtain a discharge from the French galley corps also ended up in the French Americas, but they were always sold as indentured servants. Portuguese degredados could be paid by the state for their labor and in São Tomé and Angola had sufficiently different status from slaves that most of them were involved in the slave trade that supplied São Tomé and Brazilian plantations.30 The most that can be said by way of comparison is that the spread of African slavery in the Americas coincided with the spread of forced labor in punishment systems within Europe (transportation in the English case). But no one was in any doubt about the distinctions between the two. Nevertheless, interesting differences between states emerge in the treatment of convicts. Penal servitude became a feature of the Continental European systems—effectively for life in the case of the French and Spanish. The English imposed neither lifelong banishment nor much penal servitude; the Bridewell houses held few long-term prisoners and did not always have a labor regime.31 In the Netherlands, penal servitude supplemented, then replaced, banishment, although the Dutch did hold long-term prisoners who had been incriminated, as opposed to convicted. They also used torture after that practice was discontinued in England, where in fact it had never been part of

30  Coates, “Exiles and Orphans,” 57, 98–99; Bamford, Fighting Ships and Prisons, 139–52, 250–71; Pike, Penal Servitude in Early Modern Spain, 8–14. The term “slavery” is occasionally used loosely in the penology literature. J. Thorsten Sellin in his widely read Slavery and the Penal System (New York, 1976) uses the term to cover any form of coerced labor. Spierenburg, Prison Experience, refers ambivalently to the “bondage” of convicts throughout his book (see especially page 10). Such usage skates over distinctions important to the issue of slavery in the Americas. 31  Spierenburg, Prison Experience, 1–68, 259–76; Zysberg, “La société des galeriens,” 43–75; Bamford, Fighting Ships and Prisons, 173–83; Coates, “Exiles and Orphans,” 71–72. In Spain, sentences for forzados were limited to ten years in 1653, although, as noted above, detention beyond the formal term was common (Pike, Penal Servitude in Early Modern Spain, 7–8).

Europeans And The Rise And Fall Of African Slavery

1171

the Common Law.32 For the English, transportation meant shipment across the Atlantic with no right of return for seven or, for those reprieved from death sentences, fourteen years. Once in the Americas, the vast majority of prisoners were sold into seven-year terms of servitude, but the sentence itself did not provide for this.33 Prisoners with resources could in fact buy their freedom at the point of disembarkation and, as long as they stayed away from England, convert their sentence to temporary banishment. The servitude that accompanied transportation was actually a device to pay its costs and, apart from a longer term, did not differ materially from that voluntarily experienced by non-convict indentured servants.34 In Portugal, where the state bore the cost of transportation or imposed it on others, the sentence constituted exile, with the labor required varying according to the needs of the state. The system was remarkably flexible in alternating between galleys and transportation to the colonies, with or without forced labor.35 Vagrants formed a larger group than convicts in every European country, and the destitute poor, dependent on state aid, a larger group again. From the mercantilist standpoint, the productive potential of these groups in a slave society was enormous. With the increasing breakdown of church-sponsored charity systems from the mid-sixteenth century on, workhouses mushroomed in northwestern Europe, especially England, Holland, and France—three of the leading early plantation powers.36 Penal servitude for felons was grafted onto this system in the eighteenth century.37 Hardening attitudes toward the poor were seen throughout Western Europe in the early modern period, although the exaction of labor in return for relief appeared first in Protestant areas. The Dutch system of workhouses was the most efficient and became a model for the rest of Europe. It was greatly admired, though never wholly emulated by

32  Spierenburg, Prison Experience, 41–60, 143–47; P. Spierenburg, “The Sociogenesis of Confinement and Its Development in Early Modern Europe,” Centrum voor Maatschappij Geschiedenis, 12 (1984): 38. For the ending of torture, see Langbein, Torture and the Law of Proof 10–12, 50, 134–35. 33  Ekirch, Bound for America, 16–21. 34  The longer term, equivalent to a lower price, might have reflected the greater likelihood of convicts escaping, being dishonest, a more diverse mix of ages and sexes than was the case with regular indentures, or some combination of these factors. Formal analysis has yet to be carried out on convict labor. 35  Coates, “Exiles and Orphans,” chap. 5; Boxer, Golden Age of Brazil, 140. 36  Spierenburg, “Sociogenesis of Confinement,” 9–77, esp. 31–32. 37  Spierenburg, Prison Experience, 12–86; Beattie, Crime and the Courts in England, 492–500.

1172

Eltis

the English elite, and it was also the most generous.38 The English subjected their vagrants to compulsory labor on pain of whipping and imprisonment, expected their poor to work, and sent indigent children, forcibly separated from their parents in earlier times, to masters at home and in the colonies—until the twentieth century, if we include those sent from the Dr. Barnardo homes for orphaned and destitute children. Yet the inefficiency of local government was such that the numbers caught in the system were likely modest.39 Centralized administrations in Spain, however, ensured the impressment of vagrants and gypsies into state arsenals (naval dockyards and armament factories) as late as the second half of the eighteenth century. In France, 7 percent of the galley corps at the time of its merger with the navy in 1748 were vagrants, but, as the needs of the galleys declined after 1700, this group was more likely to be sent to the West Indies.40 Overall in Europe, those sent to the Americas or put in the galleys formed a small share of the poor receiving relief; the fate of those sent to the Americas was invariably indentured servitude, not slavery; and the poor who could not avoid the state system were probably treated less harshly in the Netherlands and England (though for different reasons taken up below) than in France and Spain. Prisoners taken in the course of European military action may be divided into two groups for present purposes. Those captured in the act of rebellion against established government could expect death if they were leaders, or banishment if they were deemed followers, but never enslavement. The English sold many Irish in Barbados during the seventeenth century after military campaigns in Ireland but always as indentured servants with a maximum term of ten years. In France in 1662, nearly five hundred rebellious Boulonnais were sent to the galleys, not into slavery, despite the inability of French merchants at the time to sustain an African slave trade to the French Americas. A second group consists of those captured in conflicts between states. At least 38  Spierenburg, “Sociogenesis of Confinement,” 24; Simon Schama, The Embarrassment of Riches: An Interpretation of Dutch Culture in the Golden Age (New York, 1987), 174–75, 570–79. 39  Dorothy Marshall, “The Old Poor Law, 1662–1795,” Economic History Review, 1 (April 1937): 39; George Meriton, A Guide for Constables, Churchwardens, Overseers of the Poor … (London, 1669); Statutes at Large, 43 Eliz. c. 2. Smith, “Transportation of Convicts,” 244, indicates vagrants among a group going to Virginia in the 1660s. 40  Pike, Penal Servitude in Early Modern Spain, 67–69; Zysburg, “La société des galeriens”; Bamford, Fighting Ships, 260; Christian Huetz de Lemps, “Indentured Servants Bound for the French Antilles and Canada in the Seventeenth and Eighteenth Centuries,” in Ida Altman and James Horn, eds., “To Make America”: European Emigration in the Early Modern Period (Berkeley, Calif., 1991), 188–89.

Europeans And The Rise And Fall Of African Slavery

1173

until the French Revolutionary wars, prisoners of war were the best treated of all the European groups considered here. Detention followed by prisoner exchanges or ransoming was common. English, Dutch, and French prisoners could be consigned to the galleys in Spain, but slave status there was reserved for non-Christians. Even the latter group, however—in the case of North Africans enslaved in French and Spanish galleys—had some prospect of release in exchange for Christians held by rulers of Algiers, Tunis, and other Mediterranean Muslim powers.41 Prisoner exchanges seem to have been less common in Africa and the Americas, where captives were more likely to be killed, absorbed into the captor’s society, or traded away. Some might see the trading of captives as the distorting impact of a European-generated market for slaves, but the key point is that this option was also open to commanders of European armies and they did not use it. Finally, we might note the large differences in transatlantic shipping conditions for Europeans and for Africans. Convicts and soldiers who had been held in confined circumstances prior to embarkation could suffer as high a shipboard mortality on the Europe-America route as slaves carried from Africa to the Americas.42 There is, however, no evidence that any European transatlantic voyagers, whether steerage passengers, convicts, soldiers, or indentured servants, were ever subjected to conditions that were the norm on slave ships. Thus English and Irish alike traveled in convict ships that rarely carried in excess of 150 prisoners and on average fewer than 100. Far fewer persons-per-ton were carried on ships dedicated to convict transportation than on slave ships. 41  Alfred Vagts, A History of Militarism: Civilian and Military (New York, 1959), 113–14; Bamford, Fighting Ships and Prisons, 176–77, 261–64. Bruno S. Frey and Heinz Buhofer (in an article drawn to my attention by Stanley L. Engerman) argue that the relatively good treatment of prisoners of war came from their market value and, by implication, that the European-North African differential arose from a relative lack of interest on the part of the North African powers in buying back low status prisoners (“A Market for Men, or There Is No Such Thing as a Free Lynch,” Journal of Institutional and Theoretical Economics, 142 [1986]: 739–44). This approach can explain much, but the interesting question is why, if, for instance, neither the English nor the Algerians would buy back all their prisoners from the Spanish in the late sixteenth century, and both groups were sent to the galleys because of their lack of value on the exchange market, only the Algerians were enslaved. Likewise, to take one of their own examples, the need for labor can explain fluctuations in the treatment of death camp prisoners over the course of World War II in Germany but not the existence of the death camps themselves. 42  See Raymond L. Cohn, “Maritime Mortality in the Eighteenth and Nineteenth Centuries: A Survey,” International Journal of Maritime History, 1 (June 1989): 159–91, and the literature cited there.

1174

Eltis

Crowding was three or four times as severe on the slave ships from the seventeenth to the early nineteenth century, and only galley ships, rarely out of sight of land and never at sea in bad weather, matched the human density of a slave ship. While this difference may not have had much impact on mortality, which was not particularly affected by crowding, it may well say something about the views European merchants, and ultimately European societies, held regarding the status of different migrant groups. States, merchants, and consumers of plantation produce all stood to gain from shipping convicts in slave-like conditions.43 An almost tangible barrier thus prevented Europeans from becoming chattel slaves unless they were captured by non-Europeans. In many African societies, shared ethnicity and language might even mean an increased likelihood of enslavement—at least for women and children—given the focus on kin groups and their expansion through absorption of outsiders. In the Americas, slave raids of one Iroquois nation on another were not uncommon. But in Europe, even the most degraded member of society was spared enslavement. Only in the rather limited case of Amerindians in the Spanish Americas was such treatment extended to non-Europeans. This barrier was somewhat akin to the Muslim bar against the enslavement of non-Muslims—not in the sense that the basis of enslavement was religious but rather that in both Muslim and Christian societies slavery came to be mainly African despite the fact that in both, slaves often converted to the faith of their owners.44 But if all Europeans shared this attitude toward the insider, it is also clear from the above that some societies came closer than others to countenancing conditions that approached slavery for their own citizens. English indentured servants and convicts on average served less than seven years. Effective life sentences, though sometimes for exile without labor, were common for 43  The earliest Naval Office lists are in PRO, CO 33/13 and CO 142/13. Beckles, White Servitude and Black Slavery, 62–67; Ekirch, Bound to America, 98–100; John Mcdonald and Ralph Shlomowitz, “Mortality on Convict Voyages to Australia, 1788–1868,” Social Science History, 13 (Fall 1989): 285–313. For slave ships, see Eltis, Economic Growth and the Ending of the Transatlantic Slave Trade, 135–38, and “Productivity in the Slave Trade: A Comparative Assessment” (unpublished ms., 1993). The ships that carried the Monmouth prisoners to Barbados in 1685–1686—that is, those men who were not hanged, drawn, and quartered—averaged a density of 0.72 prisoners per ton and reported less than 10 percent mortality during the voyage. Slave ships at this time embarked 2.5 persons per ton and averaged a 20 percent loss. The higher mortality on slave ships is likely explained by the greater length of the slave voyage rather than greater crowding among those on board. 44  Bernard Lewis, Race and Slavery in the Middle East: An Historical Enquiry (New York, 1990), 5–15, 54–61; Davis, Slavery and Human Progress, 32–51.

Europeans And The Rise And Fall Of African Slavery

1175

Spanish and French convicts, and at least some French engagés were sent to the Americas by the state with a status little different from convict labor. Likewise, slave labor involving non-Europeans was much more common in southern Europe than in the north. The slave markets in Europe from the Middle Ages until to the eighteenth century were on the shores of the Mediterranean, not northwestern Europe. Given the large African slave population in Lisbon, it is hard to visualize a Portuguese counterpart to an incident in Middelburg in 1596 in which 130 African slaves were restored “their natural liberty” on the grounds that slavery did not exist in Zeeland.45 Nor is it easy to imagine equivalent action in any part of the non-European world—for example, North Africa— if the slaves had been European. Yet there are abundant instances of ethnocentric attitudes in sixteenth-century Holland, France, and England. Moreover, the greater the inability of the European to think in terms of slavery for other Europeans, the more he or she was likely to contemplate coerced labor for non-Europeans. In a further paradox, it is also possible that the more often coercion is seen to be unconscionable for people like oneself and appropriate for others, the more likely that coercion for anyone will eventually be questioned. There are two issues to be addressed next. One is the genesis of what from a transatlantic, if not global, perspective is the rather exceptional European attitude toward slavery in the early modern period—specifically, its association with the concept of race. The second is the intra-European differences in attitudes toward coerced labor that evolved and persisted into the era of abolition. Indeed, those countries that had developed the strongest antipathy to curtailment of liberty for their own people not only had the least inhibitions about creating an overseas labor system using non-European slaves but were eventually to be in the vanguard of the movement to abolish slavery worldwide. State-enforced abolition of slavery, as well as the conception of a world without slavery, originated first in Europe.46 These issues raise large questions, and the answers must be somewhat speculative. These speculations incline not so much toward chauvinism or ethnocentrism—at least, not at root—but

45  A. C. de C. M. Saunders, A Social History of Black Slaves and Freedmen in Portugal, 1441–1555 (Cambridge, 1982), 59; W. S. Unger, “Bijdragen tot de geschiedenis van de Nederlandse slaven-handel,” Economisch-Historisch Jaarboek, 26 (1956): 136. For this last reference and translation, I thank Pieter C. Emmer. For a similar incident in Bordeaux in 1571 paralleling that in Middelburg, see Charles de La Roncière, Nègres et Négriers (Paris, 1933), 15–16. 46  Stanley L. Engerman, “Coerced and Free Labor: Property Rights and the Development of the Labor Force,” Explorations in Economic History, 29 (1992): 3.

1176

Eltis

rather toward longer-term economic and ecological factors that separated Western Europe from the rest of the world before the modern era. As already noted, the difference between Europe and most of the rest of the world on the slavery issue in the early modern period was the conviction that only non-Europeans could be enslaved. Technology gave Europeans the power to sail out and impose this view of the world on others. If we follow Eric Jones’s argument, the European lead in technology was a function of the political system first and, behind that, a rather exceptional resource base and set of ecological circumstances.47 The major difference between Europe and the rest of the world at about the time that Europe began to enslave others was the diffusion of political power, both between and within European states. According to the early version of Jones’s thesis, this diffusion of power curtailed wrong-headed decision making by centralized authorities and encouraged competition between states. It also allowed economies of scale arising from the free movement of resources and the maintenance of a shared market and knowledge pool large enough to foster the possibility of a slowly rising per capita income. This tendency was further encouraged by the early evolution of the nuclear family structure and a strategy of stressing goods over children, both of which were rooted in environmental characteristics peculiar to Europe. In the long run, these social and economic patterns were hostile to enslavement of Europeans, and eventually to all slavery, in two ways. One was their tendency to stress the value of the individual, at least compared to the nonEuropean world. The system of competitive nation-states in an overall balance of power relationship with each other was complemented within each state by an implicit and unusual contract between ruler and ruled. In return, for a large share of the peasants’ small productive surplus, rulers suppressed random violence and provided stable legal institutions. The potential for individual rights was altogether greater in such a system than in a highly centralized empire or in regions where decentralized political power was associated with very low population densities. Conceptions of the individual as owner of him or herself, owing nothing to society, and of society as a series of “relations of exchange

47  Eric L. Jones, The European Miracle: Environments, Economies, and Geopolitics in the History of Europe and Asia (Cambridge, 1981). The major criticisms of this work, which the author has attempted to address in subsequent publications, hinge on questions of individual motivation and the universality of “economic man” rather than his description of how Europe differed from the rest of the world.

Europeans And The Rise And Fall Of African Slavery

1177

between proprietors,” required restraints on political power as well as the encouragement provided by an evolving market society.48 A second and perhaps more important implication of the comparative patterns that Jones describes was the opportunity they provided for the pervasive growth of market behavior. The trading of land, labor, and capital, as well as goods and services, was far more extensive in early modern Europe than in any other continent. The ramifications for slavery of an extensive market system are ambivalent and not easily summarized in an essay. Slavery was the commodification of human capital carried to extremes and might reinforce as well as draw strength from the spread of markets. But, in a series of controversial essays on the link between capitalism and abolition, Thomas Haskell has argued that, under certain conditions, markets work against slavery and the infliction of cruelty in general. Haskell has listed preconditions for the historical emergence of humanitarianism as, first, an ethical maxim that makes the alleviation of suffering “right,” second, a sense of being causally involved in the situation that gives rise to the suffering, and, lastly, possession of a recipe for intervention so easily applied that refusal to employ it might be considered “an intentional act in itself.” The market system peculiar to the late eighteenth-century North Atlantic world “expanded the range of causal perception and inspired people’s confidence in their power to intervene.” The growth and elaboration of this knowledge incorporates, but is much broader than, technical and scientific learning. For Haskell, it is the link between causal perception and moral responsibility that is important. By the late eighteenth century, “recipe knowledge and causal perception,” specifically one assumes in long-distance trade, had become so familiar that in the minds of some participants in the North Atlantic economy—notably Quakers—action against slavery (and other abuses) had become a moral imperative.49

48  C. B. Macpherson, The Political Theory of Possessive Individualism: Hobbes to Locke (Oxford, 1962), esp. 46–70. The quote is from page 3. For more specific implications of possessive individualism for free and coerced labor, see Engerman, “Coerced and Free Labor,” 1–29. 49  Thomas L. Haskell, “Capitalism and the Origins of the Humanitarian Sensibility, Part 1,” and “Capitalism and the Origins of the Humanitarian Sensibility, Part 2,” in Thomas Bender, ed., The Antislavery Debate: Capitalism and Abolitionism as a Problem in Historical Interpretation (Berkeley, Calif., 1992), 107–60, esp. 129–33 and 147–51. These essays were responses to the work of David Brion Davis and first appeared in the American Historical Review in April–June 1985.

1178

Eltis

What interests us here, however, are not the implications of the above argument for the abolition of slavery but its implications for the origins of the system. Before 1750, these same traits, again as embodied in market behavior, could have had almost exactly the opposite effect. Until the “recipe knowledge” had become so familiar that not to act became “an intentional act in itself,” such knowledge would serve to increase its possessor’s power to intervene in the lives of others without generating any feeling of moral responsibility for the suffering that might result. Causal perceptions would lead to moral restrictions on enslaving one’s weaker and poorer neighbor (or more broadly, perhaps, one’s fellow European) but not to the purchase of a non-European who was already a slave in the Mediterranean, Africa, Asia, or the Americas. The organizational and technical skills of business, and the ability to develop long-term projects around scientific knowledge and bring them to fruition, could thus disrupt or destroy societies in distant lands without eliciting anything other than indifference on the part of the perpetrators. Dutch slaving activities in the East Indies in the early seventeenth century and English and Dutch use of Africans in the Americas had a larger impact on Asia and Africa than their Iberian counterparts of the previous century but generated no more self-questioning.50 Familiarity with “recipe knowledge” would first cross the threshold of action in the domestic rather than the colonial environment and be applied to “insiders” before “outsiders.” An increasing regard for the consequences of one’s actions at home was thus not inconsistent with continued indifference toward, and indeed encouragement of, slavery overseas.51 Moreover, countries in which the market was more pervasive, such as the Netherlands and England, might tend to have greater respect for individual rights and more humane treatment of convicts, for instance, than would other countries with a slightly less invasive experience of market institutions. By the mid-seventeenth century, it was widely believed in England, the Netherlands, and France that no one entering 50  J. Fox, “‘Tor Good and Sufficient Reasons’: An Examination of Early Dutch East India Company Ordinances on Slaves and Slavery,” Anthony Reid, ed., Slavery, Bondage, and Dependency in South-East Asia (St. Lucia, Queensland, 1983), 246–62. Fox observes, “The Dutch East India Company embodied new organizational principles whose introduction into Asia had profound effects … the Company was the first formal slave holding corporation of its kind in Asia” (p. 248). Together with their English counterparts, the Dutch West India Companies probably doubled the volume of the transatlantic slave trade in the mid-seventeenth century. 51  Abolition of colonial slavery may have been the first of the humanitarian reforms of the modern era, but the basic rights of individuals under the law in northwestern Europe were established before the creation of transatlantic European slave colonies.

Europeans And The Rise And Fall Of African Slavery

1179

those countries could remain a slave. Curiously, it was thought in England that this had been the case since the advent of Christianity.52 As noted above, the French altered their laws so that those returning from the Caribbean could bring personal slaves with them, but there was no such enabling legislation in the Netherlands or England. It is therefore ironic and yet not surprising that the English and the Dutch developed the most absolute form of chattel slavery and the harshest laws against free blacks in the Americas. The ubiquity of the market and recognition of individual rights were characteristics of all of early modern Western Europe viewed from a global perspective. But these conditions evolved furthest in the Netherlands and England. Three interrelated characteristics of agriculture and family life in those countries supported such conceptions of the individual. (All were driven by and in turn reinforced market penetration of the two societies.) One was the early emergence of the nuclear family and the associated phenomena of late marriage and early entry of children into service. A second was the non-feudal nature of land ownership, which facilitated the emergence of a well-integrated market for labor. A third was the related advancements in food production that permitted the support of a larger non-agricultural population.53 These were the foundations of relative political and economic freedom (not to be confused with freedom from want) and limits on the reduction of Europeans to slave or near-slave status.

52  A handbook on England partly written for and widely used by foreigners was Edward Chamberlayne, Angliae Notitia, or, The Present State of England…: Together with Divers Reflections upon the Antient State Thereof, 6th edn. (London, 1672). It stated, “Foreign slaves in England are none, since Christianity prevailed. A foreign Slave brought into England, is upon landing ipso facto free from slavery, but not from ordinary service” (p. 331). See also Dudley North, Observations and Advices Œconomical (London, 1669), 45. For the status of blacks in England prior to the Somerset case (late eighteenth century), see Seymour Drescher, Capitalism and Antislavery: British Mobilization in Comparative Perspective (London, 1986), 25–49. 53  Alan Macfarlane, The Origins of English Individualism: The Family, Property and Social Transition (Oxford, 1978); Jan De Vries, The Dutch Rural Economy in the Golden Age, 1500–1700 (New Haven, Conn., 1974), 107–73; De Vries, European Urbanization, 1500–1800 (Cambridge, Mass., 1984), 116–18, 210–12; G. E. Fussell, “Low Countries’ Influence on English Farming,” English Historical Review, 74 (October 1959): 611–22. By the end of the eighteenth century, agricultural output per hectare in the Netherlands and England were much the highest in Europe; Paul Bairoch, “Les trois révolutions agricoles du monde développe: Rendements et productivité de 1800 à 1985,” Annales: Economies, Sociétés, Civilisations, 44 (March–April 1989): 318.

1180

Eltis

Beyond these common factors, there were important differences. In England, the notion that the privileges of the individual are ultimately of greater importance than those of the state or any group was probably more highly developed than anywhere in Europe.54 Dutch attitudes derived not so much from concern for the individual as from a sense of the fragility of the social compact at a time when the country contained many resident non-Dutch and non-Calvinists. The greater prosperity of the Netherlands attracted large numbers of immigrants, but, despite illiberal edicts issued at the behest of the Calvinist church, the Dutch developed a practical tolerance of others unrivaled in seventeenthcentury Europe.55 It would be tempting to see in this a capacity for absorbing “outsiders” in the sense used above, but very few of these migrants were in fact non-Europeans. Far more non-Europeans lived in England, where their legal status was clarified only in the late eighteenth century.56 The foregoing paragraph suggests the impossibility of using convicts as a source of permanent colonial labor in England or the Netherlands. Penal transportation did not exist among the Dutch. In the case of England, scholars have until recently disagreed in their interpretations of criminal sanctions. They all nowadays seem to accept the view that judges and juries were reluctant to apply the letter of the more ferocious laws, even though scholars are not yet in accord on the reasons. If, for example, judges and juries were slow to enforce branding on the cheek for felons at the outset of the eighteenth century, it is doubtful that they would have sent convicts to chattel slavery in the Americas. If transportation had meant a lifetime of servitude for English people, juries might not have convicted anyone at all.57 This reluctance may be understood as a function of shared community values or of fears about the resistance that would have inevitably followed. But what seems incontestable is that, in regard to slavery, the sense of the appropriate was shared across social divisions and cannot easily be explained by ideological differences or power relationships between classes. Outrage at the 54  Macpherson, Possessive Individualism, 263–71. See Macfarlane, Origins of English Individualism, 165–88, on the differences between early modern England and Continental Europe. Criticisms of his attempt to carry the comparison back into medieval times have tended to dominate evaluations of this book. 55   E. H. Kossmann, “Freedom in Seventeenth-Century Dutch Thought and Practice,” Jonathan I. Israel, ed., The Anglo-Dutch Moment: Essays on the Glorious Revolution and Its World Impact (Cambridge, 1991), 281–98. 56  Drescher, Capitalism and Antislavery, 25–49. 57  See, for example, Beattie, Crime and the Courts in England, 450–519, esp. 490–92; Peter Linebaugh, The London Hanged: Crime and Civil Society in the Eighteenth Century (Cambridge, 1992).

Europeans And The Rise And Fall Of African Slavery

1181

treatment of Africans was rarely expressed at any level of society before the late eighteenth century. The moral economy of the English crowd and the various Christian churches were preoccupied with other issues. When the immorality of coerced labor was finally recognized, the recognition appeared among all social groups at about the same time. And attempts to account for the failure of Europeans to enslave their own in terms of solidarity among the potential slaves do not seem promising.58 If the elite could kill Irish, Huguenots, Jews, prisoners of war, convicts, and many other marginalized groups, why could they not enslave them? The English considered those from the Celtic fringe very different from themselves but not different enough to enslave.59 For elite and non-elite alike, enslavement remained a fate for which only nonEuropeans were qualified. The central development shaping Western plantation slavery from the sixteenth century onward was the extension of European attitudes to the nonEuropean world. If, by the sixteenth century, it had become unacceptable for Europeans to enslave other Europeans, by the end of the nineteenth century, it was unacceptable to enslave anyone. Put in relative perspective, before the eighteenth century, Europeans, in common with most peoples in the world, were unable to include those beyond the oceans in their conception of the social contract.60 Unlike most other peoples in the world, Europeans had the power to impose their own version of that contract on others, which for three centuries meant African slavery.61 58  “Popular revulsion to bondage and untrammelled private power,” which had “long preceded … critiques of colonial slavery,” is one of three sources of anti-slavery sentiment singled out by Blackburn, Overthrow of Colonial Slavery, 36. He and Drescher, Capitalism and Antislavery, see popular involvement as vital to abolition. Like the churches, the intellectuals, and the political elite, however, those without property clearly had other priorities before the later eighteenth century. Attempts to enslave Europeans and put them to work on sugar plantations would perhaps have attracted their attention, but were plantation owners, metropolitan merchants, and mercantilist statesmen so apprehensive of popular opinion that not a single one of them dared advocate European slavery in public? A more convincing explanation of the silence, and inaction, is simply a conception, shared by all classes, of what was appropriate. 59  George M. Fredrickson, White Supremacy: A Comparative Study in American and South African History (New York, 1981), 15. 60  Drescher, Capitalism and Antislavery, 1–24. There is a striking complementarity between the shift in English perceptions outlined by Drescher and the market-driven emergence of humanitarianism that Haskell argues for in “Capitalism and the Origins of the Humanitarian Sensibility.” 61  A somewhat attenuated version of the same process might be seen in Japan, the nonEuropean society with the most “Western” family and social structure before the

1182

Eltis

The first step toward abolition of slavery generally may have been the idea that the enslavement of Europeans anywhere was a wrong that needed to be righted. In sixteenth-century Spain and Portugal, religious orders dedicated to the redemption of Spanish and Portuguese captives held as slaves in North Africa raised huge sums privately for that purpose, and the state was heavily involved by the end of the century. This activity and the funds behind it were far more significant than similar efforts made during the Middle Ages. They were much greater, too, than equivalent efforts in North Africa to redeem Muslim captives in Iberia; Muslim exchanges for Christians did not begin on a large scale until after 1750.62 Farther north, almost every coastal town in the Netherlands had a “slave fund” for redeeming Dutch sailors from the galleys of the Barbary States.63 European seafaring states signed a series of treaties with North African powers and the Ottoman Turks to safeguard ships and crews from capture and enslavement. Most involved the issuing of safe-conduct passes to merchant ships.64 The irony that some of the main beneficiaries of such arrangements were Dutch and English slave traders on their way to Africa appears to have escaped historians and contemporaries, among them the earl of Inchquin, who was once held captive in Algiers before becoming governor of Jamaica. When the passes proved ineffective and seamen were captured and enslaved anyway, petitions to the British government seeking their release demanded action in the cause of “Christian charity and humanity”—long before abolitionists began to invoke similar principles.65 Indeed, the Lady Mico Trust, twentieth century. Slaves in Japan were overwhelmingly Japanese—drawn from criminals and the poor—before the modern period; Patterson, Slavery and Social Death, 127. In the nineteenth and twentieth centuries, the Japanese increasingly drew on foreign sources for slaves, although the numbers were much lower than those carried across the Atlantic. Both before and after Japan imposed the abolition of slavery on Korea in 1910, Korea had come to be the major source of forced laborers taken to Japan. The absorption of foreigners into the Japanese view of the social contract—a process that some would argue is not yet complete—might account for the gradual disappearance of coercion of foreigners, even though in this case there are extraneous geo-political developments to be reckoned with in the form of world wars. Mikiso Hane, Peasants, Rebels, and Outcasts: The Underside of Modern Japan (New York, 1982), 236–37. 62  Ellen G. Friedman, Spanish Captives in North Africa in the Early Modern Age (Madison, Wis., 1983), 105–28, 157. 63  Personal communication from Pieter C. Emmer, December 8, 1992. 64  David Richardson, The Mediterranean Passes in the Public Records Office, London (East Ardsley, 1981). Among the first duties of ships commissioned by the early American Republic at the end of the eighteenth century was protection of U.S. shipping against the Barbary pirates. 65  PRO, James Kirkwood to Secretary of State, June 6, 1709, CO 388/12, K25.

Europeans And The Rise And Fall Of African Slavery

1183

used in the nineteenth century to fund education in the British West Indies in the aftermath of slavery, was first established in 1670 to redeem Christians from North Africa. Like Samuel Pepys and presumably the earl of Inchquin, the early Mico trustees could see the suffering of Europeans in the Barbary States but not of Africans in America.66 The relevant question is, at what point did Europeans in large enough numbers to make a difference begin to understand that “Christian charity and humanity” should also encompass those of African descent? The gradual removal of the barriers that kept non-Europeans from insider status was a very slow process and in some respects has never been completed.67 As noted above, it was extended to Native Americans before Africans and perhaps to non-Europeans living in Europe before their overseas counterparts. The widening of this insider status can be traced in intellectual developments. While no major thinker after Locke was able to defend the practice of enslaving those outside the social contract, the ambivalence of the Enlightenment on the issue of slavery for non-Europeans is evident.68 It can also be traced at the level of personal experience in both the AngloSaxon and the Hispanic Atlantic, where proximity to the exploitation could, for a minority, breed sensitivity as well as indifference.69 It is possible that among the European working class, the first abolitionists, or at least those who were prepared to identify with blacks, were seamen, rather than industrial

66  Frank Cundall, The Mico College, Jamaica (Kingston, 1914), 7–9. Samuel Pepys’ comments on the European slaves held by the Barbary States are in his diaries dated January 8 and November 28, 1661. 67  Philip D. Morgan argues that white attitudes toward blacks hardened during the colonial period. He does not take up the emancipation issue, which arose in the northern states during and after the revolution, or attitudes toward blacks in the abolition movement in the wider Atlantic region during the eighteenth century. “British Encounters with Africans and African-Americans, circa 1600–1780,” in Bernard Bailyn and Philip D. Morgan, eds., Strangers within the Realm: Cultural Margins of the First British Empire (Chapel Hill, N.C., 1991), 157–219. 68  Davis, Problem of Slavery in Western Culture, 137–39, 319–479; Davis, Slavery and Human Progress, 107–16, 154–58. Also see the discussion of Montesquieu in Seymour Drescher, “Capitalism and Abolition: Values and Forces in Britain, 1783–1814,” in Roger Anstey and P. E. H. Hair, eds., Liverpool, the African Slave Trade, and Abolition (Liverpool, 1976), 182. It would seem that for Locke the universe of the social contract was not the state but rather the European nation-state. 69  John Newton and Alexander Falconbridge are among the better-known slave ship officers who turned against the trade.

1184

Eltis

artisans.70 Certainly, the European country with the largest and most successful slave trade and slave colonies in both relative and absolute terms was the center of the strongest abolition movement and the leader of world abolitionism. British subjects, moreover, had an extraordinary proclivity to migrate, most of them before 1800 going to plantation regions that formed the most integrated of all eighteenth-century colonial systems. As early as 1700, “communication and community” across the British Atlantic had attained a depth, richness, and reliability of contact unrivaled among European powers and unprecedented in the history of long-distance migration.71 Such characteristics have clear implications for metropolitan awareness of events “beyond the line.” The Netherlands, by contrast, despite its early commitment to social and economic freedom, was among the last countries to take action and did not experience a mass antislavery movement of any kind. Also by contrast, it was without a slave-trade fleet after the 1780s, and the slave sectors of its empire were not only small but of declining significance from the late eighteenth century. More important, in the two centuries before 1800, the Dutch experienced net immigration of half a million people, almost all of them from other parts of northwestern Europe. Perhaps a mere 30,000 people left the Netherlands for the Americas and Africa on a net basis, and many of these were not Dutch. The English over the same period had a net emigration of 1.25 million, almost all of whom settled around the Atlantic basin. Dutch transatlantic networks and contacts with non-Europeans were simply many times less dense and frequent. Even if we include the Dutch East Indies, communication between England and the transoceanic world—and, accordingly, daily awareness of the world outside Europe—must have dwarfed that of the Netherlands.72 This is 70  Peter Linebaugh and Marcus Rediker, “The Many Headed Hydra: Sailors, Slaves and the Atlantic Working Class in the Eighteenth Century,” in Colin Howell and Richard J. Twomey, eds., Jack Tar in History: Essays in the History of Maritime Life and Labour (Fredericton, N. B., 1991), 11–36. For the wide range of attitudes of Europeans and Africans toward each other, see Morgan, “British Encounters with Africans and African-Americans,” 157–219. 71  Ian K. Steele, The English Atlantic, 1675–1740: An Exploration of Communication and Community (New York, 1986). 72  Jan Lucassen, Dutch Long Distance Migration: A Concise History, 1600–1900 (International Institute for Social History, IISH research paper, Amsterdam, 1991), 20–23, 35–42; and E. A. Wrigley and R. S. Schofield, The Population History of England, 1541–1871: A Reconstruction (London, 1981), 528–29. Probably fewer than 1,000 Dutch individuals a year returned from the East Indies, most of them sailors and soldiers. Most of those who did not return died (Lucassen, Dutch Long Distance Migration, 41). For the New Netherlands, Seymour Drescher points to a lack of support networks and leaders who wished to reproduce Dutch culture in the process of pursuing the issue of why the Dutch were such tardy

Europeans And The Rise And Fall Of African Slavery

1185

not, however, the place to reassess the origins of abolitionism. My intention is to suggest only that the key counterpoint is not slavery and abolition but rather the enslavement of non-Europeans and abolition.73 The above arguments have some further implications. One is for the wellknown debate on the origin of white attitudes toward black people. As previously noted, the evidence presented here points to major non-economic factors in the decision to use African slaves. While it certainly became profitable to replace European indentured servants with African slaves, the main issue was not relative profits but rather the inability of colonists to conceive of Europeans as chattel slaves. Such a conception might well have developed given enough time, but the sugar revolution proceeded too quickly to allow Europeans to adjust perceptions of insiders and outsiders. If Europeans had been able to accord Africans the same rights as themselves in the early modern period, they would not have enslaved them and brought them to the plantation Americas. The absence of European slaves, like the dog that did not bark, is perhaps the clue to understanding the slave trade and the system it supported. A second implication, less obvious but just as important, is that Europeans, and more particularly the English, failed to take advantage of two significant economic opportunities. For the first, if they had emulated the sixteenthcentury Russian aristocracy by creating an ideological distance between the common people and themselves and enslaving some members of their own society, they would have enjoyed lower labor costs, a faster development of the Americas, and higher exports and income levels on both sides of the Atlantic. For those who see European, especially English, economic power built on overseas colonies, it might be argued that, for the underpopulated tropical Americas at least, exploitation of the periphery and the transfer of surplus to the core would have been far more rapid with white slave labor. The second failure to maximize an economic advantage occurred when Europeans gradually widened their perception of what constituted an insider by beginning in the late eighteenth century to include transoceanic peoples. This change in perception brought a very profitable institution to an end. The first “missed abolitionists; Seymour Drescher, “The Long Goodbye: Dutch Capitalism and Antislavery in Comparative Perspective,” forthcoming, AHR, February 1994. 73  For the recent literature and a judicious discussion of the roots of antislavery in the English context, see David Turley, The Culture of English Antislavery, 1780–1860 (London, 1991), 1–46, 227–36. For the decline of Dutch slavery in the East Indies, where it might be noted that the institution was not related to export production, see Susan Abeyasekere, “Slaves in Batavia: Insights from a Slave Register,” in Reid, Slavery, Bondage, and Dependency, 286–314, esp. 308–10.

1186

Eltis

opportunity” created the Atlantic slave trade from Africa; the second ended not only the slave trade but slavery in the Americas as well. A third implication—really an extension of the previous point—consists of the notion that seventeenth-century capitalism, mercantile or not, was hardly as unrestrained and voracious as many students of early modern Europe have portrayed it. Profit-maximizing behavior occurred within agreed-upon limits, limits defined at least as much by shared values as by the resistance of the less-propertied classes. This should not be surprising. There have always been extensive areas of personal life defined as beyond the limits of the market and profit maximization even in the Western world, and the expansion and contraction of such an area is a major field of study, particularly as it pertains to wage labor. By the seventeenth century, enslavement of fellow Europeans was beyond the limits. More interesting, however, is that the peoples with the most advanced capitalist culture, the Dutch and the English, were also the Europeans least likely to subject their own citizens to forced labor. The fourth implication is the corollary of this last point. The countries least likely to enslave their own had the harshest and most sophisticated system of exploiting enslaved non-Europeans. Overall, the English and Dutch conception of the role of the individual in metropolitan society ensured the accelerated development of African chattel slavery in the Americas (and Asian slavery in the East Indies) because their own subjects could not become chattel slaves or even convicts for life. Further implications touch the English Revolution as well as its American counterpart of nearly a century later. There may be something to be said for expanding a variation of Edmund Morgan’s argument to cover the whole of the British Atlantic, in the sense that the celebration of British liberties—more specifically, liberties for Englishmen—depended on African slavery. But if this route leads to an interpretation of history driven purely by whichever interest group was dominant, then we should pause. Capitalists could have made more by selling European convicts for life than African slaves, and more again if the progeny of those convicts, like Africans, had also been bound to a lifetime of service. In the end, the absence of European slavery suggests that narrowly economic interpretations of history often miss the point.

The Living Dead Aboard the Slave Ship at Sea Stephanie Smallwood Like the Atlantic market of which it was a part, the slave ship at sea produced two competing narratives of the experience of the transatlantic voyage. The dominant European narrative—that of merchant investors—represented the ship as the height of maritime commercial endeavor: a useful conveyance that linked markets in a known seascape, an instrument of commerce that carried human cargoes from African sites of production to American sites of consumption. As the African captives lacked a culture of maritime travel, the ship produced in them an experience of motion without discernible direction or destination. On the ship at sea the logic of commodification reached its nadir. It was here, on the ocean crossing, that the practices of commodification most effectively muted the agency of the African subject and thereby produced their desired object: an African body fully alienated and available for exploitation in the American marketplace. The slave ship at sea produced an African narrative of persistent and often lonely attempts among the captives to continue to function as subjective beings—persons possessing independent will and agency. Women who exhausted themselves to death in their futile efforts to attend to the needs of their infants; captives who helped care for one another when disease invariably struck; and strangers who facilitated communication between speakers of mutually unintelligible languages—by such simple acts, all demonstrated the determination required to live as a human being rather than exist as an object aboard the slave ship at sea.

“This Hollow Place”

Remembering his childhood experience of slavery, Olaudah Equiano recalled the slave ship at sea as a site of profound displacement. Trying to make sense

Source: Smallwood, Stephanie, “The Living Dead aboard the Slave Ship at Sea,” in Stephanie Smallwood, Saltwater Slavery: A Middle Passage from Africa to American Diaspora, Cambridge, Mass.: Harvard University Press, 2007, 122–152. Copyright © 2007 by the President and Fellows of Harvard College.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_038

1188

Smallwood

of his captivity aboard an English ship, he asked others older and ostensibly more knowledgeable than he a simple question: “What was to be done with us?” They told him, he wrote many years later, “We were to be carried to these white people’s country to work for them.” Equiano tried to quiet his fears of imminent death and struggled to convince himself, “If it were no worse than working, my situation was not so desperate.” But these assurances did not get to the root of his anxiety. “I asked them,” he wrote, whether “these people had no country, but lived in this hollow place?”1 News that his captors indeed hailed from a “distant” land prompted more queries: “Then,” said I, “how comes it in all our country we never heard of them?” They told me because they lived so very far off. I then asked where were their women? Had they any like themselves? I was told that they had. “And why,” said I, “do we not see them?” They answered, because they were left behind. I asked how the vessel could go? They told me they could not tell; but that there was cloth put upon the masts by the help of the ropes I saw, and then the vessel went on; and the white men had some spell or magic they put in the water when they liked, in order to stop the vessel. I was exceedingly amazed at this account, and really thought they were spirits. I therefore wished much to be from amongst them, for I expected they would sacrifice me; but my wishes were vain—for we were so quartered that it was impossible for any of us to make our escape.2 In this often-cited passage from his Narrative, Equiano’s language compels understanding through its analytical precision. For whatever the balance of personal experience, ethnographic report, and imagination in his published account, Equiano’s description of the ship at sea is strikingly specific in its analysis of the puzzle the slave ship at sea represented for African captives.3 In 1  Olaudah Equiano, The Interesting Narrative of the Life of Olaudah Equiano, Written by Himself (London, 1789; reprint, ed. Robert J. Allison, Boston: Bedford, 1995), p. 55, emphasis mine. 2  Ibid. 3  On recent debate questioning whether Equiano was a native of Africa or born in slavery in the Americas, see Vincent Carretta, “Olaudah Equiano or Gustavus Vassa? New Light on an Eighteenth-Century Question of Identity,” Slavery & Abolition 20, no. 3 (1999): 96–105; Carretta, “Questioning the Identity of Olaudah Equiano, or Gustavus Vassa, the African,” in The Global Eighteenth Century, ed. Felicity Nussbaum (Baltimore: Johns Hopkins University Press, 2003), pp. 226–235; Alexander X. Byrd, “Eboe, Country, Nation, and Gustavus Vassa’s Interesting Narrative,” William and Mary Quarterly, 3rd series, 63, no. 1 (January 2006): 123–148.

The Living Dead Aboard The Slave Ship At Sea

1189

the first instance, there was the persistent problem of European intentions— “What was to be done with us?” If slave ships carried some who confidently claimed to know that the slavers were an instrument of labor recruitment, surely for others (no doubt even from among the sophisticated Akan) the weight of inexperience was too great to counter fear and anxiety. And though we cannot know, we must suspect that even those who began the crossing in confidence struggled to square what they could claim to know with what they encountered as the slave ship carried them deeper into Atlantic waters. Additionally, African captives confronted the problem of the European merchant ship itself, which presented them with challenges both physical and metaphysical. With regard to physical challenges, its cavernous form signaled an eerie emptiness demanding to be filled, a powerful and dangerous capacity to consume. As for the metaphysical aspect, the very habitat of the ship—the open sea—challenged African cosmographies, for the landless realm of the deep ocean did not figure in precolonial West African societies as a domain of human (as opposed to divine) activity—just as it had not figured as such in medieval European systems of knowledge.4 In its guiding principle—the proposition that life can be lived at sea—the ship presented an oxymoron. Now that land not only lay far distant but had, more ominously, vanished from the horizon altogether, the ship’s relentless motion pulled the captives ever deeper into temporal and spatial entrapment. The sheer scale of the unknown element disabled many of the cognitive tools supplied by African epistemologies, which attributed dangerous supernatural powers to the watery realm.5 The slave ship at sea reduced African captives to an existence so physically atomized as to silence all but the most elemental bodily articulation, so socially impoverished as to threaten annihilation of the self, the complete disintegration of personhood. Here their commodification built toward a crescendo that threatened never to arrive, but to leave the African captives suspended in an agony whose language no one knew.

4  Here I am speaking of the deep waters of the Atlantic, as distinct from the inland waters or the coastline, whose contours and patterns African fishermen knew intimately. 5  Little has been written on this dimension of African experience, beyond William D. Piersen’s examination of suicide among Africans enslaved in the Atlantic system, “White Cannibals, Black Martyrs: Fear, Depression, and Religious Faith as Causes of Suicide among New Slaves,” Journal of Negro History 62 (1977): 147–159. Especially important for this reason is John Thornton, “Cannibals, Witches, and Slave Traders in the Atlantic World,” William and Mary Quarterly, 3rd series, 60, no. 2 (April 2003): 273–294.

1190

Smallwood

Where, then, was agency to be found—or some affirmation that there yet remained a self to be preserved? Of the relationship of pain to language, Elaine Scarry writes, “Through his ability to project words and sounds out into his environment, a human being inhabits, humanizes, and makes his own a space much larger than that occupied by his body alone.” Indeed, the only means to survive in this realm was to divine a means to explain it, to define and delimit it. And the only means to achieve that was to speak of it—to probe its contours with words spoken among strangers. Words were the glue that made the crowd to which Equiano belonged into a collective “us,” whose fate stood in the balance during the journey. Agency aboard the slave ship took refuge above all in the voice, the means by which the “self” finds realization “across the bridge of the body in the world.”6 For African captives, it was their wholeness as fully embodied subjects that was at stake in the Atlantic setting. Entering the open sea signaled the end of one contest—whether captives would go into circulation as commodities in the Atlantic market—and the beginning of another—how captives would sustain their humanity in the uniquely inhumane spatial and temporal setting of the slave ship at sea. This truth lends substance to Equiano’s poetic insight that the ship packed beyond capacity with a “full complement” of African captives was a “hollow place”: a place distinguished by its many lacks—its material and social poverty, its cognitive dissonance, and its defenselessness in the face of the supernatural.

Atlantic Time-Space and Its Reckoning

Having had no reason to develop a body of knowledge and ideas about the sea as an arena for the activities of the living (as opposed to those of the ancestors—the dead living), enslaved Africans entered the Atlantic without the information and background that enabled their European captors to navigate the open sea. When confronted with the phenomenon, African captives responded to the Atlantic as Europeans had done: they made it knowable in their own terms. But the conditions of their Atlantic experience shaped that process, just as the particular conditions of maritime exploration had shaped the integration of the Atlantic into European culture and consciousness. With their charts of winds and currents, chronicling of important events such as storms and encounters with enemy vessels, and accounts of daily 6  Elaine Scarry, The Body in Pain: The Making and Unmaking of the World (New York: Oxford University Press, 1987), p. 49.

The Living Dead Aboard The Slave Ship At Sea

1191

activities at sea, artifacts such as Captain Peter Blake’s “Journall” of his voyage aboard the James provide abundant evidence of the ways European mariners made sense of the Atlantic. Applied statistical analysis, epidemiology, and modern biological science, too, have proved important tools for historical interpretation of the ocean crossing. But neither the European nautical instruments of that era nor modern Western historical explanations can help us fathom Africans’ own understanding of the Atlantic and the slave ship traversing it. We can be sure, however, that for early modern peoples everywhere in the Atlantic basin, the initial encounter with the Atlantic as an arena for human activity was profoundly transformative. The ship under sail was a world unto itself, where the passengers had to rely only on the expertise of the sailors aboard—and on whatever spiritual power they might be able to summon. No group of early modern Atlantic sea travelers had greater experience of the Atlantic than European mariners. Over the long course of Europe’s gradual reach into the Atlantic, generations of sailors developed a rich maritime culture all their own. From the knowledge and skills of observation and interpretation that enabled the “old salts” to find meaning in the mysterious events at sea, the navigational science of maps, charts, and tools of measurement was distilled—an acquired body of knowledge without which the European westward expansion would not have been possible. “Seamen,” as historian Marcus Rediker has explained, “created a rich store of knowledge, much of it genuinely scientific and reliably predictive, from the heavens and the earth.”7 European travelers—those who did not build their lives around life at sea but chose to take up life as emigrants to the Americas—also accumulated a different kind of knowledge about the Atlantic. Describing the seventeenthcentury Puritan migrations to New England, David Cressy writes, “English migrants to North America often approached their journey with apprehension and fear. Stories of storms and wrecks effectively dissuaded some prospective migrants from voyaging to America; others battled anxieties as they encountered the strange and terrifying world of the sea.” With regard to the ambience on the oceangoing vessel, Cressy notes, “Skeins, coils and tangles of rope hung with mysterious complexity. Unfamiliar nautical tackle and the puzzling seafarer’s vocabulary sparked both wariness and curiosity.”8 7  Marcus Rediker, Between the Devil and the Deep Blue Sea: Merchant Seamen, Pirates, and the Anglo-American Maritime World, 1700–1750 (Cambridge: Cambridge University Press, 1987), p. 179. 8  David Cressy, Coming Over: Migration and Communication between England and New England in the Seventeenth Century (Cambridge: Cambridge University Press, 1987), pp. 146, 152.

1192

Smallwood

Indeed, “apprehensions about ‘the casualties of the seas, which none can be freed from’ and ‘the length of the voyage … such as the weak bodies of women and other persons worn out with age and travail … could never be able to endure,’ had turned some of William Bradford’s associates away from the Pilgrim venture.” Even after English colonization of North America was well under way, “similar concerns dissuaded some prospective emigrants” from leaving England. “Good workmen … are fearful to go to sea for fear they shall not live to come to your land,” one writer observed in 1639. “ ‘Were it not for the danger of the seas you might have enough,’ ” he went on, explaining the challenge of drawing newcomers to Connecticut. Similarly, a writer from Newcastle, England, lamented that “ ‘the sad discouragement in coming by sea [was] enough to hinder’ ” his return to join his family in New England.”9 The point, too easily glossed over in familiar narrative formulations attributing to all Europeans the mariner’s hard-won mastery of the sea, is that “before the great migration of the 1630s only professional travellers, merchants and mariners knew the ropes of blue-water sailing.” Many of the English emigrants came from coastal communities possessed of a rich maritime tradition based on the navigation and exploitation of inshore waters. But that store of knowledge did not mitigate English fears. “The ocean was another matter, unknown and therefore feared,” writes Cressy, noting further that “to the popular imagination the ocean suggested hazard and uncertainty. It conjured an alien and frightful environment of commotion and discomfort, fraught with ‘daily expectations of swallowing waves and cruel pirates.’ ” Cressy observes that “as some colonial promoters pointed out, open-water sailing was much less hazardous than coasting,” and concludes that “much of this thalassophobia was overblown,” as “remarkably few people drowned on the way to New England, and many found the journey more exhilarating than tormenting.”10 But this perhaps misses the point: English emigrants believed transatlantic travel to entail unprecedented hazards, and reports of successful, even “exhilarating,” crossings did not necessarily relieve prospective emigrants of their fears. Through the networks of communication that linked New World colonies to Old World metropolises, various forms of information about the ocean crossing made its way back through European communities. The content of letters written to friends and family back home and of promotional pamphlets designed to entice new migrants coalesced into a kind of rumor-based knowledge—information that was reliable because it came from someone trusted, or someone who spoke with authority. Thus, the prospective European 9  John Wollcott and William Cutter, cited ibid., p. 147. 10  Ibid., pp. 146, 147.

The Living Dead Aboard The Slave Ship At Sea

1193

emigrant who could not fathom the idea of crossing the Atlantic could seek information, and find comfort, or not, in the idea that “a ship at sea may well be compared to a cradle rocked by a careful mother’s hand, which though it be moved up and down is not in danger of falling.”11 Like Europeans before the age of Columbus, Africans developed their knowledge of the ocean from familiarity with coastal lands and inshore waters. The sea provided a vital source of animal protein, and the salt deposited on the shore made it possible to preserve food.12 Most important, Africans knew that the sea was controlled by powerful deities whose benevolence was the real source of the sea’s gifts to them, and whose disfavor was the source of the sea’s destructive potential.13 That the earth and the sea were sacred sites of supernatural power was an African truth that De Marees explained thoroughly, if unwittingly, in his exegesis of African ignorance of the Christian God: Although they do not know him, he has given them gold, Palm wine, Millie and Maize, Chickens and Oxen, Sheep, Bannanas, Yams and other Fruits for their upkeep. But this they were not willing to concede, and they could not understand that such things came from God, saying it was not God who gave them the Gold but rather the earth, in which they seek and find it; according to them, God does not give them the Millie or Corn which they sow and reap either, but the earth gives them these things. Thirdly, Fruits are given by the Trees which they plant and which were first brought by the Portuguese. Fourthly, young Sheep come from old ones; the Sea gives Fish and they catch them; and there are many such things which they do not wish to acknowledge as coming from God, but rather say they come from the earth and the sea, each giving of its own.14 11  Colonial promoter William Wood, as quoted ibid., p. 148. 12  See Wilhelm Johann Müller, cited in Adam Jones, ed., German Sources for West African History, 1680–1700 (Wiesbaden: Franz Steiner Verlag, 1985), pp. 233–239, on the variety of fish harvested by Gold Coast fishermen, for example. See, on techniques employed to extract salt from seawater, ibid., p. 244, and Robin Law, The Slave Coast of West Africa, 1570–1750: The Impact of the Atlantic Slave Trade on an African Society (Oxford: Oxford University Press, 1985), p. 25. 13  See William Bosman, A New and Accurate Description of the Coast of Guinea (London, 1705; reprint, New York: Barnes & Noble, 1967), 368v, 383; Michael Hemmersam, cited in Jones, German Sources, pp. 118–119. 14  Pieter de Marees, Description and Historical Account of the Gold Kingdom of Guinea (1602), trans. and ed. Albert van Dantzig and Adam Jones (Oxford: Oxford University Press), pp. 72–73.

1194

Smallwood

Their relationship to the sea thus reflected the underlying logic found throughout traditional African systems of thought: that the sacred and the secular, the physical and the metaphysical were not separate spheres but rather integrally bound together and manifest in the material world. Also like European cultures, coastal African communities developed a rumor-based body of knowledge of their own about the Atlantic arena. Though experience in the Atlantic beyond the coastal waters was relatively rare among resident Africans before the nineteenth century, small numbers of them crossed the Atlantic aboard European vessels and subsequently returned to Africa as witnesses to the world beyond inshore waters. Particularly along the Gold Coast, where European settlement was so extensive, African men, occasionally taken to Europe to learn English, returned to the Gold Coast and found employment as interpreters. Others from the region found their way across the Atlantic and back as accidental passengers aboard slave ships. In 1719, three canoemen and another employed as a “Gold taker” traveled aboard a Royal African Company ship to Barbados, where they passed at least four months before being returned to Africa. “Captain Ayerst brought with him 3 Cannoe Men,” the Royal African Company agent at Barbados reported on 7 March 1719, “which shall be return’d by the first as also a Gold taker named Quamina.”15 On 30 June 1719, the company agent at Barbados “shipd on bd. the John & Elizabeth, Captain Jacob Burgesson for Cape Coast, the following Slaves, Accraw, Quaw, Cuffee, & Cobiner.”16 Having crossed the Atlantic and spent several months on American territory, these four, when they walked once more on African soil, shared with others accounts that no doubt spread quickly among Cape Coast residents. The structure of African involuntary migration was distinguished most by the general impossibility of such informational feedback, for persons who witnessed life in the Atlantic world and returned to Africa to tell about it represented a tiny minority when measured against the millions of deportees whose voices could not reach Africa’s shores. With no network of communication linking their experience in the Atlantic world, the African captives who boarded European slave ships carried a range of expectations shaped by the interplay of their own beliefs regarding the men who had purchased them and whatever rumors circulated among the coastal communities about these strangers and the land from which they came.

15  James Gohier, Barbados, to RAC, 7 March 1719, T70/8, f. 215; see also Gohier to RAC, 25 May 1719, T70/8, f. 217. 16  James Gohier to RAC, 30 August 1719, T70/8, f. 218.

The Living Dead Aboard The Slave Ship At Sea

1195

For African emigrants, as we have seen, the slave ship was not just a setting for brutality and death, but also a locus of unparalleled displacement. As the sight of land grew faint, or as the land disappeared suddenly on the closing of the hatch, the disorientation that for many had begun with the process of procurement on the African coast became more marked. Out of sight of any land, enslaved Africans commenced a march through time and space that stretched their own systems of reckoning to the limits. By the time Europeans began to colonize the New World, voyagers to the west were confident that their journey would follow a linear path, with known beginning and end points. European seamen translated the land-based systems of time-space reckoning of medieval Europe to the wider temporal and spatial context of life in a “new” Atlantic world. Hourglasses and astrolabes measured time and space; and applied mathematics and geometry turned these into the Cartesian coordinates seamen used to recognize place in the seemingly formless arena of the sea.17 Ship captains like Peter Blake could know when their ships had reached designated spatial coordinates at sea—virtual landmarks such as the equator; and whether six months or a year had elapsed since their departure from England.18 Measured against this standard, the dramatic dimensions of the displacement Africans experienced as they moved farther out into the Atlantic, their own knowledge of time and place rapidly losing its utility, can be grasped at least intellectually.19 Like many nonliterate peoples, Africans relied on the regular cycle of climatic events to locate themselves in time and space. The dry season of Harmattan winds had begun when most of the Africans on the James had come aboard, 17  J. H.  Parry, The Age of Reconnaissance (Berkeley: University of California Press, 1963), pp. 83–113. 18  Stephen Greenblatt, Marvelous Possessions: The Wonder of the New World (Chicago: University of Chicago Press, 1991); Mary W. Helms, Ulysses’ Sail: An Ethnographic Odyssey of Power, Knowledge and Geographical Distance (Princeton, N.J.: Princeton University Press, 1988); Anthony Grafton and Ann Blair, eds., The Transmission of Culture in Early Modern Europe (Philadelphia: University of Pennsylvania Press, 1990); Hildegard Binder Johnson, “New Geographical Horizons: Concepts,” in First Images of America: The Impact of the New World on the Old, ed. Fredi Chiappelli, 2 vols. (Berkeley: University of California Press, 1976); Peter Hulme, Colonial Encounters: Europe and the Native Caribbean, 1492–1797 (London: Methuen, 1986). 19  There is a rich literature on the anthropology of time. I have benefited in particular from the essays collected in John Bender and David E. Wellbery, eds., Chronotypes: The Construction of Time (Stanford, Calif.: Stanford University Press, 1991); and Johannes Fabian, Time and the Other: How Anthropology Makes Its Object (New York: Columbia University Press, 1983).

1196

Smallwood

and if it was the middle of the month of March by European reckoning when Blake removed the slaves’ irons, no doubt some among the Africans knew that the planting season now was under way in the communities they had left behind.20 Aspects of the calendrical system used by Akan-speaking peoples of the Gold Coast drew the attention of many European visitors to the region in the seventeenth century. At the most superficial level, Europeans recognized that certain days held ritual importance among the coastal Akan communities and noted the day of the week in the Julian calendar to which these days in the Akan system of reckoning time corresponded. In 1603–1604, Ulsheimer observed that the day of the week Europeans identified as Tuesday was significant. “Every Tuesday (which they celebrate as we do Sunday),” he explained, “people must bring the king all the wine from his whole territory, and he in turn must give it out liberally to his people.”21 During his residence at the Dutch fort at Mori, Brun likewise noted, “Tuesday is their sabbath and on this day they do not go out to sea; for their god Fytysi [fetiso] had forbidden them to do so.”22 Müller observed the same during his stay among the Fetu of Cape Coast. “Instead of Sunday, Tuesday is celebrated throughout the country,” he reported. It is called ohinne da, the king’s day, by them. The fishermen living on the coast hold this day so sacred that they believe a great disaster would befall them if they went to sea to fish with hooks on that day. Nevertheless they are free to put out and fetch in their nets on Tuesday. This day is celebrated in honour of the o-bossums or supposed gods of the sea. When the sea becomes rough, so that the fishermen either suffer loss or cannot

20  See Philip F. W. Bartle, “Forty Days: The Akan Calendar,” Africa 48, no. 1 (1978): 80–84; Joseph K. Adjaye, “Time, the Calendar, and History among the Akan of Ghana,” Journal of Ethnic Studies 15, no. 3 (1987): 71–100; T. C. McCaskie, “Time and the Calendar in Nineteenth Century Asante: An Exploratory Essay,” History in Africa 7 (1980): 179–200. See also Paul Bohannan, “Concepts of Time among the Tiv of Nigeria,” Southwestern Journal of Anthropology 9 (1953): 251–262; K. K. Bunseki Fu-Kiau, “Ntangu-Tandu-Kolo: The Bantu-Kongo Concept of Time,” in Time in the Black Experience, ed. Joseph K. Adjaye (Westport, Conn.: Greenwood Press, 1994), 17–34; Adjaye, “Time, Identity, and Historical Consciousness in Akan,” ibid., 55–77; Mechal Sobel, The World They Made Together: Black and White Values in Eighteenth-Century Virginia (Princeton, N.J.: Princeton University Press, 1987), pp. 15–67. 21  Andreas Josua Ulsheimer in Jones, German Sources, pp. 30–31. 22  Samuel Brun in Jones, German Sources, p. 86.

The Living Dead Aboard The Slave Ship At Sea

1197

even fish, they get the priests to question these gods and appease them with a sacrifice.23 In addition to learning the Akan term ohinne da, which he translated as “Tuesday,” Müller added Quassi-da (Sunday), Egwju da (Monday), Ejauda (Thursday), and Efi-dà (Friday) to the vocabulary list he compiled. Müller found that Adà represented the unit of time he knew as a “day,” and on learning that the Akan employed a larger unit of seven days called Dansun, he translated that term as “week.” He further identified Osran, which he translated as “month,” as the Akan word for “moon.” Finally, Müller concluded that the Akan word Affi corresponded to the European concept of a year.24 In the Akan system for reckoning time, the units of temporal measure identified by Müller comprised a forty-day cycle known as adaduanan (literally, “forty days”).25 The cycle actually consisted of forty-two days, however, and appears to have derived from the fusion of two systems of temporal measure—one recognizing a six-day week, the other featuring a seven-day week—observed by the different groups that settled the region well before the era of European exploration. From an apparent process of cultural amalgamation a calendrical system had been derived in which the six-day cycle and seven-day cycle ran concurrently, so that “when the six-day week is counted side-by-side with the seven-day week it takes a total of forty-two days to reach all combinations.”26 Within the forty-day cycles certain days were designated for ritual observance of the ancestors—the special days remarked on by Müller and other seventeenth-century European commentators.27 Beyond its primary role in ordering the cycle of ritual days associated with ancestor worship, the 40-day adaduanan cycles also gave shape to the annual agricultural calendar. Multiple

23  Müller found that among the Fetu agriculturalists, however, Sunday was the appointed day of rest. Müller, ibid., p. 167. 24  Müller, ibid., Appendix A, items 39–66. 25  Bartle, “Forty Days,” p. 81; Adjaye, “Time, the Calendar, and History,” p. 79. 26  Bartle, “Forty Days,” pp. 81–82. Bartle suggests that the six-day week was observed by Guan-speaking peoples, while the seven-day week “may have been brought south with itinerant traders from the savanna” (p. 81). 27  These days of ritual observance also were referred to as bad days (da bone), on account of the limitations placed on extensive activity on such days and the expectation that violation of the prohibition would bring misfortune. The precise number of ceremonial, or “bad,” days within the forty-day cycle remains a subject of debate: see Adjaye, “Time, the Calendar, and History,” p. 83.

1198

Smallwood

adaduanan cycles completed a year, which was marked by the annual harvest festival, odwira. The timing of the odwira celebration and other annual ­agricultural rites was determined by priests “in consultation with the gods and ancestors,” and also on the basis of the “ripening of the crops.” Since the number of aduadanan within the year varied between eight and nine such 42-day cycles in a 365-day agricultural calendar, “annually celebrated rites of the different Akan groups … are therefore celebrated each year on different days of the year.”28 Within the year, each adaduanan cycle was tied to the particular ecological demands of the season. According to one historian who has studied Akan time reckoning, the various adaduanan cycles within the year are given a number of appellations, which are not the same from place to place, and of course never quite the same from year to year, since there are less than nine and more than eight cycles in any one year. Opepon (Ope = harmattan, dry season; pon = supreme), for example, more or less corresponds to the adaduanan which appears about January–February in the middle of the dry season.29 Shaped by the seasonal demands of the landscape, the Akan system for reckoning time was no more up to the challenge of giving order and structure to life in a maritime arena than had been those of Europeans prior to their fifteenth century.30 Even for those among African captives who were accustomed to travel—traders or fishermen, say—Atlantic displacement was unlike anything with which they were familiar. Travel overland, for no matter how great a distance, took place in daily increments: as each day came to a close, the travelers found a place to rest and pass the night before commencing the journey again the following morning. Similarly, fishermen and traders who followed routes connected by coastal waters were accustomed to travel only by daylight, nighttime being the occasion to stop and to rest. Travel aboard the slave ship, however, would require that captives remain as watchful at night as during the

28  Bartle, “Forty Days,” p. 83. There is debate also regarding the number of adaduanan within an annual cycle. Whereas Bartle maintains that the number varied, Adjaye has argued that the measure was more precise, the year being “composed of nine ‘monthly’ cycles of forty days each” (p. 79). 29  Bartle, “Forty Days,” p. 83. 30  See Parry, Age of Reconnaissance, pp. 83–99, on the slow and painstaking process by which European systems of time-space reckoning were adapted for use in a maritime arena.

The Living Dead Aboard The Slave Ship At Sea

1199

daylight hours, for nightfall brought no halt to the forward motion of the ship at sea.31 Once they were onboard ship, the African captives measured time at sea as best they could by marking the cycles of the moon.32 The rising and setting of the sun and the moon’s regular appearances were familiar markers, but without a template for ordering time in this maritime arena, so distinct from the landscape from which they had been torn, Africans struggled to maintain their bearings. Counting the moon’s cycles, they knew what time it was by the agricultural calendar that had ordered their lives—whether it was time to harvest or time to plant. But that information did not help them know where, precisely, they were, how far they had traveled, how to situate this place in their own mental maps of the world. Their travel as captives thus made it enormously difficult for Africans to clearly distinguish the phases of their journey, or to anticipate the end of one phase or the beginning of another. The events that gave shape to their time aboard the slave ship at sea were random, indeterminate signs to be observed and interpreted as they appeared. Always in motion but seeming to never reach any destination, the ship plowed forward in time without ever getting anywhere, always seeming to be in the same place as the day before. It was as if time were standing still. Time was lived in motion, but at no discernible rhythm by which Africans could orient their movement in time or readily measure time’s passage.

The Accounting of the Dead

In few settings were human beings exposed to a greater number of pathogens in the early modern world than aboard slave ships en route to the Americas.33 Any sailing vessel was an enclosed space, where it was impossible to create 31  Ivor Wilks, “On Mentally Mapping Greater Asante: A Study of Time and Motion,” Forests of Gold: Essays on the Akan and the Kingdom of Asante (Athens: Ohio University Press, 1993), pp. 189–214. 32  Ibid. See also McCaskie, “Time and the Calendar”; Walter J. Ong, Orality and Literacy: The Technologizing of the Word (London: Routledge, 1982). On Africans’ lunar measurement of time in the Atlantic world, see Philip D. Morgan, Slave Counterpoint: Black Culture in the Eighteenth-Century Chesapeake & Lowcountry (Chapel Hill, N.C.: University of North Carolina Press, for the Omohundro Institute of Early American History and Culture, 1998), p. 201. 33  See Philip D. Curtin, “Epidemiology and the Slave Trade,” Political Science Quarterly 83 (1968): 190–216.

1200

Smallwood

physical distance between the infected and the noninfected, African or European, when disease broke out. But the crowded conditions on slaving vessels made for a level of human density unmatched on other types of oceangoing vessels.34 With crowding came lack of sanitation, and the enslaved Africans found that none of the familiar habits of personal hygiene could be observed. Thus, illness was nearly impossible to avoid in that setting. Exhaustion, malnutrition, fear, and seasickness resulted in depressed immune systems and increased vulnerability to disease. Particularly among the African prisoners, crowding, poor hygiene, and often contaminated food and water supplies, together made the slave ship a breeding ground for airborne pathogens (smallpox and tuberculosis) as well as those spread via fecal-to-oral pathways (bacillary and amoebic dysentery) or by direct physical contact (yaws).35 The spatial dimensions of the Atlantic and the relatively slow pace of oceanic travel in the age of sail meant that the assault on health was sustained without relief for the duration of the sea voyage. The Atlantic passage was above all else a test of African endurance, with no known end in sight. Every African who boarded a European slave ship brought his or her own strengths and weaknesses to the encounter—a reflection of the life that had preceded expulsion into the Atlantic. The basic nutritional profile of the community of origin distinguished inhabitants of the Gold Coast, for example, whose primary staple crop was maize, from those of Biafra, whose primary crop was the less nutritious yam.36 And the particular events that carried 34  See Stanley L. Engerman and Herbert S. Klein, “Experiences on Slave Ships Compared with those in Other Long-Distance Oceanic Migrations” (paper presented at conference, “Transatlantic Slaving and the African Diaspora,” Omohundro Institute of Early American History and Culture, Williamsburg, Va., 11–13 September 1998), p. 11. 35  Alfred W. Crosby, “Smallpox,” in The Cambridge World History of Human Disease, ed. Kenneth F. Kiple (Cambridge: Cambridge University Press, 1993), p. 1008; K. David Patterson, “Bacillary Dysentery” and “Amebic Dysentery,” ibid., pp. 604–606, 568–571; William D. Johnston, “Tuberculosis,” ibid., pp. 1059–1061; Don R. Brothwell, “Yaws,” ibid., pp. 1096–1100. 36  On a related subject, examination of height data for slaves has shown that “slaves from the Gold Coast were about an inch taller than the average Africans entering the West Indies.” Their stature reflected their superior nutritional profile; by contrast, “those regions such as the Bight of Biafra and Central Africa, where the diet centered on yams or cassava, produced the shortest slaves introduced to the West Indies.” Kenneth F. Kiple, The Caribbean Slave: A Biological History (Cambridge: Cambridge University Press, 1984), p. 58. For full presentation of the data, see B. W. Higman, “Growth in Afro-Caribbean Slave Populations,” American Journal of Physical Anthropology 50 (1979): 373–386.

The Living Dead Aboard The Slave Ship At Sea

1201

individual men, women, and children aboard the European ships made for varying combinations of exhaustion, injury, hunger, exposure to contagious disease and vulnerability to pathogens that were endemic to local African populations. Within every cargo, some were more physically or emotionally exhausted, some more nutritionally deprived, and some more vulnerable to illness than others. Conditions thus made the slave ship a deadly place, so much so that in Africa the language of death became part of nomenclature for it. Slave ships were called tumbeiros in the eighteenth-century Angolan trade, for example, a term historians have translated as “floating tombs” or “undertakers.” The symbolism of the tomb also was used to designate the personnel involved in the activities of slaving, other historians having translated tumbeiros to mean “bear[ers] to the tomb,” finding it used also to refer to those “who brought slaves down to the coast.”37 For the most part, though, historians have described the slave ship’s lethal nature the same way the slave traders did: by calculating the number of dead. Indeed, as long as the transatlantic slave trade has been a subject of intellectual inquiry, mortality and its quantitative measurement have functioned as a key with which scholars have labored to unlock the secrets of the slave ship and expose to view the stories that played out belowdecks. By tallying the dead to measure the toll the voyage took on African life, we have made that body count the most potent symbolic measure of the horrors of the middle passage. We can gauge the trauma to the extent that we know how many lives it extinguished: the greater the number, the deeper the perceived tragedy. A great deal has been learned from the data on mortality aboard the slave ship, but overall numbers—and our interpretation of them—correspond only loosely to the ways African captives experienced and understood shipboard mortality. Narrative texts like Blake’s journal written aboard the James offer another window onto the slave ship at sea, one that helps us recognize the limitations of quantitative analysis. This alternate analysis of shipboard mortality, in alluding to the temporal and spatial realities of African captives’ transatlantic voyage, puts the statisticians’ numbers back into the context in which death occurred (Fig 5.1).

37  Joseph C.  Miller, Way of Death: Merchant Capitalism and the Angolan Slave Trade, 1730– 1830 (Madison: University of Wisconsin Press, 1988), pp. 314, 314n, citing Robert Conrad’s rendering of the term in Robert Conrad, ed., World of Sorrow: The African Slave Trade to Brazil (Baton Rouge: Louisiana State University Press, 1986).

1202

Smallwood

[To view this image, refer to the print version of this title.]

Figure 5.1 “Account of the Mortality” aboard the James, 1675–1676, T70/1211.

The Living Dead Aboard The Slave Ship At Sea

1203

On 6 September 1675, Peter Blake penned the following in his journal: “This day I had a neagg.r man departed this life whoe died suddenly.” Coming just over a week since the first captives had come aboard the vessel on 28 August, this first death of an African aboard the James occasioned the beginning of a supplemental documentary record Blake would keep within his “Journall.” Marking the first entry in his “Account of Mortallity of Slaves aboard the Shipp James,” Blake entered a “1” in the column labeled “Men,” and in a separate column for explanatory notes, he added the brief phrase “departed this Life suddenly.”38 Blake entered numbers and words in his account of mortality in language adequate to the needs and tasks of a slaving captain. The numeral stood for the property now lost, and succinct phrases reducing death to a simple statement of cause and effect were sufficient to explain the event: “departed this Life suddenly,” “departed this life of Convulsion fitts,” “Received from Wyemba [Winneba] very thin & wasted to Nothing & soe dyed.” Blake’s language not only explained the circumstances of death, but also identified the agent of death. The captives themselves bore the responsibility and had the agency—it was they who “departed this life.” The language of accounting thus rationalized shipboard mortality, portraying the European agent of commodification as the passive victim of the Africans who died—as an investor robbed of his property by that property. But for the person whose earthly life had ended here in the water-borne wooden vessel, the picture was far more complicated and troubling. Death represented not just a discrete event but rather a shift in social relations that had wide reverberations. Why did death weigh more heavily on one group of captives than another? Why did one man suffering from flux die, while another did not? Whose supernatural power intervened to save the woman ill with dysentery or smallpox? Why did the white doctor’s suspicious-looking medicine bring relief to one and death to another? Far from simple equations based on circumstances, the patterns of life and death aboard ship were complex and mysterious. The circumstances of death were not explanations in themselves but rather signs—indicators of activity among the supernatural forces that possessed the real power.

38  Peter Blake, “A Journall of my Intended Voyage for ye Gold Coast kept by mee Peter Blake Commander of ye Royall Companys ship James in ye searvis of ye Royall African Company of England,” 28 November 1695, T70/1211, f. 34.

1204

Smallwood

The “departure” that Blake had unambiguously recorded in fact put the departed very much in jeopardy, according to the understanding the African captives brought to the occurrence, for it was marked by the absence of traditional mortuary ritual—a distressing failure to act.39 If the prescribed ritual was not performed, death threatened to have far-reaching consequences for both the deceased and the living. Nowhere was this truer than in the Atlantic, where the most central element of mortuary practice was lacking: earth. Properly memorialized, death afforded the opportunity to join the living community through a protective web of connections to the ancestors and the not-yet-born. Mourning and interment mitigated the disruptive threat that death posed, by channeling the sacral power of death into the renewal of life. Suitable rituals thus protected the community from the unmitigated loss of its members and protected the individual from the threat of annihilation; the “necral space” of the burial ground and the rituals associated with it served as the medium through which these vital connections were maintained.40 Ancestors “consecrated” the ground in which they were buried, “and continuous rituals connecting them with their heirs created a single community consisting of the dead, their heirs, and the soil they shared.”41 The soul’s departure and migration to the realm of the ancestors could only be carried to completion by the performance of mortuary practices that affirmed the close affinity between these two domains. Death without a funeral compromised the journey on to a new realm. With no food and drink to sustain the deceased in the domain of the ancestors, neither clothing nor tools with 39  My interpretation of ritual is influenced especially by Katherine Verdery, The Political Lives of Dead Bodies: Reburial and Post-socialist Change (New York: Columbia University Press, 1999); Jonathan Z. Smith, To Take Place: Toward Theory in Ritual (Chicago: University of Chicago Press, 1987); and Mary Douglas, Natural Symbols: Explorations in Cosmology (London: Routledge, 1970). My thinking on ritual in the context of migration is influenced by an emerging literature on religion and geography. See especially Chris C. Park, Sacred Worlds: An Introduction to Geography and Religion (London: Routledge, 1994). 40  See Park, Sacred Worlds, p. 213, on “necral space” and “landscapes of death.” The fundamental grounding that a proper mortuary ritual accomplished was of course not unique to the Akan or other precolonial African societies, but rather was (and is) common to many cultures. My goal here is to take seriously the particular practices of Akan subjects. See also Achille Mbembe “Necropolitics,” trans. Libby Meintjes, Public Culture, 15, no. 1 (2003): 11–40. 41  Verdery, Political Lives of Dead Bodies, p. 104, describing mortuary rituals of Ancient Greece and Rome. For a discussion of the rituals developed by seamen to adapt “burial” practices to the maritime setting, see also Rediker, Between the Devil and the Deep Blue Sea, pp. 194–198.

The Living Dead Aboard The Slave Ship At Sea

1205

which to continue the activities of earthly life in the new realm, and no earth to receive the dead bodies, how were the deceased to find their way out of the watery realm to the land of the ancestors? In essence, a fully realized death could not be accomplished alone. Nor was it something one could attain at sea. Far more than the economic event Blake casually recorded in his journal, this first death of a slave aboard the James was an event that held singularly traumatic consequences: for the deceased, death at sea meant an unfulfilled journey to the grave and therefore also to the realm of the ancestors; for the kinsmen of the deceased, his death meant that a thread of the special power and protection only ancestral members of the community could provide was lost to them forever. For the collective of African captives remaining aboard the James, the death of one of their number left them with the burden of a tormented soul, trapped here among them because its migration to join the ancestors had been thwarted. Two more deaths followed in the next eight weeks, a time when only small numbers of captives were onboard. While Blake bartered for gold, his crew prepared the ship to receive the large numbers that would “complete” its human cargo. Once the cargo on the James approached capacity in January, the forward motion of the vessel would be charted by the dispersal of bodies committed without ceremony to the sea, an accumulation of displaced souls. One week following receipt of the group of “very thin ordinary slaves” from Winneba (Chapter 3), the death march continued. 20 January: a man, “received from Wyemba thin and Consumed to Nothing & soe dyed.” Then a reprieve of five days. 26 January: a woman, “received from Wyemba very thin & wasted to Nothing & soe dyed.” Then a lapse of twelve days. 8 February: a man, “received from Wyemba very thin & dropsicall & soe departed this life.” Then a full two weeks’ respite from the recurrence of death. 23 February: a woman, “bought to Windward & departed this life of a Consumption & Wormes.”

1206

Smallwood

24 February: a boy, “received from Wyemba with a dropsy & departed this life of ye same disease.” 26 February: a woman, “received from Wyemba thin & soe Continued Untill death.” And then a break of six days. 5 March: a woman, “miscarryed & the Child dead within her & Rotten & dyed 2 days after delivery.”42 On 6 March, Blake prepared to leave the African coast, his supplies of food, water, and fuel finally complete. His ship was now, he observed, “put … in a posture to sail tomorrow morning with the land breeze.” That following day, the James thus began its transatlantic voyage toward Barbados.43 With land fast disappearing in its wake, the James carried its cargo of captives forward into the open sea. The last point of spatial reference vanished once the ship was surrounded by an empty horizon on all sides; only the vessel itself offered fixity and the sense of place. Its scale dwarfed by the unbounded expanse of ocean, the ship became, in this astonishingly unearthly realm, a world unto itself. Following the death of the pregnant woman on 5 March, the captives’ first month at sea continued with a stretch of seven days during which no one perished aboard the James. No longer was death at sea a novel occurrence, but thus far it had targeted the people from Winneba with a persistence that could not have gone unnoticed. The pattern continued when the procession of departing souls resumed. 13 March: a man “received from Wyembah.” 15 March: a man, “received from Wyembah very thin & fell into a flux.” As the slave ship made its way into deeper waters, the daily cycles of land and sea breezes gave way to the more random, and powerful, winds and ocean swells on the open sea. Not long after the land disappeared from view, 42  Blake, “Account of the Mortallity of Slaves aboard the Shipp James,” in “A Journall of my Intended Voyage,” ff. 100–101. 43  Blake, “A Journall of my Intended Voyage,” f. 70v, 7, 8 March 1676.

The Living Dead Aboard The Slave Ship At Sea

1207

crewmen aboard most slave ships began to remove the iron shackles that the enslaved men had worn for weeks and sometimes months by this time. “When our slaves are aboard,” wrote Thomas Phillips, “we shackle the men two and two, while we lie in port, and in sight of their own country, for ’tis then they attempt to make their escape, and mutiny.” But “when we come to sea,” he explained, “we let them all out of irons, they never attempting then to rebel, considering that should they kill or master us, they could not tell how to manage the ship, or must trust us, who would carry them where we pleas’d.”44 Likewise, Jean Barbot reported that women were allowed to move about freely on deck, while “many of the males had the same liberty by turns, successively; few or none being fetter’d or kept in shackles, and that only on account of some disturbances, or injuries, offer’d to their fellow-captives, as will unavoidably happen among a numerous crowd of such savage people.”45 It was one week following the ship’s departure from the African coast before captives aboard the James regained the use of their limbs. In the midst of a storm of two days’ duration and sufficient strength to cause the main topsail yard to break, Blake decided to remove the shackles that bound their arms and legs. “This day,” Blake wrote, as the “turnadoe” continued with “much thund’r lightening and raine” on 16 March, “I putt all my slaves out of Irons.”46 Removal of iron restraints restored tiny freedoms that brought captives enormous relief. Whereas before actions as simple as rolling onto one’s side had required coordination with another person, now a measure of independent movement was possible. Even so, movements larger than small bodily gestures demanded choreographed effort. The reduction of so many bodies alive with kinetic energy to disquieting stillness was among the worst of the violence done to slaves on board ship. 18 March: a man, “received from Wyembah very thin & soe fell into a Consumption & departed this life.” 44  Thomas Phillips, Voyage Made in the Hannibal of London, Ann. 1693, 1694 … in A Collection of Voyages and Travels, 6 vols., ed. Awnsham Churchill and John Churchill (London, 1732), 6:229. 45  P. E. H. Hair, Adam Jones, and Robin Law, eds., Barbot on Guinea: The Writings of Jean Barbot on West Africa, 1678–1712 (London: Hakluyt Society, 1992), 2:779. 46  Blake, “A Journall of my Intended Voyage,” f. 71v. The majority of the slaves aboard the James having boarded in January, most had worn shackles for two months. Unlike the Africans aboard the Hannibal, commanded by Thomas Phillips, and Barbot’s vessel, it appears that both male and female slaves aboard the James may have worn irons.

1208

Smallwood

On 22 March, Blake recorded that the James had crossed to the south of the equator. This event, significant as an indication of the ship’s progress toward its destination, supplies a spatial referent not even perceptible to the African captives aboard the James. Of great meaning for the captives was the freshly caught fish Blake added to their diet. Having promised on 28 March to reward his crew with one pint of brandy “for every 10 fish” caught, he henceforth made albacore a regular part of the rations his captives received. Notations like the one he entered on 28 March, “Gave my slaves 10 fish in ye Suppie [Supper],” started to appear routinely in Blake’s journal. The eleven days beginning 19 March passed without event when it came to death, making this the longest such stretch of days during the entire voyage thus far. The loss of a captive from Winneba broke the lull. 30 March: a man, “received from Wyembah very thin & soe Continued Wasting Untill death.” 31 March: a “very sick” boy “fell overboard in the night & was Lost.” 6 April: a man, “received from Wyembah thin & Consumed very low & after dyed of a Great swelling of his face & head.” 14 April: a man, “received from Wyemba thin & dyed of a flux.” 15 April: a woman, “received from Wyemba Sickened & would not eat nor take anything.” 16 April: a man, “bought by mee & dyed of a flux.” 17 April: two men, “the one received from Wyembah & dyed of a flux. The other received ditto who Leaped over board & drowned himself.” The second man chose his moment to act in the afternoon hours. The ship’s boat quickly pursued him, a “stout manslave” to a trader’s eye, but “just as they came upp with him,” Blake wrote, “hee sunke downe,” and when the coxswain “runn downe his oare betweene his armes,” the man “would not take hould of it & soe drowned himselfe.” Thirteen weeks had passed since he had boarded the ship, and in this time he had seen seventeen bodies committed to the sea, thirteen of which belonged

The Living Dead Aboard The Slave Ship At Sea

1209

to people who also had boarded at Winneba. Why did that slave choose to take his own life? Did he seek to follow someone to whom he was connected by kinship? Did he somehow trust that his journey beyond the ship would lead to communion with his ancestors, although he had violated prohibitions against suicide? Or was it his choice to accept a different fate: Was his refusal to take the oar extended to bring him back to the ship a choice to risk the eternal wandering of his soul, rather than remain in the limbo of a ship that could sustain neither life nor death? 20 April: a woman, “received thin att Wyembah & dyed of a Consumption.” 21 April: a boy, “received from Wyemba with a dropsey & soe dyed.” 26 April: a woman, “bought by my selfe & being very fond of her Child Carrying her up & downe Wore her [self] to nothing by which meanes [she] fell into a feavour & dyed.” On 27 April, the ship crossed the equator for the second time. Sailing once again in northern latitudes, the James had reached another milestone in its transatlantic voyage: on a westerly tack, the ship now approached the Americas. Captives who had not lost track of the moon’s phases knew that the ship was entering its third month at sea; meanwhile, Blake turned his calendar to the month of May. At that point, those who had boarded the ship at other Gold Coast ports—Amersa, Anomabu, and Agga—began to join the community of the deceased. 1 May: a man and woman, both “received from Anamaboe departed this life of a flux.” 2 May: a woman, “received from Agga & departed this life of a flux.” 3 May: a man, “received from Wyembah & departed this life of a dropsey.” 4 May: a man, “received from Mr. Ballwood att Amyssa [Amersa] & dyed of a feavour by Lying in the [ship’s] Longboat in the rain in the night which noe man knew of for hee went into her privately.”

1210

Smallwood

5 May: a woman, “received from Wyembah very thin & old & departed this Life of the flux.” 6 May: a man, from Anomabu, who likewise “departed this life of a flux.” 8 May: a woman, “received from Wyembah with a Dropsey & departed this life of the same disease.” 9 May: a man, “bought by mee at Anamaboe & departed this life of the flux.” 12 May: a woman, “received from Wyembah thin & consumed away untill life departed from her.” 13 May: three men, one “received from Wyembah thin & departed of a flux”; another “received from Anamaboe & dyed of the Cramp in all his joynts & all over his body, being lately recovered of the flux”; and the third “bought my mee” and also “dyed of the Cramp.” Also on this day, Blake began to close in on his destination. “I directed my course Northwest and then by West distance 55.5 miles until 10 o clock, and then supposed myself to be near the latitude of the Island,” Blake recorded in his journal. Sixteen days after crossing back into the northern hemisphere, he was on the lookout for the tiny island of Barbados. 14 May: two women, “the one received from Wyembah very thin & departed this life of the flux” and “the other received from Anamaboe & departed this life of ye flux.” 16 May: a man, “received from Anamaboe & departed this life of a flux.” On 18 May the James was passed by a ship from New York, traffic that confirmed for Blake his proximity to the island. For the captives on board the James, signs that change was imminent were less clear. “Wee put our Slaves to noe allowance of Water” was the additional notation Blake entered into his journal that day.47 Meanwhile, the most oppressive of the ship’s rhythms continued with a steady and relentless beat. 47  Blake, “A Journall of my Intended Voyage,” ff. 76, 77, 78v.

The Living Dead Aboard The Slave Ship At Sea

1211

20 May: a woman purchased by Blake “departed this life of Convultions.” 21 May: a man “received from Agga & departed this life of a flux.” Also on this day, the island of Barbados finally appeared on the horizon in the early hours of the morning, a Sunday on Blake’s calendar.48 Steering a course for Barbados was not an easy task. The seventeenthcentury traveler Sir Henry Colt likened it to finding “sixpence throwne downe uppon newmarkett heath.” The easternmost of the Lesser Antilles by a distance of ninety miles, Barbados stands alone in the southwestern corner of the North Atlantic, marking the first doorway to the Caribbean Sea for those who could find the tiny island at all.49 Seventeenth-century sailors, who possessed not much more ability to determine their position in the Atlantic than Columbus had three hundred years earlier, relied on a combination of guesswork, science, and luck to find their way to the Caribbean islands. Latitude was relatively easy to calculate with some degree of accuracy. With the help of the astrolabe, mariners were able to determine their place on the vertical axis connecting the North and South Poles. The measurement of longitude, however, remained an extremely imprecise matter. Without an accurate measure of distance traveled east or west, mariners were forced to guess at their longitudinal position. Knowing the average crossing time from his African port of departure to the Caribbean, a skilled ship captain paid close attention to the passage of time in the Atlantic and aligned his vessel with the latitude of Barbados when the calendar indicated that it was time to begin searching for the tiny island’s silhouette on the western horizon. Once the ship was in the vicinity of the island (if calculations and guesswork had proved accurate), a well-trained watchman’s eye would enable him to complete the navigational task. Even after it was positioned in the correct latitude, the vessel that carried Colt to Barbados in 1631 “tacked warily for several days” before the island was sighted “in the small hours of the morning” by a lookout whose cry of “Land!” signaled the successful completion of the voyage, forty days after the departure from England.50 Barbados, because of its easterly position, was the first available landfall in the Americas. For this reason, whatever the final destination, most slavers crossing the Atlantic sailed first to Barbados. There, captains, even if bound 48  Ibid., f. 79. 49  Richard S.  Dunn, Sugar and Slaves: The Rise of the Planter Class in the English West Indies, 1624–1713 (Chapel Hill: University of North Carolina Press, 1972; reprint, New York: Norton, 1973), p. 5. 50  Ibid.

1212

Smallwood

for Jamaica or the Chesapeake colonies, could take on provisions when their own supplies were either spent or too spoiled to use. When the Lady Francis reached Barbados from New Calabar in 1678, for instance, the slaves aboard were given 1,000 pounds of potatoes, 4,000 limes, and 84 gallons of rum before putting out to sea again en route to Virginia.51 The 240 Africans aboard the Marygold received 2,450 pounds of potatoes, 28 gallons of rum, and an allotment of tobacco when the vessel arrived at Barbados in May 1677.52 Even vessels headed for the nearby Leeward Islands might occasionally stop first at Barbados. The Dragon took onboard 31 bushels of peas, 1,288 pounds of fish, 425 pounds of potatoes, 63 gallons of rum, and 4,000 limes at Barbados before sailing for Nevis in 1680.53 The stop at Barbados was often especially necessary when ship captains had not taken advantage of the opportunity to fortify their store of provisions at São Tomé or one of the other islands in the Gulf of Guinea, before undertaking to cross the long stretch of open sea to the west. After a voyage of nine weeks from Cape Coast Castle, the Blossom reached Barbados on 29 May 1679, “in pretty good Condition,” but in need of “water and Refreshments for the Negroes, not having touched at the Islands.” After one week’s stay, the vessel continued on its way to Virginia with 244 Africans aboard.54 51  Invoices of Goods sent home from the Factors abroad, Invoice Books Homewards of the Royal African Company of England [hereafter Invoice Books, Homeward], 25 May 1678, T70/937, f. 118. 52  Invoice Books, Homeward, 24 May 1677; 14 June 1677, T70/937, T70/937, f. 70v, ff. 79v–80. Scurvy is a disease associated with malnutrition, caused by a deficiency of vitamin C. The difficulty of maintaining a steady supply of fresh fruits and vegetables on long sea voyages made scurvy a chronic condition in the maritime arena, one whose cause was not formally understood until the publication of Dr. James Lind’s “A Treatise of the Scurvy” in 1753. Nonetheless, the curative effect of citrus fruits was widely observed and noted throughout the Atlantic (and Mediterranean) maritime worlds from an earlier date. The requisite use of limes to provision RAC ships suggests that English slave traders may well have possessed incidental awareness (if not full understanding) of the causal relationship. See Miller, Way of Death, p. 435, who notes that in the eighteenth-century Portuguese slave trade, at least, citrus and other forms of fresh produce were used more as a remedy after symptoms of scurvy had appeared than as a prophylactic or a strategy of prevention. See also Robert Harms, The Diligent: A Voyage through the Worlds of the Slave Trade (New York: Basic, 2002), p. 317, on theories that circulated, prior to Lind’s work, about the cause of scurvy; and Kiple, Caribbean Slave, pp. 59–60, 90–91. 53  Invoice Books, Homeward, 7 May 1680, T70/938, f. 133. The Leeward Islands lay three hundred miles to the northwest of Barbados, the voyage there from Barbados took around three days. Dunn, Sugar and Slaves, pp. 6–7. 54  Edwyn Stede and Stephen Gascoigne to RAC, 10 June 1679, T70/15, ff. 17v–18; Invoice Books, Homeward, 14 June 1679, T70/938, f. 73v.

The Living Dead Aboard The Slave Ship At Sea

1213

Just days after the Blossom departed, another Virginia-bound vessel entered the harbor. Once laden with 179 Africans at New Calabar, the Swallow had “touched for refreshments” at Annabon Island (southwest of São Tomé). Nonetheless, the “yeams being all rotten” by the time the captain reached the Americas, he was forced to stop at Barbados. There, company agents agreed to “furnish them with all speed with such provisions as this country at present affords, and … dispatch them with all speed to their designed port of Virginia,” putting on board 52 bushels of peas, 2,278 pounds of flour, and 96.5 gallons of rum.55 Standing alone in the eastern Caribbean Sea and having a markedly low “silhouette,” Barbados might easily be missed by European mariners anxiously searching the horizon for its outline. So Africans perhaps did not take note of the impending landfall until the island swung into view. By the time daylight came, the startling sight of the coastline, “a white line of breakers, sand, and rocks, with wooded land rising steeply beyond,” had replaced the endless expanse of the open sea.56 As the sun climbed in its familiar path, the panorama spread out before the captives aboard the James filled with detail. From the sparsely inhabited eastern shore of the island, the ship passed around Easting’s Point, and the shoreline grew thicker with dwellings, until finally the ship stood athwart a busy waterside town hugging the edge of a wide bay, called by the English after the Earl of Carlisle.57 Sale of the captives delivered on the James began four days after the ship’s arrival on 25 May, by which date four more captives were dead. By the time the seventeen unsold “refuse” slaves left the ship on 6 June, another seven persons had expired. In all, fifty-one had perished aboard the James during the nine months it had served as a waterborne prison for African captives. On average, 20 percent of the Africans carried into the Atlantic in the seventeenth century died at sea, and 40 percent of cargoes experienced mortality levels above that benchmark.58 Seen through this lens, mortality among the captives aboard the James appears notably benign. Since captivity on the James proved considerably less deadly than aboard other ships navigating the Atlantic at this time, the fifty-one people who died on the ship leave only a faint mark in the slave traders’ comparative calculation of loss. But look again. Equally convincing was the competing narrative framed by the captives themselves, those aboard the slave ship in transatlantic motion. 55  Ibid. 56  Dunn, Sugar and Slaves, p. 5, describing Colt’s first view of the eastern shore of Barbados. 57  Ibid., p. 49. 58  Engerman and Klein, “Experiences on Slave Ships,” p. 26, n. 8.

1214

Smallwood

They also evaluated death on the Atlantic, but by very different systems of value and measurement. Through the words he entered in his journal and his account of mortality, Blake unwittingly offers a glimpse of the captives’ experience. Like epitaphs, his words convey the traumatic content that shaped the captives’ own “account”—a man’s determination to commit his own body to the sea (and confront the danger that action entailed for his soul), a woman’s unwillingness to abandon her duties as a mother aboard ship, a man’s desperate search for a private place to endure his illness, and others’ profound inability to heal their disintegrating bodies or counter the overwhelming forces that made them so ill. European investors in African human cargoes focused constant attention on the conditions of the slave ship under sail. They tried different strategies for providing food sufficient in quantity and quality, worried about the purity of water supplies, fretted about crowding (“tight packing”), disease, depression, and physical atrophy. Each of these being a problem demanding a rational solution, the investors devised strategies to cope with all these concerns: extra provisions, methods for protecting water from contamination, mandatory exercise, regulations on the number of captives a ship could carry, inoculation against smallpox when the procedure became available early in the eighteenth century, and even recourse to indigenous African healing practices. Indeed, so pressing was European anxiety and effort to manage the innumerable risks that attended the business of slaving that it is easy to confuse European interest in preserving life to prevent economic loss with positive concern for the captives’ human welfare. But to interpret the regime of the slave ship in that way is to be duped by the slave traders’ rhetoric—a language of concealment that allowed European slaving concerns to portray themselves as passive and powerless before the array of forces (including the agency of the captives themselves) outside their control.59 Slave merchants and their backers disguised from themselves the ugly truth that the Atlantic regime of commodification took captives from fully realized humanity and suspended them in a purgatory in between tenuous life and dishonorable death. The James held

59  Orlando Patterson, Slavery and Social Death: A Comparative Study (Cambridge, Mass.: Harvard University Press, 1982), p. 18, on concealment as one of “two polar extremes” according to which the social (as distinct from the conceptual) “idiom of power” is articulated.

The Living Dead Aboard The Slave Ship At Sea

1215

a collective of 423 people in psychic terror over a journey in which some-one died fifty-one times. Each thwarted life, each departure, was distinct in its details, and each was connected to and compounded by the accumulating residue of those which had preceded and by those which would follow. Mortality on the Atlantic produced a crisis of enormous proportions, as Africans labored under the cumulative weight of these deaths that remained unresolved. For the Akan captives, the James was the site of a relentless accumulation of incomplete deaths, each one holding its own tragic meanings. Entrapped, Africans confronted a dual crisis: the trauma of death, and the inability to respond appropriately to death. This indirect violence, arguably, was the most abject experience of the captives’ Atlantic crossing. The cost for Africans of that crossing cannot be adequately represented by any statistics. More fundamentally, on the sea voyage, even the African dead were enslaved and commodified, trapped in a time-space regime in which they were unable fully to die.

From Creole to African: Atlantic Creoles and the Origins of African-American Society in Mainland North America Ira Berlin In 1727, Robert “King” Carter, the richest planter in Virginia, purchased a handful of African slaves from a trader who had been cruising the Chesapeake. The transaction was a familiar one to the great planter, for Carter owned hundreds of slaves and had inspected many such human cargoes, choosing the most promising from among the weary, frightened men and women who had survived the transatlantic crossing. Writing to his overseer from his plantation on the Rappahannock River, Carter explained the process by which he initiated Africans into their American captivity. “I name’d them here & by their names we can always know what sizes they are of & I am sure we repeated them so often to them that every one knew their name & would readily answer to them.” Carter then forwarded his slaves to a satellite plantation or quarter, where his overseer repeated the process, taking “care that the negros both men & women I sent … always go by the names we gave them.” In the months that followed, the drill continued, with Carter again joining in the process of stripping newly arrived Africans of the signature of their identity.1 Renaming marked Carter’s initial endeavor to master his new slaves by separating them from their African inheritance. For the most part, he designated them by common English diminutives—Tom, Jamey, Moll, Nan—as if Source: Berlin, Ira, “From Creole to African: Atlantic Creoles and the Origins of African-American Society in Mainland North America,” William & Mary Quarterly, 53(2) (1996): 251–288. Published by the Omohundro Institute of Early American History and Culture. * Ira Berlin is professor of history and dean of the College of Arts and Humanities at the University of Maryland, College Park. A version of this article was presented as the E. P. Thompson lecture at the University of Pittsburgh, to the Department of History at Carnegie-Mellon University, and to the Washington-Area Seminar in Early American History. I would like to thank Martha Berlin, W. Jeffrey Bolster, William Bravman, Steven Hahn, Ronald Hoffman, Patrick Manning, Joseph Miller, Marie Perinbam, Marcus Rediker, Sarah Russell, Marie Schwartz, Gabrielle Speigel, John Thornton, and James Walvin for helpful suggestions. 1  Carter to Robert Jones, Oct. 10, 1727 [misdated 1717], Oct. 24, 1729, quoted in Lorena S. Walsh, “A ‘Place in Time’ Regained: A Fuller History of Colonial Chesapeake Slavery through Group Biography,” in Larry E. Hudson, Jr., ed., Working toward Freedom: Slave Society and the Domestic Economy in the American South (Rochester, N.Y., 1994), 14. © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_039

From Creole To African

1217

to consign them to a permanent childhood. But he tagged some with names more akin to barnyard animals—Jumper, for example—as if to represent their distance from humanity, and he gave a few the names of some ancient deity or great personage like Hercules or Cato as a kind of cosmic jest: the most insignificant with the greatest of names. None of his slaves received surnames, marks of lineage that Carter sought to obliterate and of adulthood that he would not admit.2 The loss of their names was only the first of the numerous indignities Africans suffered at the hands of planters in the Chesapeake. Since many of the skills Africans carried across the Atlantic had no value to their new owners, planters disparaged them, and since the Africans’ “harsh jargons” rattled discordantly in the planters’ ears, they ridiculed them. Condemning new arrivals for the “gross bestiality and rudeness of their manners, the variety and strangeness of their languages, and the weakness and shallowness of their minds,” planters put them to work at the most repetitive and backbreaking tasks, often on the most primitive, frontier plantations. They made but scant attempt to see that slaves had adequate food, clothing, or shelter, because the open slave trade made slaves cheap and the new disease environment inflated their mortality rate, no matter how well they were tended. Residing in sex-segregated barracks, African slaves lived a lonely existence, without families or ties of kin, isolated from the mainstream of Chesapeake life.3 2  For the names of Carter’s slaves see the Carter Papers, Alderman Library, University of Virginia, Charlottesville. The naming of Chesapeake slaves is discussed in Allan Kulikoff, Tobacco and Slaves: The Development of Southern Cultures in the Chesapeake, 1680–1800 (Chapel Hill, 1986), 325–26, and John Thornton, “Central African Names and African-American Naming Patterns,” William and Mary Quarterly, 3d Ser., 50 (1993), 727–42. For surnames see Walsh, “A ‘Place in Time’ Regained,” 26–27 n. 18. The pioneering work on this subject is Peter H. Wood, Black Majority: Negroes in Colonial South Carolina from 1670 through the Stono Rebellion (New York, 1974), 181–86. Henry Laurens, the great South Carolina slave trader and planter, followed a similar routine in naming his slaves; Philip Morgan, “Three Planters and Their Slaves: Perspectives on Slavery in Virginia, South Carolina, and Jamaica, 1750–1790,” in Winthrop D. Jordan and Sheila L. Skemp, eds., Race and Family in the Colonial South (Jackson, Miss., 1987), 65. 3  Gerald W.  Mullin, Flight and Rebellion: Slave Resistance in Eighteenth-Century Virginia (New York, 1972), chaps. 1–3; Kulikoff, Tobacco and Slaves, esp. 319–34; Russell R. Menard, “The Maryland Slave Population, 1658 to 1730: A Demographic Profile of Blacks in Four Counties,” WMQ, 3d Ser., 32 (1975), 29–54; Lois Green Carr and Walsh, “Economic Diversification and Labor Organization in the Chesapeake, 1650–1820,” in Stephen Innes, ed., Work and Labor in Early America (Chapel Hill, 1988), 144–88. Quotation in Hugh Jones, The Present State of Virginia, from Whence Is Inferred a Short View of Maryland and North Carolina, ed. Richard L. Morton (Chapel Hill, 1956), 36–38, and Philip Alexander Bruce, Institutional History of

1218

Berlin

So began the slow, painful process whereby Africans became AfricanAmericans. In time, people of African descent recovered their balance, mastered the circumstances of their captivity, and confronted their owners on more favorable terms. Indeed, resistance to the new regime began at its inception, as slaves clandestinely maintained their African names even as they answered their owner’s call.4 The transition of Africans to African-Americans or creoles5—which is partially glimpsed in the records of Carter’s estate— would be repeated thousands of times, as African slavers did the rough business of transporting Africa to America. While the transition was different on the banks of the Hudson, Cooper, St. Johns, and Mississippi rivers than on the Rappahannock, the scenario by which “outlandish” Africans progressed from “New Negroes” to assimilated African-Americans has come to frame the history of black people in colonial North America.6 Virginia in the Seventeenth Century …, 2 vols. (New York, 1910), 1:9. See a slightly different version in “The Journal of the General Assembly of Virginia,” June 2, 1699, in W. N. Sainsbury et al., eds., Calendar of State Papers, Colonial Series, America and West Indies, 40 vols. (London, 1860–1969), 17:261. 4  In the summer of 1767, when two slaves escaped from a Georgia plantation, their owner noted that one “calls himself GOLAGA,” although “the name given him [was] ABEL,” and the other “calls himself ABBROM, the name given him here BENNET.” For evidence that the practice had not ended by 1774 see Lathan A. Windley, comp., Runaway Slave Advertisements: A Documentary History from the 1730s to 1790, 4 vols. (Westport, Conn., 1983), 4:22 ([Savannah] Georgia Gazette, June 3, 1767), 62 (ibid., Apr. 19, 1775). 5  “Creole” derives from the Portuguese crioulo, meaning a person of African descent born in the New World. It has been extended to native-born free people of many national origins, including both Europeans and Africans, and of diverse social standing. It has also been applied to people of partly European but mixed racial and national origins in various European colonies and to Africans who entered Europe. In the United States, creole has also been specifically applied to people of mixed but usually non-African origins in Louisiana. Staying within the bounds of the broadest definition of creole and the literal definition of African American, I use both terms to refer to black people of native American birth; John A. Holm, Pidgins and Creoles: Theory and Structure, 2 vols. (Cambridge, 1988–1989), 1:9. On the complex and often contradictory usage in a single place see Gwendolyn Midlo Hall, Africans in Colonial Louisiana: The Development of Afro-Creole Culture in the Eighteenth Century (Baton Rouge, 1992), 157–59, and Joseph G. Tregle, Jr., “On that Word ‘Creole’ Again: A Note,” Louisiana History, 23 (1982), 193–98. 6   See, for example, Mullin, Flight and Rebellion, and Mullin, Africa in America: Slave Acculturation and Resistance in the American South and the British Caribbean, 1736–1831 (Urbana, Ill., 1992), 268, which examines the typology of “African” and “creole” from the perspective of resistance. See also the 3 stages of black community development proposed by Kulikoff, “The Origins of Afro-American Society in Tidewater Maryland and Virginia, 1700–1790,” WMQ, 3d Ser., 35 (1978), 226–59, esp. 229, and expanded in his Tobacco and Slaves,

From Creole To African

1219

Important as that story is to the development of black people in the plantation era, it embraces only a portion of the history of black life in colonial North America, and that imperfectly. The assimilationist scenario assumes that “African” and “creole” were way stations of generational change rather than cultural strategies that were manufactured and remanufactured and that the vectors of change moved in only one direction—often along a single track with Africans inexorably becoming creoles. Its emphasis on the emergence of the creole—a self-sustaining, indigenous population—omits entirely an essential element of the story: the charter generations, whose experience, knowledge, and attitude were more akin to that of confident, sophisticated natives than of vulnerable newcomers.7 Such men and women, who may be termed “Atlantic creoles”8 from their broad experience in the Atlantic world, flourished prior to chaps. 8–10. Although the work of Sidney Mintz and Richard Price, which has provided the theoretical backbone for the study of African acculturation in the New World, begins by breaking with models of cultural change associated with “assimilation” and, indeed, all notions of social and cultural change that have a specific end point, it too frames the process as a progression from African to creole; Anthropological Approach to the Afro-American Past: A Caribbean Perspective, Institute for the Study of Human Issues Occasional Papers in Social Change (Philadelphia, 1976). Others have followed, including those sensitive to the process of re-Africanization. See, for example, Wood, Black Majority; Kulikoff, Tobacco and Slaves; Ira Berlin and Ronald Hoffman, eds., Slavery and Freedom in the Age of Revolution (Charlottesville, 1983); and Berlin, “Time, Space, and the Evolution of Afro-American Society on British Mainland North America,” American Historical Review, 85 (1980), 44–78. Thornton’s work represents an important theoretical departure. He distinguishes between the African and the Atlantic experiences, maintaining the “Atlantic environment was … different from the African one.” He extends the Atlantic environment to the African littoral as well as the Americas, in Africa and Africans in the Making of the Atlantic World, 1440–1680 (Cambridge, 1992), quotation on 211. 7  The use of charter groups draws on T. H. Breen, “Creative Adoptions: Peoples and Cultures,” in Jack P. Greene and J. R. Pole, eds., Colonial British America: Essays in the New History of the Early Modern Era (Baltimore, 1984), 203–08. Breen, in turn, borrowed the idea from anthropologist John Porter. 8  “Atlantic creole,” employed herein, designates those who by experience or choice, as well as by birth, became part of a new culture that emerged along the Atlantic littoral—in Africa, Europe, or the Americas—beginning in the 16th century. It departs from the notion of “creole” that makes birth definitive (see n. 5 above). Circumstances and volition blurred differences between “African” and “creole” as defined only by nativity, if only because Africans and creoles were connected by ties of kinship and friendship. They worked together, played together, intermarried, and on occasion stood together against assaults on their freedom. Even more important, men and women could define themselves in ways that transcended nativity. “African” and “creole” were as much a matter of choice as of birth. The term “Atlantic creole” is designed to capture the cultural transformation that sometimes preceded g­ enerational

1220

Berlin

the triumph of plantation production on the mainland—the tobacco revolution in the Chesapeake in the last third of the seventeenth century, the rice revolution in the Carolina lowcountry in the first decades of the eighteenth century, the incorporation of the northern colonies into the Atlantic system during the eighteenth century, and finally the sugar revolution in the lower Mississippi Valley in the first decade of the nineteenth century. Never having to face the cultural imposition of the likes of Robert “King” Carter, black America’s charter generations took a different path—despite the presence of slavery and the vilification of slave masters and their apologists. The Atlantic creole’s unique experience reveals some of the processes by which race was constructed and reconstructed in early America. Black life in mainland North America originated not in Africa or America but in the netherworld between the continents. Along the periphery of the Atlantic—first in Africa, then in Europe, and finally in the Americas—AfricanAmerican society was a product of the momentous meeting of Africans and Europeans and of their equally fateful encounter with the peoples of the Americas. Although the countenances of these new people of the Atlantic— Atlantic creoles—might bear the features of Africa, Europe, or the Americas in whole or in part, their beginnings, strictly speaking, were in none of those places. Instead, by their experiences and sometimes by their persons, they had become part of the three worlds that came together along the Atlantic littoral. Familiar with the commerce of the Atlantic, fluent in its new languages, and intimate with its trade and cultures, they were cosmopolitan in the fullest sense. Atlantic creoles originated in the historic meeting of Europeans and Africans on the west coast of Africa. Many served as intermediaries, employing their linguistic skills and their familiarity with the Atlantic’s diverse commercial practices, cultural conventions, and diplomatic etiquette to mediate between African merchants and European sea captains. In so doing, some Atlantic creoles identified with their ancestral homeland (or a portion of it)— be it African, European, or American—and served as its representatives in negotiations with others. Other Atlantic creoles had been won over by the power and largesse of one party or another, so that Africans entered the employ of European trading companies and Europeans traded with African potentates. change and sometimes was unaffected by it. Insightful commentary on the process of creolization is provided by Mintz, “The Socio-Historical Background to Pidginization and Creolization,” in Dell Hymes, ed., Pidginization and Creolization of Languages: Proceedings of a Conference Held at the University of the West Indies, Mona, Jamaica, April 1968 (Cambridge, 1971), 481–96.

From Creole To African

1221

Yet others played fast and loose with their diverse heritage, employing whichever identity paid best. Whatever strategy they adopted, Atlantic creoles began the process of integrating the icons and ideologies of the Atlantic world into a new way of life.9 The emergence of Atlantic creoles was but a tiny outcropping in the massive social upheaval that accompanied the joining of the peoples of the two hemispheres. But it represented the small beginnings that initiated this monumental transformation, as the new people of the Atlantic made their presence felt. Some traveled widely as blue-water sailors, supercargoes, shipboard servants, and interpreters—the last particularly important because Europeans showed little interest in mastering the languages of Africa. Others were carried— sometimes as hostages—to foreign places as exotic trophies to be displayed before curious publics, eager for firsthand knowledge of the lands beyond the sea. Traveling in more dignified style, Atlantic creoles were also sent to distant lands with commissions to master the ways of newly discovered “others” and to learn the secrets of their wealth and knowledge. A few entered as honored guests, took their places in royal courts as esteemed councilors, and married into the best families.10 9  For ground-breaking works that argue for the unity of working peoples in the Atlantic world see Peter Linebaugh, “All the Atlantic Mountains Shook,” Labour/Le Travailleur, 10 (1982), 82–121, and Linebaugh and Marcus Rediker, “The Many-Headed Hydra: Sailors, Slaves, and the Atlantic Working Class in the Eighteenth Century,” Journal of Historical Sociology, 3 (1990), 225–52. Thornton, Africa and Africans in the Making of the Atlantic World, adopts a similar perspective in viewing the making of African-American culture. A larger Atlantic perspective for the formation of black culture is posed in Paul Gilroy, The Black Atlantic: Modernity and Double Consciousness (Cambridge, Mass., 1993). 10  A. C. de C. M. Saunders, A Social History of Black Slaves and Freedmen in Portugal, 1441–1555 (Cambridge, 1982), 11–12, 145, 197 n. 52, 215 n. 73 (for black sailors and interpreters in the African trade); P. E. H. Hair, “The Use of African Languages in Afro-European Contacts in Guinea, 1440–1560,” Sierra Leone Language Review, 5 (1966), 7–17 (for black interpreters and Europeans’ striking lack of interest in mastering African languages); George E. Brooks, Landlords and Strangers: Ecology, Society, and Trade in West Africa, 1000–1630 (Boulder, Colo., 1993), chap. 7 (particularly for the role of grumetes), 124, 136–37; Wyatt MacGaffey, “Dialogues of the Deaf: Europeans on the Atlantic Coast of Africa,” in Stuart B. Schwartz, ed., Implicit Understandings: Observing, Reporting, and Reflecting on the Encounters between Europeans and Other Peoples in the Early Modern Era (Cambridge, 1994), 252 (for hostages); Kwame Yeboa Daaku, Trade and Politics on the Gold Coast, 1600–1720: A Study of the African Reaction to European Trade (Oxford, 1970), chap. 5, esp. 96–97 (for ambassadors to the United Provinces in 1611); Anne Hilton, The Kingdom of Kongo (Oxford, 1985), 64; Paul Edwards and James Walvin, “Africans in Britain, 1500–1800,” in Martin L. Kilson and Robert I. Rotberg, eds., The African Diaspora: Interpretive Essays (Cambridge, Mass.,

1222

Berlin

Atlantic creoles first appeared at the trading feitorias or factories that European expansionists established along the coast of Africa in the fifteenth century. Finding trade more lucrative than pillage, the Portuguese crown began sending agents to oversee its interests in Africa. These official representatives were succeeded by private entrepreneurs or lançados, who established themselves with the aid of African potentates, sometimes in competition with the crown’s emissaries. European nations soon joined in the action, and coastal factories became sites of commercial rendezvous for all manner of transatlantic traders. What was true of the Portuguese enclaves (Axim and Elmina) held for those later established or seized by the Dutch (Fort Nassau and Elmina), Danes (Fredriksborg and Christiansborg), Swedes (Karlsborg and Cape Apolina), Brandenburgers (Pokoso), French (St. Louis and Gorée), and English (Fort Kormantse and Cape Coast).11 The transformation of the fishing villages along the Gold Coast during the sixteenth and seventeenth centuries suggests something of the change wrought by the European traders. Between 1550 and 1618, Mouri (where the Dutch constructed Fort Nassau in 1612) grew from a village of 200 people to 1,500 and to an estimated 5,000–6,000 at the end of the eighteenth century. In 1555, Cape Coast counted only twenty houses; by 1680, it had 500 or more. Axim, with 500 inhabitants in 1631, expanded to between 2,000 and 3,000 by 1690.12 Small but growing numbers of Europeans augmented the African fishermen, craftsmen, village-based peasants, and laborers who made up the 1976), 173–205 (for African royalty sending their sons to be educated in Europe). See also Shelby T. McCloy, “Negroes and Mulattoes in Eighteenth-Century France,” Journal of Negro History, 30 (1945), 276–92. For the near-seamless, reciprocal relationship between the Portuguese and the Kongolese courts in the 16th century see Thornton, “Early KongoPortuguese Relations, 1483–1575: A New Interpretation,” History in Africa, 8 (1981), 183–204. 11  For an overview see Thornton, Africa and Africans in the Making of the Atlantic World, chap. 2, esp. 59–62. See also Daaku, Trade and Politics on the Gold Coast, chap. 2; Brooks, Landlords and Strangers, chaps. 7–8 (see the Portuguese crown’s penalties against lançados for illegal trading, 152–54); Philip D. Curtin, Economic Change in Precolonial Africa: Senegambia in the Era of the Slave Trade (Madison, Wis., 1975), chap. 3; Ray A. Kea, Settlements, Trade, and Polities in the Seventeenth-Century Gold Coast (Baltimore, 1982); John Vogt, Portuguese Rule on the Gold Coast, 1469–1682 (Athens, Ga., 1979); C. R. Boxer, Four Centuries of Portuguese Expansion, 1415–1825: A Succinct Survey (Johannesburg, 1961); and Boxer, The Dutch Seaborne Empire, 1600–1800 (New York, 1965). Lançados comes from a contraction of lançados em terra (to put on shore); Curtin, Economic Change in Precolonial Africa, 95. As the influence of the Atlantic economy spread to the interior, Atlantic creoles appeared in the hinterland, generally in the centers of trade along rivers. 12  Kea, Settlements, Trade, and Polities, chap. 1, esp. 38.

From Creole To African

1223

population of these villages. Although mortality and transiency rates in these enclaves were extraordinarily high, even by the standards of early modern ports, permanent European settlements developed from a mobile body of the corporate employees (from governors to surgeons to clerks), merchants and factors, stateless sailors, skilled craftsmen, occasional missionaries, and sundry transcontinental drifters.13 Established in 1482 by the Portuguese and captured by the Dutch in 1637, Elmina was one of the earliest factories and an exemplar for those that followed. A meeting place for African and European commercial ambitions, Elmina—the Castle São Jorge da Mina and the town that surrounded it—became headquarters for Portuguese and later Dutch mercantile activities on the Gold Coast and, with a population of 15,000 to 20,000 in 1682, the largest of some two dozen European outposts in the region.14 The peoples of the enclaves—both long-term residents and wayfarers— soon joined together genetically as well as geographically. European men took African women as wives and mistresses, and, before long, the offspring of these unions helped people the enclave. Elmina sprouted a substantial cadre of EuroAfricans (most of them Luso-Africans)—men and women of African birth but shared African and European parentage, whose combination of swarthy skin, European dress and deportment, knowledge of local customs, and multilingualism gave them inside understanding of both African and European ways while denying them full acceptance in either culture. By the eighteenth century, they numbered several hundred in Elmina. Farther south along the coast of Central Africa, they may have been even more numerous.15 13  Ibid.; Vogt, Portuguese Rule on the Gold Coast; Harvey M. Feinberg, Africans and Europeans in West Africa: Elminans and Dutchmen on the Gold Coast during the Eighteenth Century, American Philosophical Society, Transactions, 79, No. 7 (Philadelphia, 1989). For mortality see Curtin, “Epidemiology and the Slave Trade,” Political Science Quarterly, 83 (1968), 190–216, and K. G. Davies, “The Living and the Dead: White Mortality in West Africa, 1684–1732,” in Stanley L. Engerman and Eugene D. Genovese, eds., Race and Slavery in the Western Hemisphere: Quantitative Studies (Princeton, 1975), 83–98. 14  Kea, Settlements, Trade, and Polities, chap. 1, esp. 38–50, 133–34; Vogt, Portuguese Rule on the Gold Coast; Feinberg, Africans and Europeans in West Africa. Eveline C. Martin, The British West African Settlements, 1750–1821: A Study in Local Administration (New York, 1927), and Margaret Priestley, West African Trade and Coast Society: A Family Study (London, 1969), describe the English enclaves in the 18th and 19th centuries, casting light on their earlier development. 15  Brooks, Landlords and Strangers, chaps. 7–9, esp. 188–96, and Brooks, “Luso-African Commerce and Settlement in the Gambia and Guinea-Bissau Region,” Boston University African Studies Center Working Papers (1980), for the connection of the Luso-Africans

1224

Berlin

People of mixed ancestry and tawny complexion composed but a small fraction of the population of the coastal factories, yet few observers failed to note their existence—which suggests something of the disproportionate significance of their presence. Africans and Europeans alike sneered at the creoles’ mixed lineage (or lack of lineage) and condemned them as knaves, charlatans, and shameless self-promoters. When they adopted African ways, wore African dress and amulets, and underwent ritual circumcision and scarification, Europeans declared them outcasts (tangomãos, renegades, to the Portuguese). When they adopted European ways, wore European clothing and crucifixes, employed European names or titles, and comported themselves in the manner of “white men,” Africans denied them the right to hold land, marry, and inherit property. Yet, although tangomãos faced reproach and proscription, all parties conceded that they were shrewd traders, attested to their mastery of the fine points of intercultural negotiations, and found advantage in dealing with them. Despite their defamers, some rose to positions of wealth and

with the Cape Verde Islands; Daaku, Trade and Politics on the Gold Coast, chaps. 5–6; Vogt, Portuguese Rule on the Gold Coast, 154; Feinberg, Africans and Europeans in West Africa, 32, 88–90; Curtin, Economic Change in Precolonial Africa, 95–100, 113–21 (for AfroFrench). For the development of a similar population in Angola see Joseph C. Miller, Way of Death: Merchant Capitalism and the Angolan Slave Trade, 1730–1830 (Madison, Wis., 1988), esp. chaps. 8–9, and Miller, “A Marginal Institution on the Margin of the Atlantic System: The Portuguese Southern Atlantic Slave Trade in the Eighteenth Century,” in Barbara L. Solow, ed., Slavery and the Rise of the Atlantic System (Cambridge, Mass., 1991), 125, 128–29. By the mid-17th century, the hierarchy of Kongolese Catholics was largely mixed African and European ancestry or pombeiros; Hilton, Kingdom of Kongo, 140–41, 154. See also Allen F. Isaacman, Mozambique: The Africanization of a European Institution: The Zambezi Prazos, 1750–1902 (Madison, Wis., 1972). The number of such individuals in west Africa is difficult to estimate. Brooks, in his study of the Grain Coast and its interior, estimates “hundreds of Portuguese and Cabo Verdean traders were admitted to western African communities by the close of the fifteenth century.” Probably the same could be said for other portions of the African coast at that time. By the middle of the 16th century, Atlantic creoles were more numerous. In 1567, when the English adventurer John Hawkins launched a raid on an African settlement on the Cacheu River, he was repulsed by a force that included “about a hundred” lançados; Brooks, Landlords and Strangers, 137, 230–31. By the 19th century, the Afro-Europeans had become to a “remarkable extent soundly and politically integrated” and “occupied their own ‘quarter’ of the town” of Elmina; Larry W. Yarak, “West African Coastal Slavery in the Nineteenth Century: The Case of Afro-European Slaveowners of Elmina,” Ethnohistory, 36 (1989), 44–60, quotation on 47; J. T. Lever, “Mulatto Influence on the Gold Coast in the Early Nineteenth Century: Jan Nieser of Elmina,” African Historical Studies, 3 (1970), 253–61.

From Creole To African

1225

power, compensating for their lack of lineage with knowledge, skill, and entrepreneurial derring-do.16 Not all tangomãos were of mixed ancestry, and not all people of mixed ancestry were tangomãos. Color was only one marker of this culture-in-the-making, and generally the least significant.17 From common experience, conventions of personal behavior, and cultural sensibilities compounded by shared ostracism and mercantile aspirations, Atlantic creoles acquired interests of their own, apart from their European and African antecedents. Of necessity, Atlantic creoles spoke a variety of African and European languages, weighted strongly toward Portuguese. From the seeming babble emerged a pidgin that enabled Atlantic creoles to communicate widely. In time, their pidgin evolved into creole, borrowing its vocabulary from all parties and creating a grammar unique unto itself. Derisively called “fala de Guine” or “fala de negros”—“Guinea speech” or “Negro Speech”—by the Portuguese and “black Portuguese” by others, this creole language became the lingua franca of the Atlantic.18 16  Daaku, Trade and Politics on the Gold Coast, chaps. 4–5; Brooks, Landlords and Strangers, chaps. 7–9, esp. 188–96; Curtin, Economic Change in Precolonial Africa, 95–100. See also Miller’s compelling description of Angola’s Luso-Africans in the 18th and 19th centuries that suggests something of their earlier history, in Way of Death, 246–50. Brooks notes the term tangomãos passed from use at the end of the 17th century, in “Luso-African Commerce and Settlement in the Gambia and Guinea-Bissau,” 3. 17  Speaking of the Afro-French in Senegambia in the 18th century, Curtin emphasizes the cultural transformation in making this new people, noting that “the important characteristic of this community was cultural mixture, not racial mixture, and the most effective of the traders from France were those who could cross the cultural line between Europe and Africa in their commercial relations,” in Economic Change in Precolonial Africa, 117. 18  Holm, Pidgins and Creoles; Thornton, Africa and Africans in the Making of the Atlantic World, 213–18; Saunders, Black Slaves and Freedman in Portugal, 98–102 (see special word—ladinhos—for blacks who could speak “good” Portuguese, 101); Brooks, Landlords and Strangers, 136–37. See also Robert A. Hall, Jr., Pidgin and Creole Languages (Ithaca, 1966); David Dalby, “The Place of Africa and Afro-America in the History of the English Language,” African Language Review, 9 (1971), 280–98; Hair, “Use of African Languages in Afro-European Contacts in Guinea,” 5–26; Keith Whinnom, “Contacts De Langues et Emprunts Lexicaux: The Origin of the European-Based Pidgins and Creoles,” Orbis, 14 (1965), 509–27, Whinnom, “Linguistic Hybridization and the ‘Special Case’ of Pidgins and Creoles,” in Hymes, ed., Pidginization and Creolization of Languages, 91–115, and Whinnom, “The Context and Origins of Lingua Franca,” in Jürgen M. Meisel, ed., Langues en Contact—Pidgins—Creoles (Tübingen, 1975); and J. L. Dillard, “Creole English and Creole Portuguese: The Early Records,” in Ian F. Hancock, ed., Readings in Creole Studies (Ghent, 1979), 261–68. For another theory on the origins of the west African pidgin see Anthony J. Naro, “The Origins of West African Pidgin,” in Claudia Corum, T. Cedric

1226

Berlin

Although jaded observers condemned the culture of the enclaves as nothing more than “whoring, drinking, gambling, swearing, fighting, and shouting,” Atlantic creoles attended church (usually Catholic), married according to the sacraments, raised children conversant with European norms, and drew a livelihood from their knowledge of the Atlantic commercial economy. In short, they created societies of their own, of but not always in, the societies of the Africans who dominated the interior trade and the Europeans who controlled the Atlantic trade. Operating under European protection, always at African sufferance, the enclaves developed governments with a politics as diverse and complicated as the peoples who populated them and a credit system that drew on the commercial centers of both Europe and Africa. Although the trading castles remained under the control of European metropoles, the towns around them often developed independent political lives—separate from both African and European domination. Meanwhile, their presence created political havoc, enabling new men and women of commerce to gain prominence and threatening older, often hereditary elites. Intermarriage with established peoples allowed creoles to construct lineages that gained them full membership in local elites, something that creoles eagerly embraced. The resultant political turmoil promoted state formation along with new class relations and ideologies.19 New religious forms emerged and then disappeared in much the same manner, as Europeans and Africans brought to the enclaves not only their commercial and political aspirations but all the trappings of their cultures as well. Priests and ministers sent to tend European souls made African converts, some of whom saw Christianity as both a way to ingratiate themselves with their trading partners and a new truth. Missionaries sped the process of christianization and occasionally scored striking successes. At the beginning of the sixteenth century, the royal house of Kongo converted to Christianity. Catholicism, in various syncretic forms, infiltrated the posts along the Angolan coast and spread northward. Islam filtered in from the north. Whatever the sources of the new religions, most converts saw little cause to surrender their own deities. They incorporated Christianity and Islam to serve their own needs and gave Jesus and Mohammed a place in their spiritual pantheon. New religious practices, polities, and theologies emerged from the mixing of Christianity, Islam, Smith-Stark, and Ann Weiser, eds., Papers from the Ninth Regional Meeting, Chicago Linguistic Society (1973), 442–49. 19  Daaku, Trade and Politics on the Gold Coast, chaps. 3–4; Feinberg, Africans and Europeans in West Africa, chap. 6, quotation on 86; Kea, Settlements, Trade, and Polities, esp. pt. 2; Curtin, Economic Change in Precolonial Africa, 92–93.

From Creole To African

1227

polytheism, and animism. Similar syncretic formations influenced the agricultural practices, architectural forms, and sartorial styles as well as the cuisine, music, art, and technology of the enclaves.20 Like the stone fortifications, these cultural innovations announced the presence of something new to those arriving on the coast, whether they came by caravan from the African interior or sailed by caravel from the Atlantic. Outside the European fortifications, settlements—the town of Elmina as opposed to Castle São Jorge da Mina, for example—expanded to provision and refresh the European-controlled castles and the caravels and carracks that frequented the coast. In time, they developed economies of their own, with multifarious systems of social stratification and occupational differentiation. Residents included canoemen who ferried goods between ships and shore; longshoremen and warehousemen who unloaded and stored merchandise; porters, messengers, guides, interpreters, factors, and brokers or makelaers (to the Dutch) who facilitated trade; inn keepers who housed country traders; skilled workers of all sorts; and a host of peddlers, hawkers, and petty traders. Others chopped wood, drew water, prepared food, or supplied sex to the lonely men who visited these isolated places. African notables occasionally established residence, bringing with them the trappings of wealth and power: wives, clients, pawns, slaves, and other dependents. In some places, small manufactories grew up, like the salt pans, boatyards, and foundries on the outskirts of Elmina, to supply the town and service the Atlantic trade. In addition, many people lived outside the law; the rough nature and transient population of these crossroads of trade encouraged roguery and brigandage.21 20  Vogt, Portuguese Rule on the Gold Coast, 54–58; Daaku, Trade and Politics on the Gold Coast, 99–101; Thornton, “The Development of an African Catholic Church in the Kingdom of Kongo, 1491–1750,” Journal of African History, 25 (1984), 147–67; Hilton, Kingdom of Kongo, 32–49, 154–61, 179, 198; MacGaffey, Religion and Society in Central Africa; The BaKongo of Lower Zaire (Chicago, 1986), 191–216, and MacGaffey, “Dialogues of the Deaf,” 249–67. Pacing the cultural intermixture of African and Europe was the simultaneous introduction of European and American plants and animals, which compounded and legitimated many of the cultural changes; Alfred W. Crosby, Ecological Imperialism: The Biological Expansion of Europe, 900–1900 (Cambridge, 1986). 21  The history of one element of this population, the canoemen, is discussed in Peter C. W.  Gutkind, “The Boatmen of Ghana: The Possibilities of a Pre-Colonial African Labor History,” in Michael Hanagan and Charles Stephenson, eds., Confrontation, Class Consciousness, and the Labor Process: Studies in Proletarian Class Formation (Westport, Conn., 1986), 123–66, and Gutkind, “Trade and Labor in Early Precolonial African History: The Canoemen of Southern Ghana,” in Catherine Coquery-Vidrovitch and Paul E. Lovejoy, eds., The Workers of African Trade (Beverly Hills, Calif., 1985), 25–49; Robert Smith, “The

1228

Berlin

Village populations swelled into the thousands. In 1669, about the time the English were ousting the Dutch from the village of New Amsterdam, population 1,500, a visitor to Elmina noted that it contained some 8,000 residents. During most of the eighteenth century, Elmina’s population was between 12,000 and 16,000, larger than Charleston, South Carolina—mainland North America’s greatest slave port at the time of the American Revolution.22 The business of the creole communities was trade, brokering the movement of goods through the Atlantic world. Although island settlements such as Cape Verde, Principé, and São Tomé developed indigenous agricultural and sometimes plantation economies, the comings and goings of African and European merchants dominated life even in the largest of the creole communities, which served as both field headquarters for great European mercantile companies and collection points for trade between the African interior and the Atlantic littoral. Depending on the location, the exchange involved European textiles, metalware, guns, liquor, and beads for African gold, ivory, hides, pepper, beeswax, and dyewoods. The coastal trade or cabotage added fish, produce, livestock, and other perishables to this list, especially as regional specialization developed. Everywhere, slaves were bought and sold, and over time the importance of commerce-in-persons grew.23 As slaving societies, the coastal enclaves were also societies with slaves. African slavery in its various forms—from pawnage to chattel bondage—was practiced in these towns. Both Europeans and Africans held slaves, employed them, used them as collateral, traded them, and sold them to outsiders. At Canoe in West African History,” J. African Hist., 11 (1970), 515–33; and Robin Law, “Trade and Politics behind the Slave Trade: The Lagoon Traffic and the Rise of Lagos,” ibid., 24 (1983), 321–48. See also Daaku, Trade and Politics on the Gold Coast, 103–04, 121–22. For an overview of the coastal towns see Kea, Settlements, Trade, and Polities, esp. chap. 2; Daaku, Trade and Politics on the Gold Coast, chap. 4; and Curtin, Economic Change in Precolonial Africa, 119–20, for the relation between European trading communities and African towns. Since Africans would not rent outsiders more land than needed for a house or a store, food production and other services remained in African hands; Brooks, Landlords and Strangers, 189–90. For bandits see Kea, “ ‘I Am Here to Plunder on the General Road’: Bandits and Banditry in the Pre-Nineteenth Century Gold Coast,” in Donald Crummey, ed., Banditry, Rebellion, and Social Protest in Africa (London, 1986), 109–32. 22  Feinberg, Africans and Europeans in West Africa, 84–85 (for Elmina); Joyce D. Goodfriend, Before the Melting Pot: Society and Culture in Colonial New York City, 1664–1730 (Princeton, 1992), 13 (for New Amsterdam); Peter A. Coclanis, The Shadow of a Dream: Economic Life and Death in the South Carolina Low Country, 1670–1920 (New York, 1989), 115 (for Charleston). 23  Kea, Settlements, Trade, and Polities, esp. chap. 6.

From Creole To African

1229

Elmina, the Dutch West India Company owned some 300 slaves in the late seventeenth century, and individual Europeans and Africans held others. Along with slaves appeared the inevitable trappings of slave societies—overseers to supervise slave labor, slave catchers to retrieve runaways, soldiers to keep order and guard against insurrections, and officials to adjudicate and punish transgressions beyond a master’s reach. Freedmen and freedwomen, who had somehow escaped bondage, also enjoyed a considerable presence. Many former slaves mixed Africa and Europe culturally and sometimes physically.24 Knowledge and experience far more than color set the Atlantic creoles apart from the Africans who brought slaves from the interior and the Europeans who carried them across the Atlantic, on one hand, and the hapless men and women on whose commodification the slave trade rested, on the other. Maintaining a secure place in such a volatile social order was not easy. The creoles’ genius for intercultural negotiation was not simply a set of skills, a tactic for survival, or an attribute that emerged as an “Africanism” in the New World. Rather, it was central to a way of life that transcended particular venues. The names European traders called Atlantic creoles provide a glimpse of the Creole’s cosmopolitan ability to transcend the confines of particular nations and cultures. Abee Coffu Jantie Seniees, a leading African merchant and politico of Cape Coast on the Gold Coast in the late seventeenth century, appears in various European accounts and account books as “Jan Snees,” “Jacque Senece,” “Johan Sinesen,” and “Jantee Snees.” In some measure, the renderings of his name—to view him only from the perspective of European traders—reflect phonic imperialism or, more simply, the variability of transnational spelling. Seniees probably did not know or care how his trading partners registered his name, which he may have employed for commercial reasons in any case. But the diverse renderings reveal something of Abee Coffu Jantie Seniees’s ability to trade with the Danes at Fredriksborg, the Dutch at Elmina, and the English at Cape Coast, as well as with Africans deep in the forested interior.25 24  Feinberg, Africans and Europeans in West Africa, 65, 82–83; Kea, Settlements, Trade, and Polities, 197–202, 289–90. On the Cape Verde Islands, free blacks obtained the right to hold public office in 1546. Racial distinctions did not appear until the emergence of a plantation society in the mid-16th century, when the preoccupation with skin color and hair texture emerged along with racially exclusionary policies; Brooks, Landlords and Strangers, 158–59, 186–87; Dierdre Meintel, Race, Culture, and Portuguese Colonialism in Cabo Verde (Syracuse, N.Y., 1984), 96–103. 25  Kea, Settlements, Trade, and Polities, 233–35, 315–16, 319–20. Daaku notes that “difficulties arise in establishing the exact nationalities” of Gold Coast traders, as European “writers tended to ‘Europeanize’ the names of some of the Africans with whom they traded and

1230

Berlin

The special needs of European traders placed Atlantic creoles in a powerful bargaining position, which they learned to employ to their own advantage. The most successful became principals and traded independently. They played one merchant against another, one captain against another, and one mercantile bureaucrat against another, often abandoning them for yet a better deal with some interloper, all in the hope of securing a rich prosperity for themselves and their families. Success evoked a sense of confidence that observers described as impertinence, insolence, and arrogance, and it was not limited to the fabulously wealthy like Jantie Seniees or the near sovereign John Claessen (the near-ruler of Fetu), who rejected a kingship to remain at trade, or the merchant princes John Kabes (trader, entrepreneur, and dominant politico in Komenda) and John Konny (commanding ruler in Pokoso).26 Canoemen, for example, became infamous among European governors and sea captains for their independence. They refused to work in heavy surf, demanded higher wages and additional rations, quit upon insult or abuse, and abandoned work altogether when enslavement threatened. Attempts to control them through regulations issued from Europe or from local corporate headquarters failed utterly. “These canoemen, despicable thieves,” sputtered one Englishman in 1711, “think that they are more than just labour.”27 Like other people in the middle, Atlantic creoles profited from their strategic position. Competition between and among the Africans and European traders bolstered their stock, increased their political leverage, and enabled them to elevate their social standing while fostering solidarity. Creoles’ ability to find a place for themselves in the interstices of African and European trade grew rapidly during periods of intense competition among the Portuguese, Dutch, Danes, Swedes, French, and English and an equally diverse set of African nationals. At the same time and by the same token, the Atlantic creoles’ liminality, particularly their lack of identity with any one group, posed numerous dangers. While their middling position made them valuable to African and European traders, it also made them vulnerable: they could be ostracized, scapegoated, and on occasion enslaved. Maintaining their independence amid the shifting those in their service, while some of the Africans fancifully assumed European names,” in Trade and Politics on the Gold Coast, 96. 26  Daaku, Trade and Politics on the Gold Coast, chaps. 5–6; David Henige, “John Kabes of Komenda: An Early African Entrepreneur and State Builder,” J. African Hist., 18 (1977), 1–19. 27  Gutkind, “Boatmen of Ghana,” 131–39, quotation on 137, and Gutkind, “Trade and Labor in Early Precolonial African History,” 40–41 (for canoemen who pawned themselves and later became successful traders).

From Creole To African

1231

alliances between and among Europeans and Africans was always difficult. Inevitably, some failed. Debt, crime, immortality, or official disfavor could mean enslavement—if not for great men like Jantie Seniees, Claessen, Kabes, or Konny—at least for those on the fringes of the creole community.28 Placed in captivity, Atlantic creoles might be exiled anywhere around the Atlantic—to the interior of Africa, the islands along the coast, the European metropoles, or the plantations of the New World. In the seventeenth century and the early part of the eighteenth, most slaves exported from Africa went to the sugar plantations of Brazil and the Antilles. Enslaved Atlantic creoles might be shipped to Pernambuco, Barbados, or Martinique. Transporting them to the expanding centers of New World staple production posed dangers, however, which American planters well understood. The characteristics that distinguished Atlantic creoles—their linguistic dexterity, cultural plasticity, and social agility—were precisely those qualities that the great planters of the New World disdained and feared. For their labor force they desired youth and strength, not experience and sagacity. Indeed, too much knowledge might be subversive to the good order of the plantation. Simply put, men and women who understood the operations of the Atlantic system were too dangerous to be trusted in the human tinderboxes created by the sugar revolution. Thus rejected by the most prosperous New World regimes, Atlantic creoles were frequently exiled to marginal slave societies where would-be slaveowners, unable to compete with the great plantation magnates, snapped up those whom the grandees had disparaged as “refuse” for reasons of age, illness, criminality, or recalcitrance. In the seventeenth century, few New World slave societies were more marginal than those of mainland North America.29 Liminal peoples were drawn or propelled to marginal societies. 28  For enslavement of canoemen for violation of Portuguese regulations see Gutkind, “Trade and Labor in Early Precolonial African History,” 27–28, 36. Okoyaw, a canoeman who pawned himself to the Royal African Company in 1704 to redeem a debt, agreed in return “to attend Dayly the Company’s Work”; cited in Kea, Settlements, Trade, and Polities, 243. Because there was no established system of commercial law, creditors might seize the slaves or even the fellow townsmen of their debtors to satisfy an obligation; Curtin, Economic Change in Precolonial Africa, 302–08. 29  The northern North American colonies often received “refuse” slaves. For complaints and appreciations see Goodfriend, “Burghers and Blacks: The Evolution of a Slave Society at New Amsterdam,” New York History, 59 (1978), 139; Sainsbury et al., eds., Calendar of State Papers, Colonial Series, 1708–1709, 24:110; Lorenzo J. Greene, The Negro in Colonial New England, 1620–1776 (New York, 1942), 35; Jeremias van Rensselaer to Jan Baptist van Rensselaer, ca. 1659, in A. J. F. Van Laer, ed., Correspondence of Jeremias Van Rensselaer,

1232

Berlin

During the seventeenth century and into the eighteenth, the Dutch served as the most important conduit for transporting Atlantic creoles to mainland North America. Through their control of the sea, they dominated the commerce of the Atlantic periphery. Stretching mercantile theory to fit their commercial ambitions, the Dutch traded with all comers, commissioned privateers to raid rival shipping, and dealt openly with pirates. The Dutch West India Company, whose 1621 charter authorized it to trade in both the Americas and west Africa, cast its eye on the lucrative African trade in gold, ivory, copper, and slaves even as it began to barter for furs and pelts in the North Atlantic and for gold and sugar in the South Atlantic. In 1630, the Dutch captured Portuguese capitanias in northeastern Brazil, including Pernambuco, the site of the New World’s first sugar boom. About the same time, the West India Company established bases in Curaçao and St. Eustatius. To supply their new empire, the Dutch turned to Africa, supplementing their outposts at Mouri on the Gold Coast and Gorée in Senegambia by seizing the Portuguese enclaves of Elmina and Axim in 1637, Luanda and Principé in 1641, and São Tomé in 1647. They then swept the Angolan coast, establishing trading factories at Cabinda, Loango, and Mpinda.30 Although ousted from the Gold Coast, the Portuguese never abandoned their foothold in central Africa, and they and their Brazilian successors regrouped and counterattacked. In 1648, the Portuguese recaptured Luanda and forced the Dutch to evacuate Angola. They expelled the Dutch from Pernambuco in 1645 and completed the reconquest of Brazil in 1654. Still, the short period of Dutch dominance—roughly, 1620 to 1670—had a powerful impact on the Atlantic world. During those years, the Dutch took control of Portuguese enclaves in Africa, introduced their commercial agents, and pressed their case for Dutch culture and Calvinist religion on the ruling Kongolese Catholics and other remnants of Portuguese imperialism. Although 1651–1674 (Albany, 1932), 167–68, 175; William D. Pierson, Black Yankees: The Development of an Afro-American Subculture in Eighteenth-Century New England (Amherst, Mass., 1988), 4–5; Edgar J. McManus, Black Bondage in the North (Syracuse, N. Y., 1973), 18–25, and McManus, A History of Slavery in New York (Syracuse, N. Y., 1966), 23–39; James G. Lydon, “New York and the Slave Trade, 1700 to 1774,” WMQ, 3d Ser., 35 (1978), 275–79, 281–90; Darold D. Wax, “Negro Imports into Pennsylvania, 1720–1766,” Pennsylvania History, 32 (1965), 254–87, and Wax, “Preferences for Slaves in Colonial America,” J. Negro Hist., 58 (1973), 374–76, 379–87; and Sharon V. Salinger, “To Serve Well and Faithfully”: Labour and Indentured Servitude in Pennsylvania, 1682–1800 (Cambridge, 1987), 75–78. 30  Boxer, The Dutch in Brazil, 1624–1654 (Oxford, 1957); Johannes Menne Postma, The Dutch in the Atlantic Slave Trade (Cambridge, 1990), chaps. 2–3, 8; Cornells C. Goslinga, The Dutch in the Caribbean and on the Wild Coast, 1580–1680 (Gainesville, Fla., 1971).

From Creole To African

1233

unsuccessful for the most part, the Dutch established ties with the Atlantic creoles and preserved these linkages even after the Portuguese reconquest, keeping alive their connections along the African coast and maintaining their position as the most active agents in slavery’s transatlantic expansion during the seventeenth century.31 The Dutch transported thousands of slaves from Africa to the New World, trading with all parties, sometimes directly, sometimes indirectly through their base in Curaçao. Most of these slaves came from the interior of Angola, but among them were Atlantic creoles whose connections to the Portuguese offended the Dutch. Following the Portuguese restoration, those with ties to the Dutch may have found themselves in similar difficulties. During the Dutch invasions, the subsequent wars, and then civil wars in which the Portuguese and the Dutch fought each other directly and through surrogates, many creoles were clapped into slavery. Others were seized in the Caribbean by Dutch men-of-war, privateers sailing under Dutch letters of marque, and freebooting pirates.32 While such slaves might be sent anywhere in the Dutch empire between New Netherland and Pernambuco, West India Company officers in New Amsterdam, who at first complained about “refuse” slaves, in time made known their preference for such creoles—deeming “Negroes who had been 12 or 13 years in the West Indies” to be “a better sort of Negroes.”33 A perusal of the names scattered through archival remains of New Netherland reveals something of the nature of this transatlantic transfer: Paulo d’Angola and Anthony Portuguese, Pedro Negretto and Francisco Negro, Simon Congo 31  Boxer, Four Centuries of Portuguese Expansion, 48–51, and Boxer, Dutch in Brazil; P. C. Emmer, “The Dutch and the Making of the Second Atlantic System,” in Solow, ed., Slavery and the Rise of the Atlantic System, 75–96, esp. 83–84; Thornton, Africa and Africans in the Making of the Atlantic World, 64–65, 69–77. Albert van Danzig, ed. and trans., The Dutch and the Guinea Coast, 1674–1742: A Collection of Documents from the General State Archives at the Hague (Accra, 1978), provides insight into the operation of the Dutch West India Company and the role of the Dutch on the Gold and Slave coasts. 32  Thornton, The Kingdom of Kongo: Civil War and Transition, 1641–1718 (Madison, Wis., 1983), esp. 72–74, chaps. 6–7 passim; Hilton, Kingdom of Kongo, chaps. 6–7; Ernst Van Den Boogaart and Pieter C. Emmer, “The Dutch Participation in the Atlantic Slave Trade, 1596– 1650,” in Henry A. Gemery and Jan S. Hogendorn, eds., The Uncommon Market: Essays in the Economic History of the Atlantic Slave Trade (New York, 1979), 353–71. The best survey of the Dutch trade is Postma, Dutch in the Atlantic Slave Trade. 33  Goslinga, Dutch in the Caribbean and on the Wild Coast; Van Laer, ed., Correspondence of Jeremias Van Rensselaer, 167–68, 175, quotation on 167; Elizabeth Donnan, ed., Documents Illustrative of the History of the Slave Trade to America, 4 vols. (Washington, D.C., 1930– 1935), 3:421; Goodfriend, “Burghers and Blacks,” 139.

1234

Berlin

and Jan Guinea, Van St. Thomas and Francisco Cartagena, Claes de Neger and Assento Angola, and—perhaps most telling—Carla Criole, Jan Creoli, and Christoffel Crioell.34 These names trace the tumultuous experience that propelled their owners across the Atlantic and into slavery in the New World. They suggest that whatever tragedy befell them, Atlantic creoles did not arrive in the New World as deracinated chattel stripped of their past and without resources to meet the future. Unlike those who followed them into slavery in succeeding generations, transplanted creoles were not designated by diminutives, tagged with names more akin to barnyard animals, or given the name of an ancient notable or a classical deity. Instead, their names provided concrete evidence that they carried a good deal more than their dignity to the Americas. 34  Names are drawn from E. B. O’Callaghan, comp., The Documentary History of the State of New-York, 4 vols. (Albany, 1849–1851); O’Callaghan, ed., Documents Relative to the Colonial History of the State of New-York, 15 vols. (Albany, 1853–1887) (hereafter N. Y. Col. Docs.); O’Callaghan, comp., Laws and Ordinances of New Netherland, 1638–1674 (Albany, 1868); O’Callaghan, ed., Calendar of Historical Manuscripts in the Office of the Secretary of State, 1630–1664 (Albany, 1865); Berthold Fernow, ed., The Records of New Amsterdam from 1653 to 1674, 7 vols. (New York, 1897); Fernow, ed., Minutes of the Orphanmasters Court of New Amsterdam, 1655–1663, 2 vols. (New York, 1907); Kenneth Scott and Kenn StrykerRodda, comps., New York Historical Manuscripts: Dutch, vols. 1–4 (Baltimore, 1974–); Charles T. Gehring, ed., New York Historical Manuscripts: Dutch Land Papers (Baltimore, 1980); New York Genealogical and Biographical Society, Collections, Marriages from 163c to 1801 in the Reformed Dutch Church of New York (New York, 1890); I. N. Phelps Stokes, Iconography of Manhattan Island, 1498–1909, 6 vols. (New York, 1914–1928). A few names suggest the subtle transformation as the Atlantic creoles crossed the ocean and assumed a new identity that was unfamiliar to its hosts. For example, Anthony Jansen of Salee or Van Vaes, a man of tawny complexion—“mulatto,” per below—who claimed Moroccan birth, became “Anthony the Turk,” perhaps because Turks were considered fierce—as Anthony’s litigious history indicates he surely was—but, more important, because he was alien in status and brown in pigment; Leo Hershkowitz, “The Troublesome Turk: An Illustration of Judicial Process in New Amsterdam,” N. Y. Hist., 46 (1965), 299–310. But if names of new arrivals in New Netherland reflect their lived experience rather than an owner’s designation, they also have nothing of the ring of Africa: no Quaws, Phibbis, or any of the day names that Africans later carried. Such names would become familiar to northern slaveholders when the slave trade reached into the interior of Africa. In Portugal, the names slaves bore do not seem different from those of native Portuguese; Saunders, Black Slaves and Freedmen in Portugal, 89–90. The practice of attaching a national modifier to a given name was employed for others besides Africans. See Edmund S. Morgan, American Slavery, American Freedom: The Ordeal of Colonial Virginia (New York, 1975), 153–54.

From Creole To African

1235

To such men and women, New Amsterdam was not radically different from Elmina or Luanda, save for its smaller size and colder climate. A fortified port controlled by the Dutch West India Company, its population was a farrago of petty traders, artisans, merchants, soldiers, and corporate functionaries, all scrambling for status in a frontier milieu that demanded intercultural exchange. On the tip of Manhattan Island, Atlantic creoles rubbed elbows with sailors of various nationalities, Native Americans with diverse tribal allegiances, and pirates and privateers who professed neither nationality nor allegiance. In the absence of a staple crop, their work—building fortifications, hunting and trapping, tending fields and domestic animals, and transporting merchandise of all sorts—did not set them apart from workers of European descent, who often labored alongside them. Such encounters made a working knowledge of the creole tongue as valuable on the North American coast as in Africa. Whereas a later generation of transplanted Africans would be linguistically isolated and de-skilled by the process of enslavement, Atlantic creoles found themselves very much at home in the new environment. Rather than losing their skills, they discovered that the value of their gift for intercultural negotiation appreciated. The transatlantic journey did not break creole communities; it only transported them to other sites.35 Along the edges of the North American continent, creoles found slaves’ cultural and social marginality an asset. Slaveholders learned that slaves’ ability to negotiate with the diverse populace of seventeenth-century North America was as valuable as their labor, perhaps more so. While their owners employed creoles’ skills on their own behalf, creoles did the same for themselves, trading their knowledge for a place in the still undefined social order. In 1665, when Jan Angola, accused of stealing wood in New Amsterdam, could not address the court in Dutch, he was ordered to return the following day with “Domingo the Negro as interpreter,” an act familiar to Atlantic creoles in Elmina, Lisbon, San Salvador, or Cap Françis.36

35  Nothing evidenced the creoles’ easy integration into the mainland society better than the number who survived into old age. There are no systematic demographic studies of people of African descent during the first years of settlement, and perhaps because the numbers are so small, there can be none. Nevertheless, “old” or “aged” slaves are encountered again and again, sometimes in descriptions of fugitives, sometimes in the deeds that manumit—i. e., discard—superannuated slaves. Before the end of the 17th century, numbers of black people lived long enough to see their grandchildren. 36  Fernow, ed., Records of New Amsterdam, 5:337, cited in Goodfriend, Before the Melting Pot, 252 n. 25.

1236

Berlin

To be sure, slavery bore heavily on Atlantic creoles in the New World. As in Africa and Europe, it was a system of exploitation, subservience, and debasement that rested on force. Yet Atlantic creoles were familiar with servitude in forms ranging from unbridled exploitation to corporate familialism. They had known free people to be enslaved, and they had known slaves to be liberated; the boundary between slavery and freedom on the African coast was permeable. Servitude generally did not prevent men and women from marrying, acquiring property (slaves included), enjoying a modest prosperity, and eventually being incorporated into the host society; creoles transported across the Atlantic had no reason to suspect they could not do the same in the New World.37 If the stigma of servitude, physical labor, uncertain lineage, and alien religion stamped them as outsiders, there were many others—men and women of unblemished European pedigree prominent among them—who shared those taints. That black people could and occasionally did hold slaves and servants and employ white people suggested that race—like lineage and religion—was just one of many markers in the social order. If slavery meant abuse and degradation, the experience of Atlantic creoles provided strategies for limiting such maltreatment—contrary to notions that they were libidinous heathens without family, economy, or society—and even for winning to freedom. Freedom meant not only greater independence but also identification with the larger group. Although the routes to social betterment were many, they generally involved reattachment to a community through the agency of an influential patron or, better yet, an established institution that could broker a slave’s incorporation into the larger society.38 Along the coast of Africa, Atlantic creoles often identified with the appendages of European or African power—be they international mercantile corporations or local chieftains—in hopes of relieving the stigma of otherness—be it 37  Suzanne Miers and Igor Kopytoff, eds., Slavery in Africa: Historical and Anthropological Perspectives (Madison, Wis., 1977); Paul E. Lovejoy, Transformations in Slavery: A History of Slavery in Africa (Cambridge, 1983); Patrick Manning, Slavery and African Life: Occidental, Oriental, and African Slave Trades (Cambridge, 1990); Thornton, Africa and Africans in the Making of the Atlantic World, chap 3; Claude Meillassoux, The Anthropology of Slavery: The Womb of Iron and Gold (Chicago, 1991); Martin A. Klein, “Introduction: Modern European Expansion and Traditional Servitude in Africa and Asia,” in Klein, ed., Breaking the Chains: Slavery, Bondage, and Emancipation in Modern Africa and Asia (Madison, Wis., 1993), 3–26; Toyin Falola and Lovejoy, “Pawnship in Historical Perspective,” in Falola and Lovejoy, eds., Pawnship in Africa: Debt Bondage in Historical Perspective (Boulder, Colo., 1994), 1–26. A dated but still useful critical review of the subject is Frederick Cooper, “The Problem of Slavery in African Studies,” J. African Hist., 20 (1979), 103–25. 38  Miers and Kopytoff, eds., Slavery in Africa, chap. 1, esp. 17.

From Creole To African

1237

enslavement, bastard birth, paganism, or race. They employed this strategy repeatedly in mainland North America, as they tried to hurdle the boundaries of social and cultural difference and establish a place for themselves. By linking themselves to the most important edifices of the nascent European-American societies, Atlantic creoles struggled to become part of a social order where exclusion or otherness—not subordination—posed the greatest dangers. To be inferior within the sharply stratified world of the seventeenth-century Atlantic was understandable by its very ubiquity; to be excluded posed unparalleled dangers. The black men and women who entered New Netherland between 1626 and the English conquest in 1664 exemplified the ability of people of African descent to integrate themselves into mainland society during the first century of settlement, despite their status as slaves and the contempt of the colony’s rulers. Far more than any other mainland colony during the first half of the seventeenth century, New Netherland rested on slave labor. The prosperity of the Dutch metropole and the opportunities presented to ambitious men and women in the far-flung Dutch empire denied New Netherland its share of free Dutch immigrants and limited its access to indentured servants. To populate the colony, the West India Company scoured the Atlantic basin for settlers, recruiting German Lutherans, French Huguenots, and Sephardic Jews. These newcomers did little to meet the colony’s need for men and women to work the land, because, as a company officer reported, “agricultural laborers who are conveyed thither at great expense … sooner or later apply themselves to trade, and neglect agriculture altogether.” Dutch officials concluded that slave labor was an absolute necessity for New Netherland. Although competition for slaves with Dutch outposts in Brazil (whose sugar economy was already drawing slaves from the African interior) placed New Netherland at a disadvantage, authorities in the North American colony imported all the slaves they could, so that in 1640 about 100 blacks lived in New Amsterdam, composing roughly 30 percent of the port’s population and a larger portion of the labor force. Their proportion diminished over the course of the seventeenth century but remained substantial. At the time of the English conquest, some 300 slaves composed a fifth of the population of New Amsterdam, giving New Netherland the largest urban slave population on mainland North America.39

39  Goodfriend, Before the Melting Pot, 10, chap. 6; Van Den Boogaart, “The Servant Migration to New Netherland, 1624–1664,” in Emmer, ed., Colonialism and Migration: Indentured Labour before and after Slavery (Dordrecht, 1986), 58; O’Callaghan, ed., N. Y. Col. Docs., 1:154.

1238

Berlin

The diverse needs of the Dutch mercantile economy strengthened the hand of Atlantic creoles in New Netherland during the initial period of settlement. Caring only for short-term profits, the company, the largest slaveholder in the colony, allowed its slaves to live independently and work on their own in return for a stipulated amount of labor and an annual tribute. Company slaves thus enjoyed a large measure of independence, which they used to master the Dutch language, trade freely, accumulate property, identify with Dutch Reformed Christianity, and—most important—establish families. During the first generation, some twenty-five couples took their vows in the Dutch Reformed Church in New Amsterdam. When children arrived, their parents baptized them as well. Participation in the religious life of New Netherland provides but one indicator of how quickly Atlantic creoles mastered the intricacies of life in mainland North America. In 1635, less than ten years after the arrival of the first black people, black New Netherlanders understood enough about the organization of the colony and the operation of the company to travel to the company’s headquarters in Holland and petition for wages.40 Many slaves gained their freedom. This was not easy in New Netherland, although there was no legal proscription on manumission. Indeed, gaining freedom was nearly impossible for slaves owned privately and difficult even for those owned by the company. The company valued its slaves and was willing 40  Goodfriend, Before the Melting Pot, chap. 6; Goodfriend, “Burghers and Blacks,” 125–44; Goodfriend, “Black Families in New Netherland,” Journal of the Afro-American Historical and Genealogical Society, 5 (1984), 94–107; Morton Wagman, “Corporate Slavery in New Netherland,” J. Negro Hist., 65 (1980), 34–42; McManus, Slavery in New York, 2–22; Michael Kammen, Colonial New York: A History (New York, 1975), 58–60; Van Den Boogaart, “Servant Migration to New Netherland,” 56–59, 65–71; Vivienne L. Kruger, “Born to Run: The Slave Family in Early New York, 1626 to 1827” (Ph. D. diss., Columbia University, 1985), chap. 2, esp. 46–48, chap. 6, esp. 270–77; Oliver A. Rink, Holland on the Hudson: An Economic and Social History of Dutch New York (Ithaca, 1986), 161 n. 33. Between 1639 and 1652, marriages recorded in the New Amsterdam Church represented 28% of the marriages recorded in that period—also note one interracial marriage. For baptisms see “Reformed Dutch Church, New York, Baptisms, 1639–1800,” New York Genealogical and Biographical Society, Collections, 2 vols. (New York, 1901), 1:10–27, 2:10–38; for the 1635 petition see Stokes, Iconography of Manhattan Island, 4:82, and No. 14, Notulen W1635, 1626 (19–11–1635), inv. 1.05.01. 01. (Oude), Algemeen Rijksarchief, The Hague. A petition by “five blacks from New Netherland who had come here [Amsterdam]” was referred back to officials in New Netherland. Marcel van der Linden of the International Institute of Social History in Amsterdam kindly located and translated this notation in the records of the Dutch West India Company.

From Creole To African

1239

to liberate only the elderly, whom it viewed as a liability. Even when manumitting such slaves, the company exacted an annual tribute from adults and retained ownership of their children. The latter practice elicited protests from both blacks and whites in New Amsterdam. The enslavement of black children made “half-freedom,” as New Netherland authorities denominated the West India Company’s former slaves who were unable to pass their new status to their children, appear no freedom at all.41 Manumission in New Netherland was calculated to benefit slave owners, not slaves. Its purposes were to spur slaves to greater exertion and to relieve owners of the cost of supporting elderly slaves. Yet, however compromised the attainment of freedom, slaves did what was necessary to secure it. They accepted the company’s terms and agreed to pay its corporate tribute. But they bridled at the fact that their children’s status would not follow their own. Halffree blacks pressed the West India Company to make their status hereditary. Hearing rumors that baptism would assure freedom to their children, they pressed their claims to church membership. A Dutch prelate complained of the “worldly and perverse aims” of black people who “wanted nothing else than to deliver their children from bodily slavery, without striving for piety and Christian virtues.”42 Although conversion never guaranteed freedom in New Netherland, many half-free blacks secured their goal. By 1664, at the time of the English conquest, about one black person in five had achieved freedom in New

41  Petition for freedom, in O’Callaghan, ed., Calendar of Historical Manuscripts, 269. White residents of New Amsterdam protested the enslavement of the children of half-free slaves, holding that no one born of a free person should be a slave. The Dutch West India Company rejected the claim; O’Callaghan, ed., N. Y. Col. Docs., 1:302, 343; O’Callaghan, ed., Laws and Ordinances of New Netherland, 4:36–37. For the Dutch West India Company’s “setting them free and at liberty, on the same footing as other free people here in New Netherland,” although children remained property of the company, see Van Den Boogaart, “Servant Migration to New Netherland,” 69–70. 42  For black men paying tribute to purchase their families see O’Callaghan, ed., Calendar of Historical Manuscripts, 45, 87, 105; O’Callaghan, ed., N. Y. Col. Docs., 1:343; Goodfriend, “Burghers and Blacks,” 125–44, and “Black Families in New Netherlands,” 94–107; McManus, Slavery in New York, 2–22; Wagman, “Corporate Slavery in New Netherland,” 38–39; quotation in Gerald Francis Dejong, “The Dutch Reformed Church and Negro Slavery in Colonial America,” Church History, 40 (1971), 430; Kruger, “Born to Run,” chap. 1, esp. 90–92; Henry B. Hoff, “Frans Abramse Van Salee and His Descendants: A Colonial Black Family in New York and New Jersey,” New York Genealogical and Biographical Register, 121 (1990), 65–71, 157–61.

1240

Berlin

Amsterdam, a proportion never equalled throughout the history of slavery in the American South.43 Some free people of African descent prospered. Building on small gifts of land that the West India Company provided as freedom dues, a few entered the landholding class in New Netherland. A small group of former slaves established a community on the outskirts of the Dutch settlement on Manhattan, farmed independently, and sold their produce in the public market. Others purchased farmsteads or were granted land as part of the Dutch effort to populate the city’s hinterland. In 1659, the town of Southampton granted “Peeter the Neigro” three acres. Somewhat later John Neiger, who had “set himself up a house in the street” of Easthampton, was given “for his own use a little quantity of land above his house for him to make a yard or garden.” On occasion, free blacks employed whites.44 By the middle of the seventeenth century, black people participated in almost every aspect of life in New Netherland. They sued and were sued in Dutch courts, married and baptized their children in the Dutch Reformed Church, and fought alongside Dutch militiamen against the colony’s enemies. Black men and women—slave as well as free—traded on their own and accumulated property. Black people also began to develop a variety of institutions that reflected their unique experience and served their special needs. Black men and women stood as godparents to each others’ children, suggesting close family ties, and rarely called on white people—owners or not—to serve in this capacity. At times, established black families legally adopted orphaned black children, further knitting the black community together in a web of fictive kinship.45 The patterns of residence, marriage, church membership, and 43  Goodfriend estimates that 75 of New Amsterdam’s 375 blacks were free in 1664, in Before the Melting Pot, 61. 44  Kruger, “Born to Run,” 50–55, 591–606, tells of the creation of a small class of black landowners as a result of gifts from the Dutch West India Company and direct purchase by the blacks themselves (quotation on 592); Goodfriend, Before the Melting Pot, 115–16, 253 n. 36; Peter R. Christoph, “The Freedmen of New Amsterdam,” J. Afro-Amer. Hist. Gen. Soc., 5 (1984), 116–17. See also Stokes, Iconography of Manhattan Island, 2:302, 4:70–78, 100, 104–06, 120–48, 265–66; Gehring, ed., New York Historical Manuscripts; and Van Den Boogaart, “Servant Migration to New Netherland,” 56–59, 65–71. For the employment of a white housekeeper by a free black artisan see ibid., 69; Fernow, ed., Minutes of the Orphanmasters Court, 2: 46; and Roi Ottley and William J. Weatherby, eds., The Negro in New York: An Informal Social History (New York, 1967), 12. 45  O’Callaghan, ed., Calendar of Historical Manuscripts, 87, 105, 269 (for manumission, dubbed “half slaves”), 222 (adoption), 269 (land grants). See also Goodfriend, Before the Melting Pot, chap. 6, and Fernow, ed., Records of New Amsterdam from 1653 to 1674, 3:42,

From Creole To African

1241

godparentage speak not only to the material success of Atlantic creoles but also to their ability to create a community among themselves. To be sure, the former slaves’ prosperity was precarious at best. As the Dutch transformed their settlement from a string of trading posts to a colony committed to agricultural production, the quality of freedpeople’s freedom deteriorated. The Dutch began to import slaves directly from Africa (especially after the Portuguese retook Brazil), and the new arrivals—sold mostly to individual planters rather than to the company—had little chance of securing the advantages earlier enjoyed by the company’s slaves.46 The freedpeople’s social standing eroded more rapidly following the English conquest in 1664, demonstrating the fragility of their freedom in a social order undergirded by racial hostility. Nonetheless, black people continued to enjoy the benefits of the earlier age. They maintained a secure family life, acquired property, and participated as communicants in the Dutch Reformed Church, where they baptized their children in the presence of godparents of their own choosing. When threatened, they took their complaints to court, exhibiting a fine understanding of their legal rights and a steely determination to defend them. Although the proportion of the black population enjoying freedom shrank steadily under English rule, the small free black settlement held its own. Traveling through an area of modest farms on the outskirts of New York City in 1679, a Dutch visitor observed that “upon both sides of this way were many habitations of negroes, mulattoes and whites. These negroes were 5, 172, 337–40, 7, 11 (for actions in court); Goodfriend, “Burghers and Blacks,” 125–44, and “Black Families in New Netherlands,” 94–107; Van Den Boogaart, “Servant Migration to New Netherland,” 56–59, 65–71; McManus, Slavery in New York, 2–22; DeJong, “Dutch Reformed Church and Negro Slavery in Colonial America,” 430; Kruger, “Born to Run,” 46– 48; 270–78; Hoff, “Frans Abramse Van Salee and His Descendants”; Kammen, New York, 58–60. For blacks using Dutch courts early on see Rink, Holland on the Hudson, 160–61— e. g., in 1638, Anthony Portuguese sued Anthony Jansen for damages done his hog; soon after, one Pedro Negretto claimed back wages. For adoption of a black child by a free black family see Scott and Stryker-Rodda, eds., The Register of Salmon Lachaire, Notary Public of New Amsterdam, 1661–1662 (Baltimore, 1978), 22–23; O’Callaghan, ed., Calendar of Historical Manuscripts, 222, 256; and Kruger, “Born to Run,” 44–51. 46  Until New Netherland developed an agricultural base, slavery did not seem to take hold, and settlers admitted in 1649 that slaves imported at great cost “just dripped through the fingers” and “were sold for pork and peas”; O’Callaghan, ed., N. Y. Col. Docs., 1:302. For the change that took place during the 1650s and the beginning of direct African importation in 1655 see O’Callaghan, ed., Calendar of Historical Manuscripts, 268, 289, 293, 307, 331. New York sharply limited manumission in 1712. Few slaves were freed before then. One careful enumeration counted 8 manumissions between 1669 and 1712; Kruger, “Born to Run,” 593.

1242

Berlin

formerly the property of the (West India) company, but, in consequence of the frequent changes and conquests of the country, they have obtained their freedom and settled themselves down where they thought proper, and thus on this road, where they have ground enough to live on with their families.”47 Dutch vessels were not the only ones to transport Atlantic creoles from Africa to North America. The French, who began trading on the Windward Coast of Africa soon after the arrival of the Portuguese, did much the same. Just as a creole population grew up around the Portuguese and later Dutch factories at Elmina, Luanda, and São Tomé, so one developed around the French posts on the Senegal River. The Compagnie du Senegal, the Compagnie des Indes Occidentales, and their successor, the Compagnie des Indes—whose charter, like that of the Dutch West India Company, authorized it to trade in both Africa and the Americas—maintained headquarters at St. Louis with subsidiary outposts at Galam and Fort d’Arguin.48 As at Elmina and Luanda, shifting alliances between Africans and Europeans in St. Louis, Galam, and Fort d’Arguin also ensnared Atlantic creoles, who found themselves suddenly enslaved and thrust across the Atlantic. One such man was Samba, a Bambara,49 who during the 1720s worked for the French as an interpreter—maître de langue—at Galam, up the Senegal River from St. Louis. “Samba Bambara”—as he appears in the records—traveled freely along the river between St. Louis, Galam, and Fort d’Arguin. By 1722, he received 47  James B. Bartlett and J. Franklin Jameson, eds., Journal of Jasper Danckaerts, 1679–1680 (New York, 1913), 65. See also Goodfriend, Before the Melting Pot, 115–16 (land). After the English conquest, black people continued to present their children for baptism, although they changed to the Anglican church; ibid., 131. 48  Curtin, Economic Change in Precolonial Africa, 104–05, 121–27; Jean Mettas, Répertoire des Expéditions Négrières Françaises au XVIIIe Siècle, 2 vols. (Paris, 1984); Marcel Giraud, A History of French Louisiana: The Reign of Louis XIV, 1698–1715, trans. Joseph C. Lambert (Baton Rouge, 1974); Hall, Africans in Colonial Louisiana, chaps. 2–4. For the French slave trade see Robert Louis Stein, The French Slave Trade in the Eighteenth Century: An Old Regime Business (Madison, Wis., 1979). 49  The Bambaras had complex relations with the French. Although many Bambaras—usually captives of the tribe whom the French also deemed Bambaras (although they often were not)—became entrapped in the international slave trade and were sold to the New World, others worked for the French as domestics, boatmen, clerks, and interpreters in the coastal forts and slave factories. Their proud military tradition—honed in a long history of warfare against Mandingas and other Islamic peoples—made them ideal soldiers as well as slave catchers. Along the coast of Africa, “Bambara” became a generic word for soldier; Hall, Africans in Colonial Louisiana, 42, and Curtin, Economic Change in Precolonial Africa, 115, 143, 149, 178–81, 191–92; see the review of Hall in Africa, 64 (1994), 168–71.

From Creole To African

1243

permission from the Compagnie des Indes for his family to reside in St. Louis. When his wife dishonored him, Samba Bambara called on his corporate employer to exile her from St. Louis and thereby bring order to his domestic life. But despite his reliance on the company, Samba Bambara allegedly joined with African captives in a revolt at Fort d’Arguin, and, when the revolt was quelled, he was enslaved and deported. Significantly, he was not sold to the emerging plantation colony of Saint Domingue, where the sugar revolution stoked a nearly insatiable appetite for slaves. Instead, French officials at St. Louis exiled Samba Bambara to Louisiana, a marginal military outpost far outside the major transatlantic sea lanes and with no staple agricultural economy.50 New Orleans on the Mississippi River shared much with St. Louis on the Senegal in the 1720s. As the headquarters of the Compagnie des Indes in mainland North America, the town housed the familiar collection of corporate functionaries, traders, and craftsmen, along with growing numbers of French engagés and African slaves. New Orleans was frequented by Indians, whose canoes supplied it much as African canoemen supplied St. Louis. Its taverns and back alley retreats were meeting places for sailors of various nationalities, Canadian coureurs de bois, and soldiers—the latter no more pleased to be stationed on the North American frontier than their counterparts welcomed assignment to an African factory.51 Indeed, soldiers’ status in this rough frontier community differed little from that on the coast of Africa.

50  The evidence of Samba’s participation in the Fort D’Arguin insurrection is insubstantial and contradictory, but he got himself into enough trouble to be enslaved and deported; Hall, Africans in Colonial Louisiana, 109–10; Le Page du Pratz, Histoire de la Louisiane, 3 vols. (Paris, 1758), 3:305–17; Daniel H. Usner, Jr., “From African Captivity to American Slavery: The Introduction of Black Laborers to Colonial Louisiana,” La. Hist., 20 (1979), 37. On the Afro-French community in St. Louis and other enclaves on the Senegal see Curtin, Economic Change in Precolonial Africa, 113–21. 51  The first census of the French settlement of the lower Mississippi Valley comes from Biloxi in 1699. It lists 5 naval officers, 5 petty officers, 4 sailors, 19 Canadians, 10 laborers, 6 cabin boys, and 20 soldiers; Hall, Africans in Colonial Louisiana, 3, and esp. chap. 5. Usner makes the point in comparing the use of black sailors on the Mississippi and the Senegal, in “From African Captivity to American Slavery,” 25–47, esp. 36, and more generally in Indians, Settlers, and Slaves in a Frontier Exchange Economy: The Lower Mississippi Valley before 1783 (Chapel Hill, 1992). See also James T. McGowan, “Planters without Slaves: Origins of a New World Labor System,” Southern Studies, 16 (1977), 5–20; John G. Clark, New Orleans, 1718–1812: An Economic History (Baton Rouge, 1970), chap. 2; and Thomas N. Ingersoll, “Old New Orleans: Race, Class, Sex, and Order in the Early Deep South, 1718– 1819” (Ph. D. diss., University of California at Los Angeles, 1990), chaps. 2–3.

1244

Berlin

In 1720, a French soldier stationed in New Orleans was convicted of theft and sentenced to the lash. A black man wielded the whip. His work was apparently satisfactory, because five years later, Louis Congo, a recently arrived slave then in the service of the Compagnie des Indes, was offered the job. A powerful man, Congo bargained hard before accepting such grisly employment; he demanded freedom for himself and his wife, regular rations, and a plot of land he could cultivate independently. Louisiana’s Superior Council balked at these terms, but the colony’s attorney general urged acceptance, having seen Congo’s “chef d’oeuvre.” Louis Congo gained his freedom and was allowed to live with his wife (although she was not free) on land of his own choosing. His life as Louisiana’s executioner was not easy. He was assaulted several times, and he complained that assassins lurked everywhere. But he enjoyed a modest prosperity, and he learned to write, an accomplishment that distinguished him from most inhabitants of eighteenth-century Louisiana.52 Suggesting something of the symmetry of the Atlantic world, New Orleans, save for the flora and fauna, was no alien terrain to Samba Bambara or Louis Congo. Despite the long transatlantic journey, once in the New World, they recovered much of what they had lost in the Old, although Samba Bambara never escaped slavery. Like the Atlantic creoles who alighted in New Netherland, Samba Bambara employed on the coast of North America skills he had learned on the coast of Africa; Louis Congo’s previous occupation is unknown. Utilizing his knowledge of French, various African languages, and the ubiquitous creole tongue, the rebel regained his position with his old patron, the Compagnie des Indes, this time as an interpreter swearing on the Christian Bible to translate faithfully before Louisiana’s Superior Council. Later, he became an overseer on the largest “concession” in the colony, the company’s massive plantation across the river from New Orleans.53 Like his counterparts in New Amsterdam, Samba Bambara succeeded in a rugged frontier slave society by following the familiar lines of patronage to the doorstep of his corporate employer. Although the constraints of slavery eventually turned him against the company on the Mississippi, just as he had turned against it on the Senegal River, his ability to transfer his knowledge and skills from the Old World to the New, despite the weight of enslavement, suggests that the history of Atlantic creoles in New Amsterdam—their ability to escape slavery, form families, secure property, and claim a degree of independence—was no anomaly.

52  Hall, Africans in Colonial Louisiana, 131–32. 53  Ibid., 106–12; du Pratz, Histoire de la Louisiane, 3:305–17; Usner, “From African Captivity to American Slavery,” 37, 42.

From Creole To African

1245

Atlantic creoles such as Paulo d’Angola in New Netherland and Samba Bambara in New Orleans were not the only products of the meeting of Africans and Europeans on the coast of Africa. By the time Europeans began to colonize mainland North America, communities of creoles of African descent similar to those found on the West African feitorias had established themselves all along the rim of the Atlantic. In Europe—particularly Portugal and Spain— the number of Atlantic creoles swelled, as trade with Africa increased. By the mid-sixteenth century, some 10,000 black people lived in Lisbon, where they composed about 10 percent of the population. Seville had a slave population of 6,000 (including a minority of Moors and Moriscos).54 As the centers of the Iberian slave trade, these cities distributed African slaves throughout Europe.55 With the settlement of the New World, Atlantic creoles sprouted in such places as Cap Françis, Cartagena, Havana, Mexico City, and San Salvador. Intimate with the culture of the Atlantic, they could be found speaking pidgin and creole and engaging in a familiar sort of cultural brokerage. Men drawn from these creole communities accompanied Columbus to the New World; others marched with Balboa, Cortés, De Soto, and Pizarro.56 Some Atlantic creoles crisscrossed the ocean several times, as had Jerónimo, a Wolof slave, who 54  Charles Verlinden, The Beginnings of Modern Colonization: Eleven Essays with an Introduction (Ithaca, 1970), 39–40; Saunders, Black Slaves and Freedmen in Portugal, chap. 1, esp. 55; Ruth Pike, “Sevillian Society in the Sixteenth Century: Slaves and Freedmen,” Hispanic American Historical Review, 47 (1967), 344–59, and Pike, Aristocrats and Traders: Sevillian Society in the Sixteenth Century (Ithaca, 1972), 29, 170–92; P. E. H. Hair, “Black African Slaves at Valencia, 1482–1516,” History in Africa, 7 (1980), 119–31; Thornton, Africa and Africans in the Making of the Atlantic World, 96–97; A. J. R. Russell-Wood, “Iberian Expansion and the Issue of Black Slavery: Changing Portuguese Attitudes, 1440–1770,” AHR, 83 (1978), 20. During the first two decades of the 16th century, about 2,000 African slaves annually entered Lisbon and were sold there. By the 1530s, most slaves brought to Lisbon were sent to the New World via Seville. 55  In the mid-16th century, black people entered the periphery of Europe; Verlinden, Beginnings of Modern Colonization, chap 2. England developed a small black population that grew with English involvement in the African trade; see James B. Walvin, Black and White: The Negro and English Society, 1555–1945 (London, 1973), chap. 1, and F. O. Shyllon, Black Slaves in Britain (London, 1974), and Black People in Britain 1555–1833 (London, 1977). For France see William B. Cohen, The French Encounter with Africans: White Response to Blacks, 1530–1880 (Bloomington, 1980), and Sue Peabody, “ ‘There Are No Slaves in France’: Law, Culture, and Society in Early Modern France, 1685–1789” (Ph. D. diss., University of Iowa, 1993). 56  J. Fred Rippy, “The Negro and the Spanish Pioneer in the New World,” J. Negro Hist., 6 (1921), 183–89; Leo Wiener, Africa and the Discovery of America, 3 vols. (Philadelphia, 1920–1922).

1246

Berlin

was sold from Lisbon to Cartagena and from Cartagena to Murica, where he was purchased by a churchman who sent him to Valencia. A “mulâtress” wife and her three slaves followed her French husband, a gunsmith in the employ of the Compagnie des Indes, from Gorée to Louisiana, when he was deported for criminal activities.57 Other Atlantic creoles traveled on their own, as sailors and interpreters in both the transatlantic and African trades. Some gained their freedom and mixed with Europeans and Native Americans. Wherever they went, Atlantic creoles extended the use of the distinctive language of the Atlantic, planted the special institutions of the creole community, and propagated their unique outlook. Within the Portuguese and Spanish empires, Atlantic creoles created an intercontinental web of cofradias (confradias to the Spanish), so that, by the seventeenth century, the network of black religious brotherhoods stretched from Lisbon to São Tomé, Angola, and Brazil.58 Although no comparable institutional linkages existed in the Anglo- and Franco-American worlds, there were numerous informal connections between black people in New England and Virginia, Louisiana and Saint Domingue. Like their African counterparts, Atlantic creoles of European, South American, and Caribbean origins also found their way to mainland North America, where they became part of black America’s charter generations. The Dutch were the main conduit for carrying such men and women to the North American mainland in the seventeenth century. Juan (Jan, in some accounts) Rodrigues, a sailor of mixed racial ancestry who had shipped from Hispaniola in 1612 on the Jonge Tobias, offers another case in point. The ship, one of the several Dutch merchant vessels vying for the North American fur trade before the founding of the Dutch West India Company, anchored in the Hudson River sometime in 1612 and left Rodrigues either as an independent trader or, more likely, as ship’s agent. When a rival Dutch ship arrived the following year, Rodrigues promptly shifted his allegiance, informing its captain 57  Saunders, Black Slaves and Freedmen in Portugal, 29; for sailors see 11, 71–72, 145, and Hall, Africans in Colonial Louisiana, 128. A sale of 6 slaves in Mexico in 1554 included one born in the Azores, another born in Portugal, another born in Africa, and the latter’s daughter born in Mexico; Colin A. Palmer, Slaves of the White God: Blacks in Mexico, 1570–1650 (Cambridge, Mass., 1976), 31–32; “Abstracts of French and Spanish Documents Concerning the Early History of Louisiana,” Louisiana Historical Quarterly, 1 (1917), 111. 58  Saunders, Black Slaves and Freedmen in Portugal, 152–55; Russell-Wood, “Black and Mulatto Brotherhoods in Colonial Brazil,” Hisp. Amer. Hist. Rev., 54 (1974), 567–602, and Russell-Wood, The Black Man in Slavery and Freedom in Colonial Brazil (New York, 1982), chap. 8, esp. 134, 153–54, 159–60. See also Pike, Aristocrats and Traders, 177–79. In the 16th century, some 7% (2,580) of Portugal’s black population was free; Saunders, Black Slaves and Freedmen in Portugal, 59.

From Creole To African

1247

that, despite his color, “he was a free man.” He served his new employer as translator and agent collecting furs from the native population. When the captain of the Jonge Tobias returned to the Hudson River, Rodrigues changed his allegiance yet again, only to be denounced as a turncoat and “that black rascal.” Barely escaping with his life, he took up residence with some friendly Indians.59 Atlantic creoles were among the first black people to enter the Chesapeake region in the early years of the seventeenth century, and they numbered large among the “twenty Negars” the Dutch sold to the English at Jamestown in 1619 as well as those who followed during the next half century.60 Anthony Johnson, who was probably among the prizes captured by a Dutch ship in the Caribbean, appears to have landed in Jamestown as “Antonio a Negro” soon after the initial purchase. During the next thirty years, Antonio exited servitude, anglicized his name, married, began to farm on his own, and in 1651 received a 250–acre headright. When his Eastern Shore plantation burned to the ground two years later, he petitioned the county court for relief and was granted a substantial reduction of his taxes. His son John did even better than his father, receiving a patent for 550 acres, and another son, Richard, owned a 100–acre estate. Like other men of substance, the Johnsons farmed independently, held slaves, and left their heirs sizable estates. As established members of their communities, they enjoyed rights in common with other free men and frequently employed the law to protect themselves and advance their interests. When a black man claiming his freedom fled Anthony Johnson’s plantation

59  Simon Hart, The Prehistory of the New Netherland Company: Amsterdam Notarial Records of the First Dutch Voyages to the Hudson (Amsterdam, 1959), 23–26, 74–75, quotations on 80–82; Thomas J. Condon, New York Beginnings: The Commercial Origins of New Netherland (New York, 1968), chap. 1, esp. 30; Rink, Holland on the Hudson, 34, 42; Van Cleaf Bachman, Peltries or Plantations: The Economic Policies of the Dutch West India Company in New Netherland, 1623–1639 (Baltimore, 1969), 6–7. 60  Wesley Frank Craven’s investigation determined that the first black people to arrive at Jamestown were prizes taken by a Dutch man-of-war in consort with an English ship somewhere in the eastern Caribbean. Craven maintains they were born in the West Indies and stolen from there. J. Douglas Deal suggests they may have been taken from a Portuguese or Spanish slaver. Craven, White Red, and Black: The Seventeenth-Century Virginian (Charlottesville, 1971), 77–81; Deal, Race and Class in Colonial Virginia: Indians, Englishmen, and Africans on the Eastern Shore during the Seventeenth Century (New York, 1993), 163–64. In 1708, a Virginia planter remembered “that before the year 1680 what negros were brought to Virginia were imported generally from Barbados for it was very rare to have a Negro ship come to this Country directly from Africa”; Donnan, ed., Documents Illustrative of the Slave Trade, 4:89.

1248

Berlin

and found refuge with a nearby white planter, Johnson took his neighbor to court and won the return of his slave along with damages from the white man.61 Landed independence not only afforded free people of African descent legal near-equality in Virginia but also allowed them a wide range of expressions that others termed “arrogance”—the traditional charge against Atlantic creoles. Anthony Johnson exhibited an exalted sense of self when a local notable challenged his industry. Johnson countered with a ringing defense of his independence: “I know myne owne ground and I will worke when I please and play when I please.” Johnson also understood that he and other free black men and women were different, and he and his kin openly celebrated those differences. Whereas Antonio a Negro had anglicized the family name, John Johnson—his grandson and a third-generation Virginian—called his own estate “Angola.”62 The Johnsons were not unique in Virginia. A small community of free people of African descent developed on the Eastern Shore. Their names, like Antonio a Negro’s, suggest creole descent: John Francisco, Bashaw Ferdinando (or Farnando), Emanuel Driggus (sometimes Drighouse; probably Rodriggus), Anthony Longo (perhaps Loango), and “Francisco a Negroe” (soon to become Francis, then Frank, Payne and finally Paine).63 They, like Antonio, were drawn 61  Anthony Johnson’s primacy and “unmatched achievement” have made him and his family the most studied members of the charter generation. The best account of Johnson and his family is found in Deal, Race and Class in Colonial Virginia, 217–50. Also useful are T. H. Breen and Stephen Innes, “Myne Owne Ground”: Race and Freedom on Virginia’s Eastern Shore, 1640–1676 (New York, 1980), chap. 1; Ross M. Kimmel, “Free Blacks in Seventeenth-Century Maryland,” Maryland Magazine of History, 71 (1976), 22–25; Alden Vaughan, “Blacks in Virginia: A Note on the First Decade,” WMQ, 3d Ser., 29 (1972), 475– 76; James H. Brewer, “Negro Property Owners in Seventeenth-Century Virginia,” ibid., 12 (1955), 576–78; Susie M. Ames, Studies of the Virginia Eastern Shore in the Seventeenth Century (Richmond, 1940), 102–05; John H. Russell, The Free Negro in Virginia, 1619–1865 (Baltimore, 1913); and Russell, “Colored Freemen as Slave Owners in Virginia,” J. Negro Hist., 1 (1916), 233–42. Indirect evidence of the baptism of Johnson’s children comes from the 1660s, when John Johnson replied to challenges to his right to testify in court by producing evidence of baptism. He may have been baptized as an adult. 62  Quotation in Breen and Innes, “Myne Owne Ground,” 6. The statement is generally attributed to Johnson but may have been uttered by Francis Payne. See Deal, Race and Class in Colonial Virginia, 266–67. For John Johnson’s Angola see Kimmel, “Free Blacks in Maryland,” 23. 63  Deal, Race and Class in Colonial Virginia, 205–406, 265–67 (for Payne), 305 n. 2 (for the Driggus name), and Deal, “A Constricted World: Free Blacks on Virginia’s Eastern Shore, 1680–1750,” in Lois Green Carr, Philip D. Morgan, and Jean B. Russo, eds., Colonial Chesapeake Society (Chapel Hill, 1989), 275–305; Breen and Innes, “Myne Owne Ground,” esp. chap. 4, 69 (names).

From Creole To African

1249

from the Atlantic littoral and may have spent time in England or New England before reaching the Chesapeake. At least one, “John Phillip, A negro Christened in England 12 yeeres since,” was a sailor on an English ship that brought a captured Spanish vessel into Jamestown; another, Sebastian Cain or Cane, gained his freedom in Boston, where he had served the merchant Robert Keayne (hence probably his name). Cain also took to the sea as a sailor, but, unlike Phillip, he settled in Virginia as a neighbor, friend, and sometimes kinsman of the Johnsons, Drigguses, and Paynes.64 In Virginia, Atlantic creoles ascended the social order and exhibited a surehanded understanding of Chesapeake social hierarchy and the complex dynamics of patron-client relations. Although still in bondage, they began to acquire the property, skills, and social connections that became their mark throughout the Atlantic world. They worked provision grounds, kept livestock, and traded independently. More important, they found advocates among the propertied classes—often their owners—and identified themselves with the colony’s most important institutions, registering their marriages, baptisms, and children’s godparents in the Anglican church and their property in the county courthouse. They sued and were sued in local courts and petitioned the colonial legislature and governor. While relations to their well-placed patrons—former masters and mistresses, landlords, and employers—among the colony’s elite were important, as in Louisiana, the creoles also established ties among themselves, weaving together a community from among the interconnections of marriage, trade, and friendship. Free blacks testified on each other’s behalf, stood as godparents for each other’s children, loaned each other small sums, and joined together for after-hours conviviality, creating a community 64  The nature of the slave trade in the Chesapeake was summarized by Maryland’s governor in 1708: “before the year 1698, this province has been supplyd by some small Quantitys of Negro’s from Barbados and other her Ma’tys Islands and Plantations, as Jamaica and New England Seaven, eight, nine or ten in a Sloope, and sometymes larger Quantitys, and sometymes, tho very seldom, whole ship Loads of Slaves have been brought here directly from Affrica by Interlopers, or such as have had Lycenses, or otherwise traded there.” Most of the latter had arrived in the previous decade; Donnan, ed., Documents Illustrative of the Slave Trade, 4:21–23, 88–90; Menard, “From Servants to Slaves: The Transformation of the Chesapeake Labor System,” Southern Studies, 16 (1977), 363–67; Deal, Race and Class in Colonial Virginia, 164–65; Breen and Innes, “Myne Owne Ground,” 70–71. On Phillip see Robert McColley, “Slavery in Virginia, 1619–1660: A Reexamination,” in Robert H. Abzug and Stephen E. Mazlish, eds., New Perspectives on Race and Slavery in America (Lexington, Ky., 1986), 15–16, and Vaughan, “Blacks in Virginia,” 470; on Cain see Deal, Race and Class in Colonial Virginia, 254–55, 317–19, and Robert C. Twombly and Robert H. Moore, “Black Puritan: The Negro in Seventeenth-Century Massachusetts,” WMQ, 3d Ser., 24 (1967), 236.

1250

Berlin

that often expanded to the larger web of interactions among all poor people, regardless of color. According to one historian of black life in seventeenth-century Virginia, “cooperative projects … were more likely in relations between colored freedmen and poor whites than were the debtor-creditor, tenant-landlord, or employee-employer relations that linked individuals of both races to members of the planter class.”65 The horizontal ties of class developed alongside the vertical ones of patronage. Maintaining their standing as property-holding free persons was difficult, and some Atlantic creoles in the Chesapeake, like those in New Netherland, slipped down the social ladder, trapped by legal snares—apprenticeships, tax forfeitures, and bastardy laws—as planters turned from a labor system based on indentured Europeans and Atlantic creoles to raw Africans condemned to perpetual slavery. Anthony Johnson, harassed by white planters, fled his plantation in Virginia to establish the more modest “Tonies Vineyard” in Maryland. But even as they were pushed out, many of the Chesapeake’s charter generations continued to elude slavery. Some did well, lubricating the lifts to economic success with their own hard work, their skills in a society that had “an unrelenting demand for artisanal labor,” and the assistance of powerful patrons. A few of the landholding free black families on Virginia’s Eastern Shore maintained their propertied standing well into the eighteenth century. In 1738, the estate of Emanuel Driggus’s grandson—including its slaves—was worth more than those of two-thirds of his white neighbors.66 Atlantic creoles also entered the lowcountry of South Carolina and Florida, carried there by the English and Spanish, respectively. Like the great West Indian planters who settled in that “colony of a colony,” Atlantic creoles were drawn from Barbados and other Caribbean islands, where a full generation of European and African cohabitation had allowed them to gain a knowledge of European ways. Prior to the sugar revolution, they worked alongside white indentured servants in a variety of enterprises, none of which required the discipline of plantation labor. Like white servants, some exited slavery, as the line between slavery and freedom was open. An Anglican minister who toured the English islands during the 1670s noted that black people spoke English

65  Deal, Race and Class in Colonial Virginia, 205–405, quotation on 209, and Deal, “A Constricted World,” 275–305; Michael L. Nicholls, “Passing Through This Troublesome World: Free Blacks in the Early Southside,” Virginia Magazine of History and Biography, 92 (1984), 50–70. 66  Deal, Race and Class in Colonial Virginia, 225–35, quotation on 208, and Deal, “A Constricted World,” 290; Breen and Innes, “Myne Owne Ground,” 79–82, 86, 90.

From Creole To African

1251

“no worse than the natural born subjects of that Kingdom.”67 Although Atlantic creole culture took a different shape in the Antilles than it did on the periphery of Africa or Europe, it also displayed many of the same characteristics. On the southern mainland, creoles used their knowledge of the New World and their ability to negotiate between the various Native American nations and South Carolina’s European polyglot—English, French Huguenots, Sephardic Jews—to become invaluable as messengers, trappers, and cattle minders. The striking image of slave and master working on opposite sides of a sawbuck suggests the place of blacks during the early years of South Carolina’s settlement.68 Knowledge of their English captors also provided knowledge of their captors’ enemy, some two hundred miles to the south. At every opportunity, Carolina slaves fled to Spanish Florida, where they requested Catholic baptism. Officials at St. Augustine—whose black population was drawn from Spain, Cuba, Hispaniola, and New Spain—celebrated the fugitives’ choice of religion and offered sanctuary. They also valued the creoles’ knowledge of the countryside, their ability to converse with English, Spanish, and Indians, and their willingness to strike back at their enslavers. Under the Spanish flag, former Carolina slaves raided English settlements at Port Royal and Edisto and liberated even more of their number. As part of the black militia, they, along with other fugitives from Carolina, fought against the English in the Tuscarora and Yamasee wars.69 67  Morgan Godwyn, The Negro’s and Indians Advocate (London, 1680), 101, quoted in Breen and Innes, “Myne Owne Ground,” 70, 130 n. 8. 68  Wood, Black Majority, chaps. 1, 4, esp. 97, for a reference to a slave master who “worked many days with a Negro man at the Whip saw.” See also Clarence L. Ver Steeg, Origins of a Southern Mosaic: Studies of Early Carolina and Georgia (Athens, Ga., 1975), 105–07. 69  Jane Landers, “Spanish Sanctuary: Fugitives in Florida, 1687–1790, Florida Historical Quarterly, 62 (1984), 296–302, and Landers, “Gracia Real de Santa Teresa de Mose: A Free Black Town in Spanish Colonial Florida,” AHR, 95 (1990), 9–30; John J. TePaske, “The Fugitive Slave: Intercolonial Rivalry and Spanish Slave Policy, 1687–1764,” in Samuel Proctor, ed., Eighteenth-Century Florida and Its Borderlands (Gainesville, 1975), 2–12; I. A. Wright, comp., “Dispatches of Spanish Officials Bearing on the Free Negro Settlement of Gracia Real de Santa Teresa de Mose, Florida,” J. Negro Hist., 9 (1924), 144–93, quotation on 150; Zora Neale Hurston, “Letters of Zora Neale Hurston on the Mose Settlement, and Negro Colony in Florida,” ibid., 12 (1927), 664–67; J. D. Duncan, “Slavery and Servitude in Colonial South Carolina, 1670–1776” (Ph. D. diss., Emory University, 1964), chap. 17, quotation on 664; and J. G. Dunlop, “William Dunlop’s Mission to St. Augustine in 1688,” South Carolina Historical and Genealogical Magazine, 34 (1933), 1–30. Several of the slaves who rejected freedom and Catholicism in St. Augustine and returned to South Carolina were rewarded with freedom, creating a competition between English and Spanish

1252

Berlin

Florida’s small black population mushroomed in the late seventeenth and early eighteenth centuries, as the small but steady stream of fugitives grew with the expansion of lowcountry slavery. Slaves from central Africa—­ generally deemed “Angolans”—numbered large among the new arrivals, as the transatlantic trade carried thousands of Africans directly to the lowlands. Although many were drawn from deep in the interior of Africa, others were Atlantic creoles with experience in the coastal towns of Cabinda, Loango, and Mpinda. Some spoke Portuguese, which, as one Carolinian noted, was “as near Spanish as Scotch is to English,” and subscribed to an African Catholicism with roots in the fifteenth-century conversion of Kongo’s royal house. They knew their catechism, celebrated feasts of Easter and All Saint’s Day or Hallowe’en, and recognized Christian saints. These men and women were particularly attracted to the possibilities of freedom in the Spanish settlements around St. Augustine. They fled from South Carolina in increasing numbers during the 1720s and 1730s, and, in 1739, a group of African slaves—some doubtless drawn from the newcomers—initiated a mass flight. Pursued by South Carolina militiamen, they confronted their owners’ soldiers in several pitched battles that became known as the Stono Rebellion.70 Although most of the Stono rebels were killed or captured, some escaped to Florida, from where it became difficult to retrieve them by formal negotiation or by force. The newcomers were quickly integrated into black life in St. Augustine, since they had already been baptized, although they prayed— as one Miguel Domingo informed a Spanish priest—in Kikongo.71 Much to the delight of St. Augustine’s Spanish rulers, the former Carolina slaves did more than pray. They fought alongside the Spanish against incursions by English raiders. An edict of the Spanish crown promising “Liberty and Protection” to all slaves who reached St. Augustine boosted the number of fugitives—most from Carolina—especially after reports circulated that the colonies that redounded to the fugitives’ advantage. See Duncan, “Slavery and Servitude in Colonial South Carolina,” 381–83. 70  Wood, Black Majority, chaps. 11–12; Thornton, “African Dimensions of the Stono Rebellion,” AHR, 96 (1991), 1101–11, quotation on 1102. Thornton makes a powerful case for the Kongolese origins of the Stono rebels in their military organization and the nature of their resistance. For the pretransfer conversion of slaves from central Africa to Christianity see Thornton, “Development of an African Catholic Church in the Kingdom of Kongo,” 147–50, and Kingdom of Kongo, 63–68; MacGaffey, Religion and Society in Central Africa, 198–211; and Hilton, Kingdom of Kongo, 179–98. 71  Thornton, “African Dimensions of the Stono Rebellion,” 1107; Landers, “Gracia Real de Santa Teresa de Mose,” 27; Michael Mullin, ed., American Negro Slavery: A Documentary History (New York, 1976), 84.

From Creole To African

1253

Spanish received runaways “with great Honors” and gave their leaders military commissions and “A Coat Faced with Velvet.” In time, Spanish authorities granted freedom to some, but not all, of the black soldiers and their families.72 Among the unrewarded was Francisco Menéndez, a veteran of the Yamasee War and leader of the black militia. Frustrated by the ingratitude of his immediate superiors, Menéndez petitioned the governor of Florida and the bishop of Cuba for his liberty, which he eventually received. In 1738, when a new governor established Gracia Real de Santa Teresa de Mose, a fortified settlement north of St. Augustine, to protect the Spanish capital from the English incursions, he placed Menéndez in charge. Under Captain Menéndez, Mose became the center of black life in colonial Florida and a base from which former slaves—sometimes joined by Indians—raided South Carolina. The success of the black militia in repelling an English attack on Mose in 1740 won Menéndez a special commendation from the governor, who declared that the black captain had “distinguished himself in the establishment, and cultivation of Mose.” Not one to lose an opportunity, the newly literate Menéndez promptly requested that the king remunerate him for the “loyalty, zeal and love I have always demonstrated in the royal service” and petitioned for a stipend worthy of a militia captain.73 To secure his reward, Menéndez took a commission as a privateer, with hopes of eventually reaching Spain and collecting his royal reward. Instead, a British ship captured the famous “Signior Capitano Francisco.” Although stretched out on a cannon and threatened with emasculation for alleged atrocities during the siege of Mose, Menéndez had become too valuable to mutilate. His captors gave him 200 lashes, soaked his wounds in brine, and commended him to a doctor “to take Care of his Sore A-se.” Menéndez was then carried before a British admiralty court on New Providence Island, where “this Francisco that Cursed Seed of Cain” was ordered sold into slavery. Even this misadventure hardly slowed the irrepressible Menéndez. By 1752, perhaps ransomed out of bondage, he was back in his old position in Mose.74 Meanwhile, members of the fugitive community around St. Augustine entered more fully into the life of the colony as artisans and tradesmen as well as laborers and domestics. They married among themselves, into the Native 72  Landers, “Spanish Sanctuary,” 296–302; Landers, “Gracia Real de Santa Teresa de Mose,” 9–30; Duncan, “Servitude and Slavery in Colonial South Carolina,” chap. 17, quotations on 659, 663. 73  Landers, “Gracia Real de Santa Teresa de Mose,” 15–21, quotation on 20; Larry W. Kruger and Robert Hall, “Fort Mose: A Black Fort in Spanish Florida,” The Griot, 6 (1987), 39–40. 74  Landers, “Gracia Real de Santa Teresa de Mose,” 21–22, quotations on 22.

1254

Berlin

American population, and with slaves as well, joining as husband and wife before their God and community in the Catholic church. They baptized their children in the same church, choosing godparents from among both the white and black congregants. Like the Atlantic creoles in New Amsterdam about a century earlier, they became skilled in identifying the lever of patronage, in this case royal authority. Declaring themselves “vassals of the king and deserving of royal protection,” they continually placed themselves in the forefront of service to the crown with the expectations that their king would protect, if not reward, them. For the most part, they were not disappointed. When Spain turned East Florida over to the British in 1763, black colonists retreated to Cuba with His Majesty’s other subjects, where the crown granted them land, tools, a small subsidy, and a slave for each of their leaders.75 In the long history of North American slavery, no other cohort of black people survived as well and rose as fast and as high in mainland society as the Atlantic creoles. The experience of the charter generations contrasts markedly with what followed: when the trauma of enslavement, the violence of captivity, the harsh conditions of plantation life left black people unable to reproduce themselves; when the strange language of their enslavers muted the tongues of newly arrived Africans; and when the slaves’ skills and knowledge were submerged in the stupefying labor of plantation production. Rather than having to face the likes of Robert Carter and the imposition of planter domination, Paulo d’Angola, Samba Bambara, Juan Rodrigues, Antonio a Negro, and Francisco Menéndez entered a society not markedly different from those they had left.76 There, in New Netherland, the Chesapeake, Louisiana, and Florida, they made 75  Ibid., quotations on 21, 23–30; Theodore G. Corbett, “Population Structure in Hispanic St. Augustine,” Florida Historical Quarterly, 54 (1976), 265, and “Migration to a Spanish Imperial Frontier in the Seventeenth and Eighteenth Centuries: St. Augustine,” Hisp. Amer. Hist. Rev., 54 (1974), 420–21. 76  I have been unable to locate female analogues of Paulo d’Angola, Samba Bambara, Juan Rodrigues, Antonio a Negro, and Francisco Menéndez. Their absence does not, however, reflect the experience of Atlantic creoles, as small shards of evidence indicate that women played central roles in the production of creole culture, the transmission of language, the facilitation of trade, and the accumulation of capital. The best study derives from the 18th century. See George E. Brooks, Jr., “The Signares of Saint-Louis and Gorée: Women Entrepreneurs in Eighteenth-Century Senegal,” in Nancy J. Hafkin and Edna G. Bay, eds., Women in Africa: Studies in Social and Economic Change (Stanford, Calif., 1976), 19–44. For an interpretation of 17th-century Chesapeake society that stresses the critical role of women in the shaping of race relations and the emergence of slavery see Kathleen Mary Brown, “Gender and the Genesis of a Race and Class System in Virginia, 1630–1750” (Ph. D. diss., University of Wisconsin, 1990).

From Creole To African

1255

a place for themselves, demonstrating confidence in their abilities to master a world they knew well. Many secured freedom and a modest prosperity, despite the presumption of racial slavery and the contempt of their captors. The charter generations’ experience derived not only from who they were but also from the special circumstances of their arrival. By their very primacy, as members of the first generation of settlers, their experience was unique. While they came as foreigners, they were no more strange to the new land than were those who enslaved them. Indeed, the near simultaneous arrival of migrants from Europe and Africa gave them a shared perspective on the New World. At first, all saw themselves as outsiders. That would change, as European settlers gained dominance, ousted native peoples, and created societies they claimed as their own. As Europeans became European-Americans and then simply Americans, their identification with—and sense of ownership over—mainland society distinguished them from the forced migrants from Africa who continued to arrive as strangers and were defined as permanent outsiders. The charter generations owed their unique history to more than just the timing of their arrival. Before their historic confrontation with their new owner, the men and women Robert Carter purchased may have spent weeks, even months, packed between the stinking planks of slave ships. Atlantic creoles experienced few of the horrors of the Middle Passage. Rather than arriving in shiploads totaling into the hundreds, Atlantic creoles trickled into the mainland singly, in twos and threes, or by the score. Most were sent in small consignments or were the booty of privateers and pirates. Some found employment as interpreters, sailors, and grumetes on the very ships that transported them to the New World.77 Although transatlantic travel in the seventeenth and eighteenth centuries could be a harrowing experience under the best of circumstances, the profound disruption that left the men and women Carter purchased physically spent and psychologically traumatized was rarely part of the experience of Atlantic creoles. Most important, Atlantic creoles entered societies-with-slaves, not, as mainland North America would become, slave societies—that is, societies in which

77  Writing about the forced transfer of Africans to the New World, W. Jeffrey Bolster observes that “many of the slaves who left Africa with no maritime skills acquired rudimentary ones on the Middle Passage, along with some knowledge of European work-routines and social organization,” in Black Jacks: African-American Seamen in the Atlantic World, 1740– 1865 (forthcoming).

1256

Berlin

the order of the plantation shaped every relationship.78 In North America—as in Africa—Atlantic creoles were still but one subordinate group in societies in which subordination was the rule. Few who arrived before the plantation system faced the dehumanizing and brutalizing effects of gang labor in societies where slaves had become commodities and nothing more. Indeed, Atlantic creoles often worked alongside their owners, supped at their tables, wore their hand-me-down clothes, and lived in the back rooms and lofts of their houses. Many resided in towns, as did Paulo d’Angola, Samba Bambara, and Francisco Menéndez. The proportion of the mainland’s black population living in places such as New Amsterdam, Philadelphia, Charleston, St. Augustine, and New Orleans was probably higher during the first generations of settlement than it would ever be again. Urban slaves, for better or worse, lived and worked in close proximity to their owners. The regimen imposed the heavy burdens of continual surveillance, but the same constant contact prevented their owners from imagining people of African descent to be a special species of beings, an idea that only emerged with the radical separation of master and slave and the creation of the worlds of the Big House and the Quarters. Until then, the open interaction of slave and slaveowner encouraged Atlantic creoles, and others as well, to judge their enslavement by its older meaning, not by its emerging new one. The possibility of freedom had much the same effect. So long as some black people, no matter how closely identified with slavery, could still wriggle free of bondage and gain an independent place, slavery may have carried the connotation of otherness, debasement, perhaps even transgression, iniquity, and vice, but it was not social death. The success of Atlantic creoles in rising from the bottom of mainland society contradicted the logic of hereditary bondage and suggested that what had been done might be undone. The rise of plantation slavery left little room for the men and women of the charter generations. Their efforts to secure a place in society were put at risk by the new order, for the triumph of the plantation régime threatened not inequality—which had always been assumed, at least by Europeans—but debasement and permanent ostracism of the sort Robert “King” Carter delivered on that Virginia wharf. With the creation of a world in which peoples of African descent were presumed slaves and those of European descent free, people of

78  For a useful distinction between societies with slaves and slave societies see Keith Hopkins, Conquerors and Slaves: Sociological Studies in Roman History, 2 vols. (Cambridge, 1978), 1:99, and Moses I. Finley, “Slavery,” International Encyclopedia of the Social Sciences (New York, 1968), and Ancient Slavery and Modern Ideology (New York, 1980), 79–80.

From Creole To African

1257

color no longer had a place. It became easy to depict black men and women as uncivilized heathens outside the bounds of society or even humanity.79 Few Atlantic creoles entered the mainland after the tobacco revolution in the Chesapeake, the rice revolution in lowcountry Carolina, and the sugar revolution in Louisiana. Rather than being drawn from the African littoral, slaves increasingly derived from the African interior. Such men and women possessed little understanding of the larger Atlantic world: no apprenticeship in negotiating with Europeans, no knowledge of Christianity or other touchstone of European culture, no acquaintance with western law, and no open fraternization with sailors and merchants in the Atlantic trade—indeed, no experience with the diseases of the Atlantic to provide a measure of immunity to the deadly microbes that lurked everywhere in the New World. Instead of speaking a pidgin or creole that gave them access to the Atlantic, the later arrivals were separated from their enslavers and often from each other by a dense wall of language. Rather than see their skills and knowledge appreciate in value, they generally discovered that previous experience counted for little on the plantations of the New World. Indeed, the remnants of their African past were immediately expropriated by their new masters. In the stereotypes that demeaned slaves, European and European-American slaveholders inadvertently recognized the difference between the Atlantic creoles and the men and women who followed them into bondage, revealing how the meaning of race was being transformed with the advent of the plantation. Slaveholders condemned creoles as roguish in the manner of Juan Rodrigues the “black rascal,” or arrogant in the manner of Antonio a Negro, who knew his “owne ground,” or swaggering in the manner of “Signior Capitano Francisco,” who stood his ground against those who threatened his manhood. They rarely used such epithets against the postcreole generations that labored on the great plantations. Instead, slaveholders and their apologists scorned such slaves as crude primitives, devoid of the simple amenities of refined society. The failings of plantation slaves were not those of calculation or arrogance, but of stark ignorance and dense stupidity. Plantation slaves were denounced, not for a desire to convert to Christianity for “worldly and perverse aims” as were the half-free blacks in New Netherland or because they claimed the “True Faith” as did the Carolinians who fled to St. Augustine, but because they knew nothing of the religion, language, law, and social etiquette that Europeans equated with civilization. The unfamiliarity of the post-Atlantic creole cohort with 79  Jordan, White over Black: American Attitudes toward the Negro (Chapel Hill, 1968), chaps. 1–6, traces the initial appearance of such notions among the transplanted English and their later triumph.

1258

Berlin

the dynamics of Atlantic life made them easy targets for the slaveholders’ ridicule. Like the Virginia planters who slammed Africans for the “gross bestiality and rudeness of their manners,” an eighteenth-century chronicler of South Carolina’s history declared lowcountry slaves to be “as great strangers to Christianity, and as much under the influence of Pagan darkness, idolatry and superstition, as they were at their first arrival from Africa.” Such a charge, whatever its meaning on the great lowcountry rice plantations, could have no relevance to the runaways who sought the True Faith in St. Augustine.80 In time, stereotypes made were again remade. During the late eighteenth century, planters and their apologists rethought the meaning of race as more than a century and a half of captivity remolded people of African descent. As a new generation of black people emerged—familiar with the American countryside, fluent in its languages, and conversant in its religions—the stereotype of the artful, smooth-talking slave also appeared. Manipulative to the point of insolence, this new generation of African-Americans peopled the slave quarter, confronted the master on their own terms, and, in the midst of the Revolution, secured freedom. African-Americans reversed the process of enslavement— among other things, taking back the naming process (although not the names) that “King” Carter had usurped.81 Their story—whereby Africans became creoles—was a great one and one that Americans would repeat many times in the personages of men as different as David Levinsky, the Godfather, or Kunta Kinte—as greenhorns became natives. Historians, like novelists and film makers, have enjoyed retelling the tale, but in so doing, they lost the story of another founding generation and its transit from immigrant to native. While the fathers (and sometimes the mothers) of European America, whether Puritan divines or Chesapeake adventurers, would be celebrated by their posterity, members of black America’s charter generations disappeared into the footnotes of American history. Generations of Americans lived in the shadow of John Winthrop and William Byrd, even Peter Stuyvesant and Jean Baptiste Bienville, but few learned of Paulo d’Angola, 80  Quotations in Bruce, Institutional History of Virginia in the Seventeenth Century, 1:9, and Alexander Hewatt, An Historical Account of the Rise and Progress of the Colonies of South Carolina and Georgia, 2 vols. (London, 1779), 2:100. 81  Their story is told by Mullin, Flight and Rebellion, and Africa in America. On renaming see Berlin, Slaves without Masters: The Free Negro in the Antebellum South (New York, 1974), 51–52; Gary Nash, Forging Freedom: The Formation of Philadelphia’s Black Community, 1720–1840 (Cambridge, Mass. 1988), 79–88, and “Forging Freedom: The Emancipation Experience in the Northern Seaport Cities, 1775–1820,” in Berlin and Hoffman, eds., Slavery and Freedom in the Age of Revolution, 20–27; and Cheryll Ann Cody, “Kin and Community among the Good Hope People after Emancipation,” Ethnohistory, 41 (1994), 28–33.

From Creole To African

1259

Samba Bambara, Juan Rodrigues, Antonio a Negro, and Francisco Menéndez. If Atlantic creoles made any appearance in the textbook histories, it would be as curiosities and exceptions to the normal pattern of American race relations, examples of false starts, mere tokens. The story of how creoles became Africans was lost in a chronicle that presumed American history always moved in a single direction. The assimiliationist ideal could not imagine how the diverse people of the Atlantic could become the sons and daughters of Africa. The possibility that a society-withslaves was a separate and distinct social formation, not a stage in the development of slave society, was similarly inconceivable in a nation in which wealth and power rested upon plantation slavery. The causes of creole anonymity ran deep. While Carter initiated newly arrived Africans to the world of the plantation, the descendants of the charter generations struggled to maintain the status they had earlier achieved. To that end, many separated themselves from the mass of Africans on whom the heavy weight of plantation bondage fell. Some fled as a group, as did the creole community in St. Augustine that retreated with the Spanish from Florida to Cuba following the British takeover in 1764.82 Others merged with Native American tribes and European-American settlers to create unique biracial and triracial combinations and established separate identities. In the 1660s, the Johnson clan fled Virginia for Maryland, Delaware, and New Jersey. John Johnson and John Johnson, Jr., the son and grandson of Anthony Johnson, took refuge among the Nanticoke Indians and so-called Moors, among whom the Johnson name has loomed large into the twentieth century. Near one Nanticoke settlement in Delaware stands the small village of “Angola,” the name of John Johnson’s Virginia plantation and perhaps Anthony Johnson’s ancestral home. Similar “Indian” tribes could be found scattered throughout the eastern half of the United States, categorized by twentieth-century ethnographers as “tri-racial isolates.”83 82  Corbett, “Migration to a Spanish Imperial Frontier in the Seventeenth and Eighteenth Centuries,” 420; Wilbur H. Siebert, “The Departure of the Spaniards and Other Groups from East Florida, 1763–1764,” Florida Historical Quarterly, 19 (1940), 146; Robert L. Gold, “The Settlement of the East Florida Spaniards in Cuba, 1763–1766,” ibid., 42 (1964), 216–17; Landers, “Gracia Real de Santa Teresa de Mose,” 29. For the northward migration of free people of color from the Chesapeake region see Deal, Race and Class in Colonial Virginia, 188. 83   The accepted anthropological designation for the these communities is “tri-racial isolates.” Scholars have traced their origins to Virginia and North Carolina in the 17th century and then their expansion into South Carolina, Kentucky, and Tennessee with various branches moving north and south. A recent survey by Virginia Easley DeMarce

1260

Berlin

Others moved west to a different kind of autonomy. Scattered throughout the frontier areas of the eighteenth century were handfuls of black people eager to escape the racially divided society of plantation America. White frontiersmen, with little sympathy for the nabobs of the tidewater, sometimes sheltered such black men and women, employing them with no questions asked. People of African descent also found refuge among the frontier banditti, whose interracial character—a “numerous Collection of outcast Mulattoes, Mustees, free Negroes, all Horse-Thieves,” by one account—was the subject of constant denunciation by the frontier’s aspiring planters.84 While some members of the charter generations retreated before the expanding planter class, a few moved toward it. At least one male member of every prominent seventeenth-century free black family on the Eastern Shore of Virginia married a white woman, so the Atlantic creoles’ descendants would, perforce, be lighter in color. Whether or not this was a conscious strategy, there remains considerable, if necessarily incomplete, evidence that these lightskinned people employed a portion of their European inheritance—a pale complexion—to pass into white society.85 Retreat—geographic, social, and physical—was not the only strategy members of the charter generations adopted in the face of the emergent plantation régime. Some stood their ground, confronting white authorities and perhaps setting an example for those less fortunate than themselves. In 1667, claiming “hee was a Christian and had been severall years in England,” a black man named Fernando sued for his freedom in a Virginia court. The case, initiated just as tidewater planters were consolidating their place atop Virginia society, sent Virginia lawmakers into a paroxysm that culminated in the passage of a new law clarifying the status of black people: they would be slaves for life and their status would be hereditary. In succeeding years, such Atlantic provides an excellent overview; “ ‘ Verry Slitly Mixt’: Tri-Racial Isolate Families of the Upper South—A Genealogical Study,” National Genealogical Society Quarterly, 80 (1992), 5–35. 84  Rachel N.  Klein, Unification of a Slave State: The Rise of the Planter Class in the South Carolina Backcountry, 1760–1808 (Chapel Hill, 1990), 18–21, 62–72. 85  For Johnson’s whitening see Deal, Race and Class in Colonial Virginia, 258–69, esp. 277. See, for example, the case of Gideon Gibson, a mulatto slaveholder who during the mid18th century was in the process of transforming himself from “black” to “white,” in Jordan, White over Black, 171–74; Klein, Unification of a Slave State, 69–71; Robert L. Meriwether, The Expansion of South Carolina, 1729–1765 (Kingsport, Tenn., 1940), 90, 96. As a group, free people of color were getting lighter in the Chesapeake during the late 17th century and into the 18th, perhaps as part of a conscious strategy of successful free men who married white women. See, for example, Deal, Race and Class in Colonial Virginia, 187, 276 n. 20, and Berlin, Slaves without Masters, 3–4.

From Creole To African

1261

creoles—men and women of African descent with long experience in the larger Atlantic world—would continue as Fernando continued to bedevil planters and other white Americans in and out of the court room, harboring runaway slaves, providing them with free papers, and joining together matters slaveholders viewed as subversive. In 1671, New York authorities singled out Domingo and Manuel Angola, warning the public “that the free negroes were from time to time entertaining sundry of the servants and negroes belonging to the Burghers … to the great damage of their owners.” It appears that the warning did little to limit black people from meeting, for several years later New York’s Common Council again complained about “the frequent randivozing of Negro Slaves att the houses of free negroes without the gates hath bin occasion of great disordr.” As slaveholders feared, the line between annoyance and subversion was a thin one. Atlantic creoles were among the black servants and slaves who stood with Nathaniel Bacon against royal authority in 1676.86 The relentless engine of plantation agriculture and the transformation of the mainland colonies from societies-with-slaves to slave societies submerged the charter generations in a régime in which African descent was equated with slavery. For the most part, the descendants of African creoles took their place as slaves alongside newly arrived Africans. Those who maintained their freedom became part of an impoverished free black minority, and those who lost their liberty were swallowed up in an oppressed slave majority.87 In one way or another, Atlantic creoles were overwhelmed by the power of the plantation order.

86  Quotation from William W. Hening, comp., The Statutes at Large: Being a Collection of All the Laws of Virginia, 13 vols. (Richmond, 1800–1823) 2:260; Warren M. Billings, “The Cases of Fernando and Elizabeth Key: A Note on the Status of Blacks in Seventeenth-Century Virginia,” WMQ, 3d Ser., 30 (1973), 467–74; Billings, ed., The Old Dominion in the Seventeenth Century: A Documentary History of Virginia, 1606–1689 (Chapel Hill, 1975), 165–69; David W. Galenson, “Economic Aspects of the Growth of Slavery in the Seventeenth-Century Chesapeake,” in Solow, ed., Slavery and the Rise of the Atlantic System, 271; Fernow, ed., Records of New Amsterdam, 6:146, 286; Herbert L. Osgood, ed., Minutes of the Common Council of the City of New York, 8 vols. (New York, 1905), 1:134, 276–77; J. B. Lyon, ed., Colonial Laws of New York from 1664 to the Revolution, 5 vols. (Albany, 1894–1896), 1:356–57; Goodfriend, Before the Melting Pot, 120–21. After warning Domingo and Manuel Angola not to repeat their behavior, the court ordered them to communicate its admonition to “the other remaining free negroes”; ibid. 87  Goodfriend, Before the Melting Pot, 116–17; Nicholls, “Passing Through This Troublesome World,” 50–53; Deal, “A Constricted World,” 275–305; Breen and Innes, “Myne Owne Ground,” chaps. 4–5.

1262

Berlin

Even so, the charter generations’ presence was not without substance. During the American Revolution, when divisions within the planter class gave black people fresh opportunities to strike for liberty and equality, long-suppressed memories of the origins of African life on the mainland bubbled to the surface, often in lawsuits in which slaves claimed freedom as a result of descent from a free ancestor, sometimes white, sometimes Indian, sometimes free black, more commonly from some mixture of these elements.88 The testimony summoned by such legal contests reveals how the hidden history of the charter generations survived the plantation revolution and suggests the mechanisms by which it would be maintained in the centuries that followed. It also reveals how race had been constructed and reconstructed in mainland North America over the course of two centuries of African and European settlement and how it would be remade. 88  Berlin, Slaves without Masters, 33–34; Shane White, Somewhat More Independent: The End of Slavery in New York City, 1770–1810 (Athens, Ga., 1991), 117–18; Nash and Jean R. Soderlund, Freedom by Degrees: Emancipation in Pennsylvania and Its Aftermath (New York, 1991), 115– 36. Also see the papers of the Pennsylvania Society for Promoting the Abolition of Slavery (Historical Society of Pennsylvania), and the New York Manumission Society (New-York Historical Society), for the upsurge of suits for freedom. For naming patterns within free black families that reached from the Revolutionary era back to the mid-17th century see Deal, Race and Class in Colonial Virginia, 342.

Task and Gang Systems: The Organization of Labor on New World Plantations Philip D. Morgan To think of slaves and of plantation labor is invariably to envision gangs of laborers toiling from sunup to sundown, working in unison, under direct and close supervision. To picture the gang at work is to visualize them ranged in a line, proceeding in regular rows, the pace being set by one or two key laborers, urged on by a driver or foreman. It is to see, through Janet Schaw’s eyes, “a number of Negroes follow[ing] each other’s tail the day long”; it is to see, as one southern cotton planter described it, slaves “all keeping up with their leader”; and it is to see, in Henry Harcourt’s words, slaves “drawn out in a line, like troops on a parade, with their driver and his whip close at hand … all work[ing] or paus[ing] together.” This gang system, according to U. B. Phillips, was employed on “virtually all of the tobacco, short staple cotton and sugar plantations” of the New World.1

Source: Morgan, Philip D., “Task and Gang Systems: The Organization of Labor on New World Plantations,” in Stephen Innes (ed.), Work and Labour in Early America, Chapel Hill: University of North Carolina Press, 1988, 189–220. Published for the Omohundro Institute for Early American History and Culture. Copyright © 1988 by the University of North Carolina Press. Used by permission of the publisher. www.uncpress.unc.edu. * The author wishes to thank many friends for their comments and advice on earlier drafts of this paper. Michael Craton, Barry Higman, Stephen Innes, Sidney Mintz, and Marcus Rediker made particularly valuable suggestions. Stanley Engerman, always the model of generosity and intelligence, raised so many penetrating questions that it will require another essay to answer all of them. Naturally, therefore, the author takes full responsibility for the remaining inadequacies of this version. A condensed version of this essay appeared in Élise Marienstras and Barbara Karsky, eds., Autre temps, autre espace / An Other Time, an Other Space: Études sur l’Amérique pré-industrielle (Nancy, 1986), 147–162. 1  [Janet Schaw], Journal of a Lady of Quality, Being the Narrative of a Journey from Scotland to the West Indies, North Carolina, and Portugal, in the Years 1774 to 1776, ed. Evangeline Walker Andrews and Charles McLean Andrews (New Haven, Conn., 1923), 163; James Townes, “Farm Management,” Southern Cultivator, V (Apr. 1847), 26; Henry Harcourt, The Adventures of a Sugar-Plantation (London, 1836), 90–92; Ulrich Bonnell Phillips, American Negro Slavery: A Survey of the Supply, Employment, and Control of Negro Labor as Determined by the Plantation Regime (Baton Rouge, La., 1966 [orig. publ. New York, 1918]), 247.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_040

1264

Morgan

As we will discern, however, Phillips oversimplified matters even with respect to these three staples. More important, there was a variety of other New World products—arrowroot, cocoa, coffee, long-staple cotton, hemp, naval stores, pimento, rice, and timber—for which another system, that of tasking, was generally employed. So prevalent did tasking seem to a Jamaican writing in the early nineteenth century that he thought it “practised among the American planters in Louisiana, and in some parts of Virginia and North-Carolina, and perhaps, also in Maryland, as well as in South-Carolina and Georgia.” He also believed that “the Spaniards task their negroes in gathering gold dust.”2 For this Caribbean planter, it was the sugar producer who was out of step with his contemporaries, not vice versa. And when we think of the form of labor organization known as tasking, we envisage slaves working neither in unison nor under close supervision; rather, we should picture them being assigned a certain amount of work for the day or perhaps week, upon the completion of which they were free to use their time as they pleased.3 This essay will outline some of the features of these two systems, investigate their origins, and suggest some of their consequences. The consequences, the implications of these different labor systems for those who were involved in them, constitute my ultimate quarry. Of particular concern are the domestic economies of the slaves—their attitudes toward resource allocation, their attempts to acquire property, and their patterns of consumption—which were related, in some cases very closely, to different modes of labor organization. The broader context for this essay, then, is an investigation into the relationship between the work styles of slaves (of which only one component would be the labor system) and their cultures (within which their domestic economy is my primary focus) in a number of different societies up to and immediately beyond emancipation.4 Tasking and ganging were not hermetically sealed systems. Rather, we should think of them as two extremes in a range of labor systems, represented by individual tasking at one pole and large-scale, regimented ganging at the other. In between these two ideal types were various hybrids, most notably a 2  An Essay on Task-Work: Its Practicability and the Modes to Be Adopted for Its Application to Different Kinds of Agricultural Labour (Spanish Town, Jamaica, ca. 1809), i, 8. This pamphlet is extremely informative. The only copy that I have managed to locate is in the Houghton Library, Harvard University. 3  The best study of task and gang systems, upon which this essay attempts to build, is Lewis Cecil Gray, History of Agriculture in the Southern United States to 1860, 2 vols. (Gloucester, Mass., 1958 [orig. publ. Washington, D.C., 1933]), I, 550–556. 4  I hope to write a book on this subject, the prolegomenon of which is this essay.

Task And Gang Systems

1265

system of collective tasking on the one hand and small, relatively unsupervised gangs or squads on the other. If we take these two ideal types and the two most distinctive hybrids, and if we remember that in any place and at any time more than one would likely be present (perhaps even in the cultivation of the same staple), can we identify particular plantation staples and societies with each one? Individual tasking arrangements were associated with many more slave societies than the stereotypical picture of slave labor allows. Rice and Sea Island cotton production in the lowcountry of South Carolina and Georgia generated an extensive and deeply rooted task system. According to the anonymous Jamaican writing in the early nineteenth century, “In South Carolina and Georgia, it appears that task-work is the mode in which all their negroes are put to labour, and that it is reduced to a complete system; our mode of working them in a row being altogether unknown, or at least never practised.”5 The basic task unit, the quarter-acre, became so ingrained in lowcountry life that ex-slaves in the twentieth century still thought in terms of it.6 Long-staple cotton production in various Caribbean societies, like that of the lowcountry, also generated a task system. In Bahamian cotton planting, work was, according to one visitor to the islands at the beginning of the nineteenth century, “allotted to [the slaves] … daily and individually, according to their strength; and if they are so diligent as to have finished it at an early hour, the rest of the day is allowed to them.”7 In 1823 Bahamian slaveowners claimed that task-work had been universal in the islands “ ‘within the memory of the oldest of us,’ for all slaves except those employed as domestics, sailors, or tradespeople.” In Grenada, taskwork was first introduced by the planters of Carriacou, who were cotton planters.8 Timber or naval stores production—as the Bahamian case also illustrates— was associated with tasking. In places as remote from one another as the forests of British Honduras, North Carolina, and Surinam, taskwork was the norm for slave workers. The work of mahogany cutters in British Honduras was said 5  Essay on Task-Work, 8. 6  Philip D. Morgan, “Work and Culture: The Task System and the World of Lowcountry Blacks, 1700 to 1880,” William and Mary Quarterly, 3d Ser., XXXIX (1982), 566. 7  Daniel McKinnen, A Tour through the British West Indies, in the Years 1802 and 1803. Giving a Particular Account of the Bahama Islands (London, 1804), 172. 8  An Official Letter from the Commissioners of Correspondence of the Bahama Islands (Nassau, 1823), 41; St. George’s Chronicle and Grenada Gazette, Aug. 24, 1833. Both these references are from B. W. Higman, Slave Populations of the British Caribbean, 1807–1834 (Baltimore, 1984), 179.

1266

Morgan

to be almost entirely by task.9 Each operation in the North Carolina turpentine work cycle had its own quota: an axman ought to be able to “box” 125 trees daily, “chip” about 12,000 trees a week, and “dip” at least four barrels daily. Each slave was allocated what one early nineteenth-century observer referred to as a “district” of some 3,000–5,000 trees, which formed the slave’s stint. Many slaves calculated their task by the week and had completed it by midafternoon on Friday. As a result of such a system in Surinam, slaves on timber estates enjoyed a significant “measure of freedom” and thereby “developed a strong feeling of independence.”10 In the Caribbean, coffee and tasking generally went hand in hand. James Kelly, a bookkeeper on a Jamaican coffee plantation, noted that almost all the work was “by task.” The slaves, he continued, “weed the coffee-field, picked the coffee from the trees, billed the savannah or pastures, and built stone walls, by task.” According to a Jamaican, writing in the early nineteenth century, “taskwork had been practised for many years” on the island’s coffee plantations. According to him, coffee fields were “laid off so as to form square acres or half acres” by means of stakes, known appropriately as “task-sticks.” In fact, land in Jamaican coffee plantations (like that in lowcountry rice and Sea Island cotton estates) “was laid off by what they call the task acre, which is somewhat larger than the statute-acre,” so that more regular divisions—of 105 square feet for a quarter-acre and of 150 square feet for a half-acre—could be employed. This Jamaican specified quotas for a whole host of operations from pruning the trees to picking the berries.11 Henry de La Beche thought that industrious Jamaican slaves could complete their assignment of picking coffeeberries by 9  See, once again, Higman, Slave Populations, 179. 10  Thomas Gamble, comp., Naval Stores: History, Production, Distribution, and Consumption (Savannah, Ga., 1921), 11, 26–27; Percival Perry, “The Naval Stores Industry in the AnteBellum South, 1789–1861” (Ph.D. diss., Duke University, 1947), 22–26, 30–33, 40, 74; Luigi Castiglioni, Luigi Castiglioni’s Viaggio: Travels in the United States of North America, 1785– 1787, trans. and ed. Antonio Pace (Syracuse, N.Y., 1983), 181; Adam Hodgson, Letters from North America, Written during a Tour in the United States and Canada (London, 1824), I, 39; Herbert A. Kellar, ed., Solon Robinson, Pioneer and Agriculturalist: Selected Writings (Indianapolis, 1936), II, 222. For tasking in the late-19th-century lowcountry turpentine industry, see Thomas F. Armstrong, “The Transformation of Work: Turpentine Workers in Coastal Georgia, 1865–1901,” Labor History, XXV (1984), 526–528. R. A. J. van Lier, Frontier Society: A Social Analysis of the History of Surinam (The Hague, 1971), 152. 11  James Kelly, Voyage to Jamaica; Being a Narrative of Seventeen Years’ Residence in That Island … (Belfast, 1838), 18; Essay on Task-Work, 10, 14, 21–24. See also B. W. Higman, Slave Population and Economy in Jamaica, 1807–1834 (Cambridge, 1976), 21–24. For the lowcountry task acre, see my “Work and Culture,” WMQ, 3d Ser., XXXIX (1982), 581–583.

Task And Gang Systems

1267

one or two in the afternoon. Apparently, other slaves on coffee estates also found their tasks relatively congenial. In early nineteenth-century Berbice, ten slave women reported that their master gave them as “a task one hundred [coffee] trees to be weeded and cleansed.” They noted that “with this [task] we were satisfied.”12 The planting of many lesser crops, as well as a number of ancillary plantation activities, was accomplished by slaves organized into tasks. Pimento, hemp, and arrowroot production was generally typified by task systems. Even on sugar estates, ancillary operations, like the cutting of copper wood or the building of stone walls, were completed by task.13 Moreover, artisans in many slave societies were tasked in their work. Coopers in societies as distant as Jamaica and Louisiana were set quotas of so many barrels a day. Slave women were tasked in spinning and weaving. Slave woodworkers in the island of Mauritius, noted Charles Telfair, could finish their task of sawing planks in “little more than a half a day’s labour.”14 The archetypal form of gang labor was associated with sugar production. There was no better-brigaded, better-supervised form of labor. In cutting cane, for example, William Beckford marveled at the “regular discipline in the work.” Dividing the labor force into two, perhaps three gangs (common for most of the seventeenth and eighteenth centuries), the West Indian sugar planter placed great emphasis on the abilities of his “first row” men to set the pace for the workers and on his superintendents to monitor their progress.15 12  H. T. de La Beche, Notes on the Present Condition of the Negroes in Jamaica (London, 1825), 19; Zachary Macauley, The Slave Colonies of Great Britain; or, A Picture of Negro Slavery Drawn by the Colonists Themselves; Being an Abstract of the Various Papers Recently Laid before Parliament on That Subject (London, 1825), 154. 13  Higman, Slave Population and Economy, 24–25; Gray, History of Agriculture, I, 552; J. S. Handler, “The History of Arrowroot and the Origin of Peasantries in the British West Indies,” Journal of Caribbean History, II (1971), 76; Essay on Task-Work, 2. 14  Joseph Carlyle Sitterson, Sugar Country: The Cane Sugar Industry in the South, 1753–1950 (Louisville, Ky., 1953), 112; interview with Catherine Cornelius, Federal Writers’ Project Archives, MS Dept., Earl K. Long Library, University of New Orleans, cited in Roderick Alexander McDonald, “ ‘Goods and Chattels’: The Economy of Slaves on Sugar Plantations in Jamaica and Louisiana” (Ph.D. diss., University of Kansas, 1981), 117; J. B. Moreton, Manners and Customs in the West India Islands (London, 1790), 148. Perry, “Naval Stores Industry,” 76, 78; Gray, History of Agriculture, I, 552: Charles Telfair, Some Account of the State of Slavery at Mauritius, since the British Occupation in 1810 … (Port Louis, Mauritius, 1830), 46. 15  William Beckford, A Descriptive Account of the Island of Jamaica … (London, 1790), II, 47–48; Woodville K. Marshall, ed., The Colthurst Journal: Journal of a Special Magistrate in the Islands of Barbados and St. Vincent, July 1835–September 1838 (Millwood, N.Y., 1977),

1268

Morgan

The regimentation of sugar planting was legendary or infamous, depending on the observer’s point of view. In digging cane holes, for instance, critics likened the slaves to “a regiment drawn out and exercised on parade,” with the driver at one end to “keep them in a regular string.” Others described the experience more positively: the slaves, “accompanied by rhythmic song, were said to have raised their gleaming hoes all together, and in as exact a time as the performance of a well-conducted orchestra.” Thomas Roughley recommended that the “entire gang … be spread” out; in that way, they “have a more sightly and better marshalled appearance.” But, whatever the perspective, there was no gainsaying that, in James Stephen’s words, “the whole line of holers advance together [and that] it is necessary that every hole or section of the trench should be finished in equal time with the rest.”16 Where the nature of the work did not permit the slaves to be drawn up in a line abreast, as in carrying canes or grass from the fields, they were, instead, “marched in files, each with a bundle on his head, and with the driver in the rear.” This led Richard Ligon to marvel at “the lovely sight [of] … a hundred handsome Negroes, men and women, with every one a grasse-green bunch … on their heads … all coming in a train one after another.”17 So pervasive was ganging in Caribbean sugar societies that even mulemen were organized into such groups under the supervision of a “headman,” as were various “trade 143. The literature on the 17th- and 18th-century sugar gang system is immense. For some of the more important works, see Bryan Edwards, The History, Civil and Commercial, of the British Colonies in the West Indies, 2 vols. (London, 1794), II, 132–135; J. Harry Bennett, Bondsmen and Bishops: Slavery and Apprenticeship on the Codrington Plantations of Barbados, 1710–1838, University of California Publications in History, LXII (Berkeley, 1958), 11–15; Elsa V. Goveia, Slave Society in the British Leeward Islands at the End of the Eighteenth Century (New Haven, Conn., 1965), 129–151; Michael Craton and James Walvin, A Jamaican Plantation: The History of Worthy Park, 1670–1970 (London, 1970), 95–154; Richard S. Dunn, Sugar and Slaves: The Rise of the Planter Class in the English West Indies, 1624–1713 (Chapel Hill, N.C., 1972), 188–222; Higman, Slave Population and Economy, 188–189; Jerome S.  Handler and Frederick W. Lange, Plantation Slavery in Barbados: An Archaeological and Historical Investigation (Cambridge, Mass., 1978), 33–34, 72–83. 16  James Stephen, The Slavery of the British West India Island Colonies…, 2 vols. (London, 1824), I, 54; Richard B. Sheridan, Sugar and Slavery: An Economic History of the British West Indies, 1623–1773 (Baltimore, 1974), 258; Thomas Roughley, The Jamaica Planter’s Guide; or, A System for Planting and Managing a Sugar Estate, or Other Plantations … (London, 1823), 276. See also William Dickson, Letters on Slavery (London, 1789), 23; James Grainger, The Sugar-Cane: A Poem (London, 1766), 22; James Stephen, The Crisis of the Sugar Colonies … (London, 1802), 10 (quotation). 17  Stephen, Crisis of the Sugar Colonies, 11; Richard Ligon, A True and Exact History of the Island of Barbadoes (London, 1673), 43.

Task And Gang Systems

1269

gangs” of carpenters, masons, smiths, and the like. West Indians even thought in terms of gang labor where it did not, in fact, exist. In an agreement drawn up in 1835 between the proprietor and newly emancipated slaves of the small island of Barbuda, a graduated wage scale was established. It specified payments according to the freedmen’s capabilities as members of the “first class gang,” “weeding gang,” and “grass gang.” Barbuda, however, contained no sugar estates, and Barbudan slaves had engaged in a wide variety of occupations, none of which were characterized by gang labor.18 If the association of sugar and gangs was almost axiomatic, the relationship was not a static, unchanging one. The more progressive planters conceded that an inability or unwillingness to grade the members of a gang carefully could impair efficiency and create hardships for less able slaves. The most obvious remedy—one certainly emphasized by late-eighteenth- and early nineteenthcentury treatises on plantation husbandry—was to increase the number of gangs. Consequently, some planters in those years experimented with as many as four to six gangs.19 Improvements were particularly sought in the arduous operation of cane holing. Proslavery advocate Alexander Barclay claimed that individual slaves were generally given distinct sections of land to hole and were not always constrained to keep in line; others had their slaves work in pairs of one stronger and one weaker laborer, thereby equalizing their efforts. However, the gang generally still began and ended work together and was just as much subject to constant surveillance.20 But some Caribbean sugar planters went further, for, in the late eighteenth and early nineteenth century, some introduced individual tasking to at least the cane-holing operation. In the late eighteenth century, Bryan Edwards advocated individual tasking but also noted that “this is not always practicable.” John Baillie, testifying in the early nineteenth century, described how a 18  Roughley, Jamaica Planter’s Guide, 110–112; Marly; or, The Life of a Planter in Jamaica … (Glasgow, 1828), 93–94; Bennett, Bondsmen and Bishops, 11; Michael Craton, Searching for the Invisible Man: Slaves and Plantation Life in Jamaica (Cambridge, Mass., 1978), 208. The agreement is enclosed with MacGregor’s Dispatch to Aberdeen, May 16, 1835, CO 7/41, Public Record Office, London, as cited in Douglas Hall, Five of the Leewards, 1834–1870: The Major Problems of the Post-emancipation Period in Antigua, Barbuda, Montserrat, Nevis, and St. Kitts (Barbados, 1971), 66. 19  Dr. Collins, Practical Rules for the Management and Medical Treatment of Negro Slaves in the Sugar Colonies (Freeport, N.Y., 1971 [orig. publ. London, 1803]), 151; William Dickson, Mitigation of Slavery … (Miami, Fla., 1969 [orig. publ. London, 1814]), 13–14; Craton, Searching for the Invisible Man, 139–140. 20  Alexander Barclay, A Practical View of the Present State of Slavery in the West Indies … (London, 1826), 310; Edwards, History of British Colonies, II, 210; Marly, 165.

1270

Morgan

Jamaican slave digging cane holes “does his day’s work [by about two or three in the afternoon] … and is at liberty to go where he pleases afterwards.” On the other hand, he added that they “can scarcely give task-work” in other areas of the sugar cycle.21 Clement Caines indicates the limits to which progressive minds generally extended. In 1800 he advocated dispensing with the whole process of “lining” in cane-hole digging but acknowledged that this was only feasible with a small row of slaves “on account of the difficulty of preserving a considerable and lengthened row of people in regular order, without a surer guide than their own attention.” For the most part, rows and regularity were not to be sacrificed to the new “improving” spirit.22 Bahian sugar planters do not appear to have followed the example of their Caribbean counterparts by introducing tasking into cane holing. Indeed, as Stuart Schwartz has emphasized, “field labor in Brazil was essentially gang labor.” However, as early as the seventeenth century, this region’s sugar planters introduced piecework into cane cutting as one way of stimulating labor productivity. According to Schwartz: “Slaves were assigned a certain number of ‘hands’ of cane to be cut as their tarefa or daily task…. At the completion of the tarefa de Corte (daily cane-cutting quota), the slave was free to spend time as he or she liked. Much of the labor on Bahian plantations was assigned by tarefa, not only in the cutting of cane but in the mill, at the pottery, and elsewhere.” Important as this development was, it was not individual tasking. Teams or pairs of slaves, rather than individuals, formed the basic production unit in cane cutting.23 During the early nineteenth century and in colonies where the British government could exert pressure—as in the crown colonies of Trinidad, DemeraraEssequibo, Berbice, and Mauritius—tasking was introduced more widely as part of a broader amelioration campaign to reduce the hours worked by slaves. Advocates of tasking argued that its introduction would increase profitability and reduce the costs of supervision. By the mid-1820s, the system had been adopted on many sugar estates in Trinidad and was widespread in Berbice. By the late 1820s, it was “becoming pretty general” in Demerara. James Alexander, 21  Edwards, History of British Colonies, II, 151; Great Britain, Parliamentary Papers, Minutes of Evidence Taken before the Select Committee of the House of Lords Appointed to Inquire into the Laws and Usages of the Several West India Colonies in Relation to the Slave Population, no. 127 (1831–1832), 47. See also Josiah Condor, Wages or the Whip: An Essay on the Comparative Cost and Productiveness of Free and Slave Labour (London, 1833), 75–81. 22  Clement Caines, Letters on the Cultivation of the Otaheite Cane … (London, 1801), 51. 23  Stuart B.  Schwartz, Sugar Plantations in the Formation of Brazilian Society: Bahia, 1550–1835 (New York, 1985), 109, 139–145, 154–155.

Task And Gang Systems

1271

who visited Essequibo in the early 1830s, saw slaves weeding a cane plot by tasks, which were “frequently completed at three in the afternoon.”24 Similarly, in Mauritius, occupied by the British from 1810 onward, “the adoption of tasks, whenever practicable,” noted Charles Telfair, “augmented the amount of work performed, … simplified the duties of the overseer, [and allowed] … many of the more handy negroes … to perform their portions before four o’clock, and some even before two o’clock p.m.” According to Mr. H. Hunter, testifying before the House of Commons in the mid-nineteenth century, tasks were well established for most operations on Mauritian sugar estates: the “daily tasks of a slave in digging 100 holes; in cutting cane, two to three cart-loads, cut and cleaned; and in cleaning 800 feet by four or five feet of ground.”25 In short, there was a perceptible movement, more prevalent in some of the newer British slave societies, from gang to task, even in the sugar regions. Plantations in North and South America that grew tobacco, short-staple cotton, and coffee shared many of the same features of the archetypal Caribbean gang system. During the planting season, for example, planters in the cotton South generally divided their slaves into two gangs of plow and hoe hands, respectively.26 The role of leaders (or headmen) was as critical in cotton as in sugar cultivation. As one Mississippi planter exhorted: Have a headman among your plowers; hold him responsible for the quantity of work, at least, and you might for the quality too…. Make one of your best hands leader of the hoe hands, and class your hands, so that all may keep up with the leader. Three women will carry two rows, and if first rate women, four will carry three rows…. When your rows are of

24  Higman, Slave Populations, 179; Sir James Edward Alexander, Transatlantic Sketches, Comprising Visits to the Most Interesting Scenes in North and South America, and the West Indies … (London, 1833), I, 98. 25  Telfair, Some Account of Slavery at Mauritius, 33; Great Britain, House of Commons, Reports from the Select Committee on Sugar and Coffee Planting (1848), I, 215. See also Moses D. E. Nwulia, The History of Slavery in Mauritius and the Seychelles, 1810–1875 (Rutherford, N.J., 1981). 26  Phillips, American Negro Slavery, 237, 240, 268; Gray, History of Agriculture, I, 548–549. The best way to explore upland cotton ganging is through plantation diaries: see, for example, Franklin L. Riley, “Diary of a Mississippi Planter,” Publications of the Mississippi Historical Quarterly, X (1909), 305–481; Ulrich Bonnell Phillips and James D. Glunt, eds., Florida Plantation Records from the Papers of George Noble Jones (St. Louis, Mo., 1927); Edwin Adams Davis, Plantation Life in the Florida Parishes of Louisiana, 1836–1846, as Reflected in the Diary of Bennet H. Barrow (New York, 1967 [orig. publ. 1943]).

1272

Morgan

different lengths, make the leader take alternately the longest and shortest rows—this will enable them to keep together.27 The same imperatives were at work in Chesapeake tobacco cultivation. When a line of female laborers was found to be weeding too few rows, their Virginia master, Landon Carter, promised the foreman “a sound correction” unless he “mended his pace.” Under a gang arrangement, an individual’s place in the line determined the speed at which he worked. Thus, when this same Chesapeake planter wished to demonstrate to his disgraced gardener that “the hoe [was not] to be his field of diversion,” he gave him “the place of my fourth man and have ordered my overseers to keep him to that.” He smugly observed that this “made him quicken the motion of his arm.”28 Coffee planters in nineteenth-century Brazil also organized their slave complements into gangs. During the twice-annual weedings, the four best hands acted as lead-row men, working as pacesetters for those slower workers sandwiched in between. Each gang moved in line from the bottom to the top of hillsides on which coffee trees were always planted. As Warren Dean describes coffee production in the Rio Claro region of Brazil: “This was gang labor, supervised by feitores who were sometimes slaves themselves, who maintained discipline mainly with insults and threats, with the whip as a last resort…. The work day was sun to sun, and after dark during harvest.”29 As in the case of the sugar regime, planters of tobacco, short-staple cotton, and coffee tended either to introduce elements of individual tasking into their operations or, in some cases, to adopt the system wholesale. This tendency became more widespread as time wore on and was particularly noticeable in the nineteenth century. Picking short-staple cotton, collecting coffeeberries, husking corn, and stripping and prizing tobacco were operations that lent themselves to individual tasking, even when most of the other work with these

27  Townes, “Farm Management,” Southern Cultivator, V (Feb. 1847), 61 (quotation); see also V (Apr. 1847), 26. 28  Jack P. Greene, ed., The Diary of Colonel Landon Carter of Sabine Hall, 1752–1778, 2 vols. (Charlottesville, Va., 1965), I, 397, 430. 29  Stanley J.  Stein, Vassouras: A Brazilian Coffee County, 1850–1900 (New York, 1976 [orig. publ. Cambridge, Mass., 1957]), 34, 163; Warren Dean, Rio Claro: A Brazilian Plantation System, 1820–1920 (Stanford, Calif., 1976), 64. The careful reader will have noted that coffee and cotton seem to have given rise both to task and gang systems. I will attempt to explain this in the following section.

Task And Gang Systems

1273

staples was effected by slaves organized into gangs.30 But, by the early nineteenth century, some upland cotton planters had established individual tasks for virtually all of their operations—from plowing through hoeing to picking the cotton. It would be surprising if some coffee and tobacco planters were not experimenting along similar lines.31 Although gang systems, then, had a tendency to employ tasking in some operations and even, in some cases, to evolve into full-scale task systems, there were still two identifiable intermediate stages. The first remained a gang system, but one where the gangs were small and relatively unsupervised. To some extent, this pattern could be found on many tobacco quarters, for the size of their laboring units, generally, was small, but it was most typically found on those plantations engaged in a truly diversified farm routine. Such plantations were most common in the temperate Chesapeake region, particularly in the late eighteenth and nineteenth centuries, when tobacco no longer reigned supreme, at least in the tidewater.32 Since the labor force on such plantations could be divided into as many as six or more units, each composed of about one to five slaves, routinization of labor or the formation of permanent hoe or plow gangs was impossible. Squads of slaves performed a variety of functions, and, although there might have been one slave setting the tempo of the work, there were so few slaves in a group that the notion of their being driven is probably inappropriate. In practice, such an arrangement might closely approximate individual tasking, for many a slave would work alone, completing several jobs during the day. The widespread practice of hiring slaves reinforced the trend toward individual autonomy.33 30  This assessment is based on my incomplete analysis of 19th-century southern agricultural periodicals. For a small sample, see James O. Breeden, ed., Advice among Masters: The Ideal in Slave Management in the Old South (Westport, Conn., 1980). 31  Gray, History of Agriculture, I, 551–554. 32  The literature on Chesapeake diversification is large. Some of the best works include Paul G. E. Clemens, The Atlantic Economy and Colonial Maryland’s Eastern Shore: From Tobacco to Grain (Ithaca, N.Y., 1980); Carville V. Earle, The Evolution of a Tidewater Settlement System: All Hallow’s Parish, Maryland, 1650–1783, University of Chicago Department of Geography Research Paper, no. 170 (Chicago, 1975); David Klingaman, “The Significance of Grain in the Development of the Tobacco Colonies,” Journal of Economic History, XXIX (1969), 268–278; and the essay by Lois Green Carr and Lorena Walsh in this volume. 33  Gray, History of Agriculture, I, 554–555; John T. Schlotterbeck, “The ‘Social Economy’ of an Upper South Community: Orange and Greene Counties, Virginia, 1815–1860,” in Orville Vernon Burton and Robert C. McMath, Jr., eds., Class, Conflict, and Consensus: Antebellum Southern Community Studies (Westport, Conn., 1982), 11–12; Sarah S. Hughes, “Slaves for

1274

Morgan

Despite the tendencies toward individual work, however, the small gang was still the most common work unit on diversified plantations of any size. In the early nineteenth century, James Madison noted that “slaves [were] seldom employed in regular task work” in his part of Virginia. During the pea harvest, Landon Carter placed two overseers at each end of a row so that his gang’s productivity could be closely monitored. The wheat harvest, which became the central event of the agricultural year in many parts of the Chesapeake by the late eighteenth century, led planters to investigate closely how to make their harvesting gang—this “whole machine,” as Thomas Jefferson termed it—move more expeditiously. George Washington, for instance, hoped to produce “more ease, regularity and dispatch” among his harvesters by a careful allocation of roles and division into gangs. For every two cradlers whom Washington and Jefferson employed, seven and five gatherers, respectively, were allocated. This gang system was indeed closely monitored.34 Chesapeake planters generally organized their children into “small gangs,” having them perform simple tasks like shelling corn and sowing grass seed, activities not too far removed from the experiences of Caribbean “grass gangs.” Such arrangements resemble the small gangs assembled in such diverse work situations as mining, quarrying, or shipbuilding throughout the Western world. All shared at least one feature: teamwork was valued, but the scattered nature of the work made external monitoring of effort difficult and expensive.35 A second intermediate stage between tasking and ganging was a combination of both systems—an arrangement best termed collective tasking. The jobbing or task gangs that were hired out to Caribbean planters represent one variant of this form. Henry Bolingbroke provides an admirable definition: “A Task Gang … is so called from its undertaking to do a specific quantity of work, such as clearing and preparing so many acres of land, draining and planting the same; which they are paid for by the acre.” Such groups were the hardestworked of any West Indian slaves. According to Benjamin M’Mahon, jobbing Hire: The Allocation of Black Labor in Elizabeth City County, Virginia, 1782 to 1810,” WMQ, 3d Ser., XXXV (1978), 260–286. 34  “James Madison’s Attitude toward the Negro,” Journal of Negro History, VI (1921), 89; Greene, ed., Diary of Carter, I, 497; Edwin M. Betts, ed., Thomas Jefferson’s Garden Book, 1766–1824, with Relevant Extracts from His Other Writings (Philadelphia, 1944), 228, 230; Donald Jackson and Dorothy Twohig, eds., The Diaries of George Washington, 6 vols. (Charlottesville, Va., 1976–1979), II, 172, V, 9–10. 35   Schlotterbeck, “Social Economy,” in Burton and McMath, eds., Class, Consensus, and Conflict, 11. For references, see Ralph Schlomowitz, “Melanesian Labor and the Development of the Queensland Sugar Industry, 1863–1906,” Research in Economic History, VII (1982), 335.

Task And Gang Systems

1275

gangs “were tasked to perform each a certain amount of work, without any deduction for weather; the consequence was that they had to be at it from day-light till dark, without even stopping for meals.”36 Certainly, jobbing gangs might wreak some form of revenge for their harsh treatment. In 1811 Barbadian planters complained that their hired gangs employed by the task often dug shallow cane holes. But the planters could usually counter the task gangs’ stratagems. In this case, they simply ordered the hired gangs to begin work earlier than usual.37 Collective tasking was not confined to hired gangs. When a gang lagged behind in its work, an owner might task its members to make up for lost time. An upland cotton planter from South Carolina declared that he did “not believe in tasking Negroes all the time, but I believe in tasking them when it is necessary, in case of a push, when the crop is suffering.” A Jamaican planter, who was forced to resort to this strategy, observed that the slaves “became very much dissatisfied, at being put off their ordinary routine of working.” Jonathan Steele’s experiments in late-eighteenth-century Barbados involved collective tasking. Groups of slaves were given financial inducements to hole, weed, and hoe an acre of canes per day.38 One of the most progressive of early nineteenth-century treatises on plantation husbandry—by Dr. Collins—advocated tasking, but only in holing (the preserve of the task or jobbing gangs). As James Stephens later pointed out, “Dr. Collins had in view, when he recommended taskwork, the assessment of a daily portion of the work required, not on each individual slave, but on the collective gang.”39 But, if individual tasking was not generally countenanced on Jamaican sugar plantations or those of other long-established Anglo-American sugar societies, 36  Henry Bolingbroke, A Voyage to Demerary, Account of the Settlements There, and of Those on the Essequibo and Berbice, 1799–1806 (London, 1807), 215; Benjamin M’Mahon, Jamaica Plantership (London, 1839), 86. 37   Minute Book of the Barbados Agricultural Society, 30, as cited in Higman, Slave Populations, 164. 38   Farmer and Planter, V, 229, as cited in Gray, History of Agriculture, I, 552; Marly, 153; Dickson, Mitigation of Slavery, 119–120, 123–133. See also Slavery in British Guiana in Negro Slavery (London, 1823–1826), 2–4. Task gangs were introduced into royal shipyards in Britain and were opposed by the workers for much the same reasons that jobbing gangs were hated by West Indian slaves. See James M. Haas, “The Introduction of Task Work into the Royal Dockyards, 1775,” Journal of British Studies, VIII, no. 2 (May 1969), 44–68; and Essay on Task-Work, 8. 39  Dr. Collins, Practical Rules, 152–154; Stephen, Slavery of British West India, I, 57, II, 232–235. The debate between Collins and Stephen is crucial to an understanding of what tasking usually meant in the context of Caribbean sugar production.

1276

Morgan

it was surprisingly widespread in other areas and on other kinds of estates. Not only was the task system associated with a whole range of staple crops and secondary plantation activities, it also became more widely adopted over time. By the early nineteenth century, it was even invading the sugar regimes of newly acquired Anglo-American slave societies. By the third decade of the nineteenth century, perhaps a majority of slaves in the Anglo-American world were working by task rather than in gangs. How do we account for the original association of task and gang systems with particular crops, and how do we explain the tendency for gang systems to evolve into task systems? Certainly, staple-crop requirements explain much about the original choice of labor arrangement. In particular, the degree to which direct supervision (or, to put it more technically, external monitoring of input) was required by various crop cycles or by pivotal operations within crop cycles lies at the heart of the association between crops and labor systems. Some staples required close surveillance, either because their production schedule involved a careful synchronization of operations (as with sugar) or because they were delicate plants that were difficult to grow successfully (as with tobacco).40 Wherever supervision was at a premium, gang systems seem to have arisen. Such associations were not confined only to the New World. In early seventeenth-century England, the commercial growth of special crops like woad gave rise to daily gang labor during the weeding and picking parts of the cycle. Both the laborers’ unfamiliarity with the crop and the need for careful timing in the cycle seem to have accounted for this arrangement. In the case of sugar cultivation, where the two most critical features were the careful timing of operations and the assumed benefits of teamwork, gangs were large.41 40  For the careful synchronization in sugar production, see Ward Barrett, “Caribbean SugarProduction Standards in the Seventeenth and Eighteenth Centuries,” in John Parker, ed., Merchants and Scholars: Essays in the History of Exploration and Trade (Minneapolis, Minn., 1965), 145–170; Richard Pares, Merchants and Planters, Economic History Review Supplement no. 4 (Cambridge, 1960), 21; Fernando Ortiz, Cuban Counterpoint: Tobacco and Sugar, trans. Harriet de Onís (New York, 1947), 6, 26–27. For one of many descriptions of tobacco, see Joseph M. Hernandez, “On the Cultivation of the Cuban Tobacco Plant,” Southern Agriculturalist, III (1830), 463. 41  Joan Thirsk, Economic Policy and Projects: The Development of a Consumer Society in Early Modern England (Oxford, 1978), 4–5. The best information on this subject comes from the example of sugar production in the postslavery era: teamwork was still seen as valuable, particularly in cane cutting, in spite of the negative connotations associated with gang work. See Jerome Handler, “Some Aspects of Work Organization on Sugar Plantations in Barbados,” Ethnology, IV (1965), 16–38; Ralph Schlomowitz, “Team Work and Incentives:

Task And Gang Systems

1277

Tobacco cultivation, on the other hand, laid the most emphasis on the careful husbandry of a delicate plant. Only a tobacco planter could claim, as one Virginian did, that his crop had been “under my own eye, and [I] may say I saw almost every plant from the planting to the prizing.” Such detailed attention was only practicable where gangs were, as one contemporary put it, “generally made into small parcells, not above eight or ten hands at a place.”42 Conversely, where a crop was relatively hardy (as with rice) or where there was no particular urgency in the production schedule (as with tending pimento or drawing off turpentine), a more relaxed attitude could be taken to the external monitoring of input, and individual tasking arrangements usually evolved.43 Here again, such associations were not confined to the New World. The great expansion of rice culture in seventeenth-century Lombardy, for instance, was predicated, not on a stable, sophisticated, and well-supervised labor force, but on a pool of transient labor drawn from far afield. Even New World crops produced largely by means of task labor can point up the association between gang labor and intensive supervision. Lowcountry rice was cultivated and processed by an elaborate schedule of tasks, but special work, like the combating of floodwaters, had to be closely supervised and seems to have been accomplished by organizing slaves into gangs. A Jamaican bookkeeper reported that the only work on a coffee plantation not carried out by tasks was The Origins and Development of the Butty Gang System in Queenland’s Sugar Industry, 1891–1913,” Journal of Comparative Economics, III (1979), 41–55; and Schlomowitz, “Melanesian Labor,” Res. Econ. Hist., VII (1982), 327–361. 42  John Custis to Robert Cary, [1729], John Custis Letterbook, typescript, Research Library, Colonial Williamsburg Foundation; Sir Dalby Thomas, An Historical Account of the Rise and Growth of the West-India Collonies…(London, 1690), in Jerome E. Brooks, comp., Tobacco: Its History Illustrated by the Books, Manuscripts, and Engravings in the Library of George Arents, Jr., 5 vols. (New York, 1937–1952), II, 531. One June day in 1770, Landon Carter visited one of his gangs three times and, on each occasion, spent two hours monitoring their progress in turning tobacco hills (Greene, ed., Diary of Carter, I, 422). More generally, see T. H. Breen, “The Culture of Agriculture: The Symbolic World of the Tidewater Planter, 1760–1790,” in David D. Hall, John M. Murrin, and Thad W. Tate, eds., Saints and Revolutionaries: Essays on Early American History (New York, 1984), 247–284. 43  Arrowroot, like rice, was a hardy plant, which might explain its susceptibility to tasking (Handler, “History of Arrowroot,” Jour. Carib. Hist., II [1971], 50). Other staples that, like pimento and naval stores, were produced without hard driving or regimentation include coffee and cocoa: see Augustin Cochin, The Results of Emancipation, trans. Mary L. Booth (Boston, 1863), 200–201; and Benjamin Moseley, A Treatise concerning the Properties and Effects of Coffee (London, 1792), v.

1278

Morgan

the drying of the berries, because, in this operation, “constant attention was proper.”44 Practicability and the costs of supervision, likewise, were vital considerations in the choice between task and gang systems. In the case of rice, naval stores, and, to a lesser extent, coffee and cocoa, a slave could tend a considerable acreage. As a result, the labor force was inevitably scattered, and close supervision of its input would have proved laborious and costly. An eighteenth-century lowcountry slave, for example, was expected to cultivate between three and four acres of rice; by the early nineteenth century, five acres were the norm.45 A thousand cocoa trees formed the standard employment for an adult West Indian slave, whereas, as we have already noted, three to four times this number of trees constituted the turpentine worker’s district. Even in sugar areas, there was some work, such as grass picking, that could, in James Stephen’s words, “only be done by the slaves in a state of dispersion”; recourse was, therefore, “to the mode of individual task-work.”46 Other staples, however, brought workers much closer together and made direct supervision more feasible. The density of laborers engaged in Caribbean sugar cultivation was unparalleled among New World staples. A general rule of thumb dictated one slave for every acre of cane land, although a ratio of three slaves to every two acres was not unusual. One to two acres per slave constituted the norm for tobacco cultivation.47 The two crops that gave rise to both tasking and ganging in different parts of the New World—cotton and coffee—were characterized by land and labor ratios somewhere in between that of rice, on the one hand, and sugar, on the other.48 In addition to the degree of direct supervision required by a staple and the sheer practicability of monitoring input, the facility with which the laborers’ 44  Domenico Sella, Crisis and Continuity: The Economy of Spanish Lombardy in the Seventeenth Century (Cambridge, Mass., 1979), 121–122; Phillips, American Negro Slavery, 259; Kelly, Voyage to Jamaica, 19. 45  These ratios are based on the observations of 25 different 18th-century commentators and about 20 plantation-sale advertisements in South Carolina newspapers, where the number of field hands and amount of cleared rice land are given. See also James M. Clifton, “The Rice Industry in Colonial America,” Agricultural History, LV (1981), 275. 46  Cochin, Results of Emancipation, trans. Booth, 200–201 (see n. 10); Stephen, Crisis of Sugar Colonies, 12. 47  Goveia, Slave Society, 121; Dunn, Sugar and Slaves, 198. The tobacco acreage per hand ratio is based on observations of 10 different 18th- and early 19th- century Chesapeake commentators. 48  Alexander, Transatlantic Sketches, I, 24; George Pinckard, Notes on the West Indies …, 3 vols. (London, 1806), III, 404.

Task And Gang Systems

1279

output could be measured also shaped different forms of labor organization. For example, the productivity of a single pimento worker could be measured accurately and cheaply, particularly in the harvesting cycle. In Jamaica, Henry Whiteley observed a head bookkeeper examining the baskets of pimento collected by young slaves in order to “ascertain that each slave had duly performed the task allotted” (my emphasis). An anonymous Jamaican noted that the task in pimento harvesting was to “pick half a barrel of the pimento berries daily.”49 It was easy, in other words, to weigh an individual’s baskets of pimento berries, and tasking may have developed in this stage of the crop cycle before being extended to other operations. In the case of rice, it was less the harvesting and more the planting of the crop that lent itself to inexpensive and efficient measurement. The ubiquity and long-standing history of the quarter-acre task suggest that the planting and weeding stages of the rice cycle provided the initial rationale for the task system; once tasking became firmly established, it was extended to a whole host of plantation operations. Before long, everything in rice cultivation was measured by neatly demarcated, universal standards, and an individual could largely be left to make his share of the crop in his own time with no detriment to the final product.50 Monitoring an individual’s output was much more problematic in sugar and tobacco production. The large volumes involved in the sugarcane harvest made it impracticable to measure output by the individual task. Moreover, cane, once cut, had to be hurried to the grinders. With such a premium on haste, teamwork was believed to yield a joint output sufficiently larger than the sum of the outputs of laborers working independently.51 In the case of tobacco, an individual laborer did not produce large volumes; seemingly, there was no apparent prohibition against measuring an individual’s product by the task. However, the quality of any tobacco crop was so variable that there could not be, in Fernando Ortiz’s words, “an exact unit of measurement for leaf tobacco.”52 One might count hogsheads, it is true, but the quality of the 49  Henry Whiteley, Three Months in Jamaica, in 1832: Comprising a Residence of Seven Weeks on a Sugar Plantation (London, 1833); William Naish, The Advantages of Free Labour over the Labour of Slaves, Elucidated in the Cultivation of Pimento, Ginger, and Sugar (London, [ca. 1825]), 2. Barry Higman first suggested the importance of this factor in a personal communication to me. 50  Morgan, “Work and Culture,” WMQ, 3d Ser., XXXIX (1982), 566. 51  See nn. 15 and 40 above for descriptions of the sugar cycle that stress these points. Bahian sugar plantations seem to be an exception. Slaves were assigned a daily task in cane cutting. However, teamwork was still primary. See Schwartz, Sugar Plantations, 140–141. 52  Ortiz, Cuban Counterpoint, trans. Onís, 36.

1280

Morgan

leaf, just as much as the number of hogsheads, determined the crop’s worth. Moreover, tasking a single tobacco laborer made little sense when so much depended on the collective care expended on any one crop. For largely different reasons, then, sugar and tobacco production lent themselves to the measurement of collective, rather than individual, output. Cotton and coffee, once again, were in something of an intermediate position, primarily because these two crops combined qualities that encouraged both tasking and ganging. On the one hand, both crops needed continual care in the planting and weeding parts of their cycles and, thus, proved excellent candidates for a gang system. On the other hand, harvesting of both crops was most efficiently carried out by the task. Thus, in Brazil, where most operations on coffee estates were reserved for slave gangs, the laborers were tasked during the harvest.53 Similarly, in the upland cotton South, where again the gang system was widespread, the crop was often picked by task. As one upcountry Georgia planter explained: “The usual custom of planters is to work without tasks during the cultivation of their crop; but in gathering cotton tasks are common, and experience has proven that whenever work is of that kind of character that it can be properly parceled out into tasks, it is much better to do so…[because] the overseer … will get more work, and the negro will be better satisfied…[by] gain[ing] time to devote to his own jobs or pleasure.”54 A social, rather than an economic, consideration may also have contributed to the adoption of the gang system, in particular. If social control was a pressing imperative, it made sense to gang. As Henry Drax observed as early as the seventeenth century, “The best Way that I know of to prevent Idleness, and to make the Negroes do their Work properly, will be upon the change of Work, constantly to Gang all the Negroes in the Plantations in the Time of Planting.” Or, as William Dickson put it, writing from the vantage point of the more civilized nineteenth century, ganging was a “vulgar system (which, perhaps, was the only one that was practicable 150 years ago, with an untamed set of savages).”55 On the other hand, if social control was of overriding, rather 53  Stein, Vassouras, 36. 54  Robert Collins, “Essay on the Treatment and Management of Slaves,” Southern Cultivator, XII (1854), 205–206. However, Stanley Engerman has suggested to me that the cotton harvest was not true tasking in the sense of a measure of labor input. Each worker’s output was measured, it is true, but it was work done in a group, supervised closely and done in unison, with no apparent difference in rewards. This seems to me further confirmation that cotton was an intermediate crop, veering more toward a gang, than task, system. 55   Henry Drax, “Instructions for the Management of Drax-Hall and the Irish-Hope Plantation …,” in William Belgrove, A Treatise upon Husbandry or Planting (Boston, 1755), 64; Dickson, Mitigation of Slavery, 121.

Task And Gang Systems

1281

than just of contributory, importance, it seems surprising that the rice regime, where plantations were large and social control a concern, did not employ a gang system. While a variety of staple-crop requirements, rather than concerns about social control, help explain the adoption of one labor arrangement over another, those choices were certainly not fixed or impervious to changing circumstances. The most obvious example is the transformation that began to take place on many sugar estates from gang to task labor toward the end of the slave era. The origins of this transformation can be traced, in part, to progressive metropolitan opinion opposed to slavery. In 1789, for instance, Benjamin Frossard, an antislavery French Protestant, urged planters to “fix up each slave with a task per week, and to allow him to use the time he earns in his own business, or to work on a small uncultivated plot that his master grants him, or to complete another task for a reasonable wage.” It was incontestable, argued Frossard, that a slave “would work the whole week with unimaginable zeal, if he was sure that, after finishing on Friday, Saturday belonged to him.”56 According to the author of An Essay on Task-Work, “Mr. Wilberforce observed some years ago in the House of Commons, that the planters in the West-Indies ought to work their negroes in tasks, as the planters did in the southern states of America.” This author’s main argument in support of Wilberforce was that “the negroes, having an expectation of being able to finish their work in good time in the afternoon, which would be a real premium to them, would exert themselves with spirit and alacrity, and work very differently from the drawling manner they generally do, when they have no expectation of getting away while there is any day-light remaining.” One author explicitly proposed “taskwork” as the “means of paving the way for the introduction of voluntary labour on the part of the negroes.”57 The other source of change came from below—from the slaves themselves. Naturally, this impetus is less easily documented. However, a study by Mary Turner, which explores a rich set of estate papers for the parish of Saint Thomas-in-the-East, one of the most important sugar-producing areas of Jamaica, is highly suggestive of broader processes. She finds evidence, which 56  Benjamin Sigismond Frossard, La cause des esclaves nègres et des habitans de la Guinée … (Lyon, 1789), II, 269. 57   Essay on Task-Work, ii, 6; review [“Condition of the Negroes in Our Colonies”], Quarterly Review, XXIX (1823), 500. See also Plan for the Safe and Profitable Conversion of the Colonial Slaves into Free Labourers (London, [1833?]); and Reverend George F. W. Mortimer, The Immediate Abolition of Slavery, Compatible with the Safety and Prosperity of the Colonies … (Newcastle-upon-Tyne, 1833).

1282

Morgan

can be replicated for other plantations, of a constant struggle between slaves and their managers over the labor process. Through a perpetual testing of wills, a complex process of negotiation and renegotiation, work norms were eventually hammered out which, in turn, created opportunities for the slaves to demand taskwork in the form of food or cash payments for work beyond the agreed-upon norms. According to her account, “No later than the 1820’s the slaves were pressing for task work in sugar production.”58 Perhaps the battle can be traced to earlier times, but the broader significance of this development is not in doubt. That slaves were important in the shift to taskwork is most graphically displayed in the immediate responses of recently freed bondmen. Operations that previously had been unthinkable without gang labor swiftly became the preserve of task laborers. Much of the responsibility for effecting this transformation lay with the workers themselves. The importance of their preferences can be detected both within and between individual Caribbean islands. Within certain islands—Antigua and Saint Kitts, for instance—those laborers, who lived in independent villages and who therefore enjoyed more room to bargain over their terms of employment than resident estate workers, were in the vanguard of those seeking task or job work over daywork.59 Whole islands, not just a few villages, experienced the immediate postslavery years in widely contrasting ways. A society like Montserrat—which contained particularly independent laborers, deriving their autonomy from the use of extensive provision lands allotted to them as slaves and confirmed to them as freedmen—witnessed a more rapid turn from daywork to taskwork than any of her sister Leeward Islands. Similarly, in a society like British Guiana or Trinidad (where the land-labor ratio gave the freedmen greater comparative advantages than, say, freedmen in Barbados), task gangs, less than a fortnight after the proclamation of freedom, everywhere began combining for higher wages or for working only half a day. As Donald Wood has pointed out: “To an observer from Barbados, where daily work was the custom, operations in a Trinidad cane field looked ragged and ill-disciplined; the labourers started their task where they pleased and worked away on their own. In Barbados the 58  Mary Turner, “Chattel Slaves into Wage Slaves: A Jamaican Case Study,” in Malcolm Cross and Gad Heuman, eds., Labour in the Caribbean: From Emancipation to Independence, Warwick Caribbean Series (London, forthcoming). We very much need similar work from other plantations. I attempt to explore the struggle over the labor process in my forthcoming book, Slave Counterpoint: Black Culture in the Eighteenth-Century Chesapeake and Lowcountry. 59  Hall, Five of the Leewards, 44–45.

Task And Gang Systems

1283

labourers worked in line across the field.”60 In 1841 William White-house unfavorably contrasted the rapid emergence and uniform system of tasking in British Guiana with the situation he found in Jamaica.61 Wherever the freedmen held a modicum of bargaining power, taskwork prevailed. The slaves’ and ex-slaves’ preference for tasking does not mean that its introduction proved an unqualified benefit to them. In Demerara, for instance, many planters in the 1820s inflated tasks beyond a normal day’s labor to ensure that the slaves would have no free time. As Alan Adamson has pointed out for postemancipation British Guiana: “The definition of conventional field tasks was prepared by a committee of planters immediately following the abolition of slavery and was still used in the 1850s and 1860s. It is difficult to believe that this committee, nurtured in the climate of slavery, would not have set levels of performance at a relatively high level.” In fact, the standard daily task for digging a trench in mid-nineteenth-century British Guiana would have taken the best English laborer between two and three days to complete.62 Moreover, as Walter Rodney has argued, Guyanese “workers of all categories were subjected to increased rates of exploitation by the simple device of increasing the size of the tasks.” Tasking proved most iniquitous for newly arrived Asian labor. According to Hugh Tinker, a task might require a working day of fifteen hours or more for “weaker or less experienced coolies.” He quotes Governor Wodehouse of British Guiana, an early critic of the application of tasking to Indians, as saying that task work “is payment … of work done without any necessary reference to the time expended…. What is this … but a device for obtaining a certain amount of work without any trouble of superintendence?”63 60  Ibid., 50–51; Alan H. Adamson, Sugar without Slaves: The Political Economy of British Guiana, 1838–1904 (New Haven, Conn., 1972), 34; Donald Wood, Trinidad in Transition: The Years after Slavery (London, 1968), 52. See also Claude Levy, Emancipation, Sugar, and Federalism: Barbados and the West Indies, 1833–1876 (Gainesville, Fla., 1980), 78; W. Emanuel Riviere, “Labour Shortage in the British West Indies after Emancipation,” Jour. Carib. Hist., IV (1972), 17. 61  W. F. Whitehouse, Agricola’s Letters and Essays on Sugar Farming in Jamaica (London, 1845), 17. By contrast, it is interesting that some upland cotton planters in the southern United States employed a modified gang arrangement, known as the squad system, in the immediate postwar years. See Ralph Schlomowitz, “The Squad System on Postbellum Cotton Plantations,” in Orville Vernon Burton and Robert C. McMath, eds., Toward a New South? Studies in Post—Civil War Southern Communities (Westport, Conn., 1982), 265–280. 62  Higman, Slave Populations, 179; Adamson, Sugar without Slaves, 111–112. 63  Walter Rodney, A History of the Guyanese Working People, 1881–1905 (Baltimore, 1981), 58; Hugh Tinker, A New System of Slavery: The Export of Indian Labour Overseas, 1830–1920 (London, 1974), 183. The continuance of gangs and forms of collective tasking into the

1284

Morgan

Still, despite the potential for abuse, absence of supervision was only one of the benefits tasking offered freedmen. Closely related was the opportunity it gave laborers to gain time for their own affairs. More mundanely, an industrious laborer could earn more by the job or task than by daywork. These advantages were particularly valued by a people whose “natural love of independence” drew contemporary comment. This was as true of freedmen on the Eastern Shore of Virginia in the middle of the seventeenth century as of those living in Jamaica two centuries later. Thus, “Anthony the Negro,” a resident of Northampton County, Virginia, was reported to have said, after dividing a cornfield in which he and a white farmer had an interest, “I am very glad of [the division, for] now I know myne owne, hee [the white farmer] finds fault with mee that I doe not worke, but now I know myne owne ground and I will worke when I please and play when I please.” According to a visitor to Jamaica in 1840, the freedman’s overriding priority was not much different. It was to avoid being “tied down to give so many days’ labor continuously and continually,” and it was a desire “to be at liberty to do anything else that he might want for himself.”64 The advantages of tasking, as perceived by freedmen in the Caribbean sugar colonies and elsewhere, bring me, finally, to a consideration of some of the implications of the task and gang systems under slavery. Provided that there postemancipation era also proved onerous to the freedmen. See John Davy, The West Indies, before and since Slave Emancipation, Comprising the Windward and Leeward Islands’ Military Command (London, 1854), 134; Adamson, Sugar without Slaves, 154; Great Britain, House of Commons, Reports from the Select Committee on Sugar and Coffee Planting (1848), III, 3–4. Tasking scales drawn up at emancipation by planters reveal a mixture of individual and collective tasks: see, for example, Marshall, Colthurst Journal, 246–254; and Craton and Walvin, Jamaican Plantation 328–331. 64  T. H. Breen and Stephen Innes, “Myne Owne Ground”: Race and Freedom on Virginia’s Eastern Shore, 1640–1676 (New York, 1980), 6; Great Britain, House of Commons, Report from the Select Committee on the West India Colonies (1842), 6845, as cited in Douglas Hall, “The Flight from the Estates Reconsidered: The British West Indies, 1838–42,” Jour. Carib. Hist., X–XI (1978), 19. For other sources which point to the advantages of tasking, see Davy, West Indies, 91, 121, 313, 362, 501; Adamson, Sugar without Slaves, 165; Douglas Hall, Free Jamaica, 1838–1863: An Economic History (New Haven, Conn., 1959), 44–45; Great Britain, House of Commons, Reports from the Select Committee on Sugar and Coffee Planting (1848), I, 15, 215–216, II, 17, III, 3–4, 60, 75, 1.02, 162, 185, 255, 347, 408; James McNeill and Chimman Lal, Report to the Government of India on the Conditions of Indian Immigrants in Four British Colonies and Surinam (1915), Cd. 7744, 16–17, 65, Cd. 7745, 206. For an important general survey of the ex-slaves’ responses and adjustments to freedom, see Stanley L.  Engerman, “Economic Adjustments to Emancipation in the United States and British West Indies,” Journal of Interdisciplinary History, XIII (1982–1983),: 196–220.

Task And Gang Systems

1285

were equal commercial opportunities (in terms of available land, the marketing structure, and the like) from one slave society to another, it seems logical to suppose that taskworkers would be better able to exploit these opportunities than gang workers. However onerous tasking could become for some slaves, the system at least allowed them a certain latitude to apportion their day, to work intensively in their task, and then have the balance of their time. Perhaps the best, backhanded testament to this advantage comes from Thomas Cooper, who resided in Jamaica from 1817 to 1820. The reason taskwork was so “very uncommon” on that island, he maintained, was that it was “held to be dangerous to allow the slave much spare time.”65 The advantage of having spare time was readily recalled by exslaves in the lowcountry. One freedman, for instance, reported that “a good active industrious man would finish his task sometimes at 12, sometimes at 1 and 2 o’clock and the rest of the time was his own to use as he pleased.” Lowcountry masters were careful not to encroach upon their slaves’ “own time” unless by offering payment. According to a Jamaican, South Carolina and Georgia planters were “very particular in never employing a negro, without his consent, after his task is finished, and agreeing with him for the payment which he is to receive.”66 Many lowcountry slaves put these opportunities to good effect. As one ex-slave explained, “They all worked by tasks, and had a plenty of time to work for themselves and in that way all slaves who were industrious could get around them considerable property in a short-time.” This is exactly what happened in the eighteenth- and nineteenth-century lowcountry. By the end of slavery, there was a significant internal slave economy. Some slaves were raising stock, many slave families were growing their own crops on four to five acres apiece, and slave property was being not only produced and exchanged but also inherited.67 If the task system in the lowcountry represented one pole in terms of the slaves’ domestic economy, the gang system practiced in an area like the Chesapeake represented another. Chesapeake slaves worked, in William Tatham’s words, “from daylight until the dusk of the evening, and some part of the night, by moon or candlelight, during the winter.”68 As a result, Chesapeake 65  Thomas Cooper, Facts Illustrative of the Condition of the Negro Slaves in Jamaica (London, 1824), 27. 66  Morgan, “Work and Culture,” WMQ, 3d Ser., XXXIX (1982), 586; Essay on Task-Work, 9. 67  Morgan, “Work and Culture,” WMQ, 3d Ser., XXXIX (1982), 590; and “The Ownership of Property by Slaves in the Mid-Nineteenth-Century Low Country,” Journal of Southern History, XLIX (1983), 399–420. 68  G.  Melvin Herndon, William Tatham and the Culture of Tobacco (Miami, Fla., 1969), 102. The same would apply to the upland cotton South. Interestingly, one traveler who went

1286

Morgan

slaves generally had no time of their own during the week, apart from a rare holiday, and were, therefore, only able to work on their own account on Sundays. Moreover, as Chesapeake planters accorded a much higher priority to foodstuffs production than did their lowcountry counterparts, Chesapeake slaves were much less encouraged to engage in part-time, private subsistence activities. Field hands in eighteenth-century tidewater Virginia produced twice or even three times as much corn per head as did those in the eighteenth-century lowcountry.69 These differences, in both available time and white encouragement were, in turn, translated into sharply contrasting domestic economies. The domestic plots cultivated by lowcountry slaves merited the term “little plantations,” or “private fields,” whereas, in the Chesapeake, they were never more than garden plots. This is not to say that Chesapeake slaves were reluctant, domestic cultivators. William Tatham noted in 1800 that there were crops which were “permitted (and greatly confirmed by custom) to slaves,” namely, “potatoes, garden-stuff, pumpkins, melons, a few particular fruit-trees, peas, hops, flax, and cotton.”70 However, the very diversity of the crops grown by Chesapeake slaves is revealing of their gardening orientation, a reflection of the attenuated time and land allowed them. Furthermore, those account books of nineteenthcentury Chesapeake planters that recorded purchases from their slaves reveal only minor subsistence activities by the bondmen. Finally, although slaves in

from lowcountry South Carolina (where he noted the shorter workday associated with the task system) to Mississippi remarked upon the difference. In 1820, while in the Deep South, he observed that the slaves “here …, seem to work many hours longer than in Carolina” (Adam Hodgson, Remarks during a Journey through North America in the Years 1819, 1820, and 1821 … [New York, 1823], 174). 69  This is explored at greater length in my Slave Counterpoint. 70  For the lowcountry, see “A Curious New Description,” Universal Museum, I (1762), 477; and Adele Stanton Edwards, ed., Journals of the Privy Council, 1783–1789, State Records of South Carolina (Columbia, S.C., 1971), 132. Mere garden plots continued to be reserved to Chesapeake blacks when free: typical labor contracts in the immediate postemancipation era granted freedmen the use of “garden spots” or truck patches. See, for example, Report on the Condition of Bureau Affairs in the Sub-District of Amelia and Powhatan Counties of Virginia, by Lieut. J. B. Clinton, Jan. 31, 1867 (A7829), Freedmen and Southern Society, files of documents in the National Archives, University of Maryland, College Park (hereafter references to documents read at the society will be given in parentheses). William Tatham, Communications concerning the Agriculture and Commerce of the United States of America … (London, 1800), 55.

Task And Gang Systems

1287

the Chesapeake gained a deserved reputation as the “Chicken Merchants” of the region, this virtually represented the extent of their stock-raising.71 On some of the smaller, flat or gently sloping Caribbean islands (such as Barbados or Antigua), the situation facing slaves was not too far removed from that encountered by Chesapeake bondmen. Working sunup to sundown, six days a week, under the gang system formed the pervasive reality for most field hands on these islands. Moreover, the planters’ agricultural priorities—primarily a function of high prices for sugar and restricted land supplies—demanded that slaves confine their subsistence activities to garden patches. Certainly, Barbadian slaves were renowned for the intensive cultivation of their house plots. In 1824 J. W. Jordan observed that many slaves on one Barbadian estate “reap annually 1000 lbs of yambs each, besides Guinea and Indian corn, eddoes, potatoes, cassava, ginger, and the more industrious plant arrowroot, which they prepare and sell for a good price.” Eight years later, Henry Coleridge noticed that “yams, Indian corn, plantains, or even canes, are to be seen growing round every hut.” But extensive provision grounds were largely denied these slaves.72

71  Account Book, ca. 1845, Bruce of Berry Hill Plantation Volumes, box 5; Plantation Record Book, 1848–1853, and Farm Register, 1850s, Richard Irby Papers; John Tucker Ledger, 1833– 1835, Tucker Family Account Books; John Washington Account Book, 1848, Washington Family Papers of Caroline County; Manuscript Volume of William B. Irby, 1862, Neblett and Irby Family Papers, all in Alderman Library, University of Virginia, Charlottesville. James Mercer to B. Muse, Jan. 6, 1782, Battaille Muse Papers, Duke University, Durham, N.C. Labor contracts in the postemancipation Chesapeake, unlike those in the lowcountry, often delimited the stock-raising activities of the freedmen. It is highly likely, as with much else in these contracts, that this was meant to continue the practices of slavery, therefore telling us as much about life under slavery as under freedom. A typical agreement specified that the freedmen “can raise and keep chickens and no other fowls—six to eight hens to each man and when they have a family may raise a hog but no sows or pigs or other kind of stock” (Agreement between E. W. Hubbard, Agent for Miss E. W. Eppes, and 8 Freedmen, Dec. 2, 1865 [A8140]). 72  Joseph William Jordan, An Account of the Management of Certain Estates in the Island of Barbados (London, 1826), 4; Henry Nelson Coleridge, Six Months in the West Indies, in 1825 (London, 1832), 125–126. See also Handler, “History of Arrowroot,” Jour. Carib. Hist., XI (1971), 79; Sidney W. Mintz and Douglas Hall, The Origins of the Jamaican Internal Marketing System, Yale University Publications in Anthropology no. 57 (New Haven, Conn., 1960), 4, 10; Marshall, Colthurst Journal, 140; Dickson, Letters on Slavery, 11; Davy, West Indies, 16; Stephen, Slavery of British West India, II, 261; Pinckard, Notes on the West Indies, I, 368; Thomas Rolph, A Brief Account Together with Observations Made during a Visit in the West Indies … Dundas, 1836), 52.

1288

Morgan

A similar reality faced Antiguan slaves. In 1788, when John Luffman visited the island, he observed that adult slaves were allocated patches of ground no more than “twenty five to thirty feet square.” A few years later, another visitor pointed out the disadvantages that flowed from a system of subsistence dependent on planter allocations as opposed to slave provision grounds. The quantity and quality of the food supply were inferior, he noted, but also “the negro suffers too in his poultry and little stock, which are his wealth…[and] a negro without stock, and means to purchase tobacco and other little conveniences, and some finery too for his wife, is miserable.”73 The notion of slaves as gardeners rather than as protopeasants applies in large measure to the dependent populations of both the smaller Caribbean islands and areas of North America like the Chesapeake. What clearly demarcated all Caribbean slaves (even those of the smaller islands) from their North American counterparts (the Chesapeake more so than the lowcountry) were radically different marketing opportunities. Even if some Caribbean slaves were more gardeners than large-scale cultivators, great numbers of them were soon producing a surplus that they could sell to their masters, to other slaves, or to other freemen. Once these surpluses were regularly taken to local markets and exchanged for other commodities, or sold for cash, market day became an important social and economic institution. As early as the 1740s, the manager of the Codrington plantations on Barbados remarked that nothing could keep his slaves from the markets short of “locking them up.” By the early nineteenth century, George Pinckard declared that Barbadian markets “depend almost wholly” on the slaves’ private endeavors. J. B. Moreton, when writing of Jamaica in the late eighteenth century, observed that “Sunday is the greatest market day; the negroes from all parts of the country flock to town, hundreds of them in a gang, carrying with them the product of their grounds.” Similarly, John Luffman noted that the small patches allotted Antiguan slaves proved “to be of material benefit to the country, their produce principally supplying the ‘Sunday market’ … with vegetables.” By this means, Antiguan whites were “prevented from starving.”74

73  John Luffman, A Brief Account of the Island of Antigua … (London, 1788), 94–95; Sir William Young, A Tour through the Several Islands of Barbadoes, St. Vincent, Antigua, Tobago, and Grenada, in the Years 1791 and 1792, in Bryan Edwards, History of the British Colonies … (Philadelphia, 1806), IV, 266. See also Goveia, Slave Society, 135–139; and Hall, Five of the Leewards, 29–30. 74  Bennett, Bondsmen and Bishops, 24; Pinckard, Notes on the West Indies, I, 369–370; J. B. Moreton, Manners and Customs, 36; Luffman, Brief Account, 94–95.

Task And Gang Systems

1289

In the larger, mountainous Caribbean islands (most notably, Jamaica and Saint Domingue), the slaves’ access to extensive provision grounds as well as garden plots provided for even more intensive marketing. On Jamaica, the largest of the British islands, most planters owned land that they would rarely consider using for cane. Over time, these backlands became the preserve of slaves; by the late eighteenth century, slaves were even permitted, in Bryan Edwards’s words, “to bequeath their grounds or gardens to such of their fellow-slaves as they think proper.” Moreover, the planters so fully recognized the slaves’ rights to these grounds that they offered compensation wherever it became necessary to convert an area of slave cultivation to estate use. As Sidney W. Mintz and Douglas Hall have emphasized, “It is upon the polinks that the foundations of the free peasantry were established.”75 By the late eighteenth century, slaves had become the most important suppliers of foodstuffs to all Jamaicans. Many of the island’s minor exports—everything from gums to arrowroot, from oil nuts to goatskins—were produced on their grounds, and Edward Long could estimate that one-fifth of the currency circulating on the island was in their hands.76 In the case of islands like Jamaica, Saint Vincent, and Saint Domingue, the favorable commercial opportunities brought about by a conjunction of planter preferences, soil, topography, and climate, not to mention the abilities of slaves, outweighed many of the disadvantages inherent in a gang system. Indeed, on these islands, the extensive provision ground system may be said to have arisen in spite of the gang system, although, in fact, the grounds offered the same sort of psychological and economic benefits of freedom from supervision and opportunities for choice which were elsewhere offered directly by the task system. Still, it is worth speculating whether taskworkers in the larger islands were not the best equipped to take advantage of these favorable commercial opportunities. This can only be conjecture at present, because little of the detailed research that would be necessary to test this hypothesis has been undertaken. But perhaps the gradual introduction of taskwork into some parts of the sugar cycle around the turn of the nineteenth century helps explain the simultaneous burgeoning of the provision ground economy. The development of the provision grounds undoubtedly was encouraged by what Edward Brathwaite has termed the “Humanitarian Revolution,” of which one important, and 75  Edwards, History of British Colonies, II, 137; Mintz and Hall, Origins of Jamaican Internal Marketing, 3–26 (quotation on 9). See also McDonald, “Goods and Chattels.” For the situation in Cuba, see Rebecca J. Scott, Slave Emancipation in Cuba: The Transition to Free Labor, 1860–1899 (Princeton, N.J., 1985), 15–16, 50, 149–150, 183, 244–245. 76  Mintz and Hall, Origins of Jamaican Internal Marketing, 3–26.

1290

Morgan

neglected, component was the introduction of tasking.77 What remains is to discover the precise dimensions of its role. We also will need studies of the provision ground economies in the crown colonies, where tasking was more fully introduced. We also will need to bear in mind comparative remarks like that of Daniel McKinnen, who attributed his observation that “negroes in the Bahama Islands discover, in general, more spirit and exertion than in the southern parts of the West Indies” to the daily and individual allocation of labor to which they were subject.78 The situation on Bahian sugar plantations is also instructive. There, it is clear that organizing cane cutting by quota and allowing field hands to use their time freely once tasks had been completed were closely linked to the slaves’ own provision grounds. The link is most compellingly drawn in a list of demands submitted by a group of rebellious slaves to their owner. High on their agenda were improvements in the conditions of labor. Reductions in work quotas, for instance, constituted one major claim. But these slaves did not just aim for less taxing work; they sought to control their own work pace. They demanded the right to choose their own overseers and to “play, relax, and sing any time we wish.” Leisure was not, however, their primary goal. Rather, as Schwartz observes: “Above all, the slaves’ concern [was] to have their own land, grow their own food, and market the surplus. The slaves of Santana requested Friday and Saturday of each week for their own endeavors, the right to plant rice and cut wood wherever they wished, and to be provided with fishing nets and canoes.” A search for more such documents, in which slaves lay bare heartfelt grievances and deep aspirations, would advance immeasurably our understanding of the connections between work and culture in their world.79 Another avenue of research would be to make comparisons between coffee and sugar workers. Certainly, the question of timing complicates the comparison, as the Jamaican case reveals. Coffee production became important in Jamaica only in the late eighteenth century, so it cannot be expected that the provision grounds and internal markets of the coffee regions would rival those of the long-established sugar areas. On the other hand, the general conditions of coffee estates were far more conducive to the workers’ well-being than those of sugar plantations: labor demands were lower, and slave populations were healthier. Added to this, however, was the significant advantage of being able, on occasion, to finish work by late afternoon. A Jamaican bookkeeper 77  Edward Brathwaite, The Development of Creole Society in Jamaica, 1770–1820 (Oxford, 1971), xiv. 78  McKinnen, Tour through West Indies, 172. 79  Schwartz, Sugar Plantations, 156–159.

Task And Gang Systems

1291

recalled that the slaves on his coffee plantation “quit the field, their task satisfactorily done, an hour before the usual time of field-work in the ordinary method.” Not surprisingly then, a study of missionaries in late-eighteenth- and early nineteenth-century Jamaica concludes that the missions in the mountain districts of Saint Mary, Saint Andrew, and Saint Elizabeth benefited immeasurably from the coffee workers’ comparative freedom. Weekday visits to stimulate attendance on Sundays gained a more favorable reception on coffee, as opposed to sugar, estates. Similarly, John Stewart noted that coffee workers possessed more livestock than sugar workers, in part a reflection of the greater time on their hands.80 One aspect of the slaves’ well-being that has been intensively studied in the antebellum South, namely the susceptibility of slave children to sudden infant death syndrome, reveals a marked contrast between areas of gang and of task labor. The smothering of slave infants, as contemporaries termed this phenomenon, is now thought to be closely related to the harshness of physical labor required of pregnant slave women. Death rates for smothered slave infants were much higher in areas where gang labor prevailed. In Georgia’s cotton counties, the smothering death rate was more than four times greater than that of the same state’s rice counties. Since the coastal counties were generally more unhealthy than the interior, this might seem a surprising finding. Moreover, a similar contrast can be detected in Virginia between those areas still committed to tobacco production and gang labor as against those areas that had converted to wheat production and individual or small gang labor. The author of this study concludes, “This is a pattern one would expect to find if the hard work that masters assigned to pregnant slave women were the major cause of slave smothering deaths.”81 Finally, there are even tantalizing suggestions that taskwork and property owning might even have had a bearing on slave psychology, perhaps even their propensity to rebel. One historian of slavery in Saint Domingue has noted that “the caféterie left more room for the expression of the slave’s individuality” 80  Kelly, Voyage to Jamaica, 19; Mary Turner, Slaves and Missionaries: The Disintegration of Jamaican Slave Society, 1787–1834 (Urbana, Ill., 1982), 41; John Stewart, A View of the Past and Present State of the Island of Jamaica … (New York, 1969 [orig. publ. Edinburgh, 1823]), 267–268. Because of their greater free time, slaves on a coffee estate, one contemporary explained, “therefore … have good provision grounds … [and] are much more comfortable, and less harassed than on a sugar estate” (R. Bickell, The West Indies as They Are; or, A Real Picture of Slavery … [London, 1825], 23). 81  Michael R Johnson, “Smothered Slave Infants: Were Slave Mothers at Fault?” Journal of Southern History, XLVII (1981), 493–520 (see especially 515–519; quotation on p. 518).

1292

Morgan

than sugar, and a late-eighteenth-century account of coffee planting in Saint Domingue even attributed the slave revolt of that island to the common “foolish fancy of enriching negroes,” presumably a direct outgrowth of the time available for independent production and acquisition by slaves on coffee estates.82 Similarly, Barry Higman has argued that the 1831 slave rebellion in Jamaica, while it was concentrated in the monocultural sugar areas, was, nevertheless, led by slaves “who had achieved some degree of status within slave society.” Higman quotes John Baillie, who had a slave executed for his part in the insurrection: this bondman was said to be “worth six, seven, or eight hundred pounds; he had cattle running upon my estate, and other property…. He was originally a mason, and made a great deal of money in that way.”83 Property owning did not necessarily defuse a slave’s revolutionary ardor. On the other hand, the connection between rebelliousness and tasking or marketing or property owning was never straightforward. Tasking was preferred by slaves because, at bottom, it involved their participation, since it called upon them to share in production decisions. As soon as masters recognized their slaves’ wills in this way, they were, in some senses, acknowledging the humanity of their bondmen. Eliciting this acknowledgment marked a victory for the slaves, however small. Indeed, extracting this admission was in itself a form of resistance, for slaves were now learning valuable skills and resisting the dehumanization inherent in their status. At the same time, however, it is not difficult to see how tasking (or marketing, subsistence cultivation, property owning, or any other independent activity by slaves) could be said to reduce their ability to resist, since it made the state of slavery more tolerable. A slave who had more time for his own affairs might well be a less-discontented slave. Easing the torments of slavery and yet preparing bondmen for freedom,

82  Michel-Rolph Trouillot, “Motion in the System: Coffee, Color, and Slavery in EighteenthCentury Saint-Domingue,” Review, V (1981–1982), 376; P. J. Laborie, The Coffee Planter of Saint Domingo … (London, 1798), 179–180. See also Gabriel Debien, Les esclaves aux Antilles françaises, XVIIe–XVIIIe siecles (Basse-Terre, 1974), 144. Incidentally, it is now thought that Toussaint L’Ouverture was not a privileged slave living on his master’s plantation, but rather was a freeman who owned both land and slaves (G. Debien et al., “Toussaint Louverture avant 1789: Légendes et realités,” Conjonction [1977], 67–80). 83  Higman, Slave Population and Economy, 227–228. There are other associations too, of course. Sidney W. Mintz has explored the connections between slave domestic economies and the rise of postslavery peasantries in his important “Slavery and the Rise of Peasantries,” in Michael Craton, ed., Roots and Branches: Current Directions in Slave Studies (Toronto, 1979), 213–242.

Task And Gang Systems

1293

offering a welcome respite from the rigors of hard driving and yet encouraging slaves to become self-reliant: these were the implications of tasking. And, ultimately, they were profoundly ambiguous.84 The prevailing stereotype of plantation labor, with which this essay began, not only never applied to a large range of New World staples but also became increasingly inappropriate for those crops (most notably, sugar) with which it was, at one time, closely associated. The significance of this development for the laborers has been suggested. But an even wider importance might be claimed. The introduction of piecework or taskwork into a whole range of industries employing free labor raised explosive issues. Many saw the development as a new and more subtle instrument of subordination. Karl Marx, for instance, saw piecework as the form of wage payment most suited to capitalism, because it ensured a maximum intensity of labor and stimulated competitive bidding between laborers. Caribbean task gangs, while not at first operating within a wage system, would have provided Marx with proof for his contentions. However, what Marx did not foresee, and the wider history of task labor in the plantation setting would have provided him with vital clues, was the ability of workers to regulate piecework and turn it to their own ends. The inherent ambiguity of piecework—its potential for either subordination or resistance—was well expressed by a prominent British trade union leader in 1967, when he remarked that, “although all the ills of engineering could be blamed on piece-work, engineers fight to retain it because ‘you have the man on the floor determining how much effort he will give for a given amount of money.’ ”85 Making allowances for a different context, a slave taskworker could easily have echoed the same sentiment.

84  Sidney W. Mintz has been my mentor on these matters, both formally and informally. See his “Slavery and the Slaves,” Caribbean Studies, VIII (June 1969), 69–70; and “Labor Exaction and Cultural Retention in the Antillean Region,” in James Schofield Saeger, ed., Essays on Eighteenth-Century Race Relations in the Americas (Bethlehem, Pa., 1987), 31–52. 85  Quoted in Richard Price, “Rethinking Labour History: The Importance of Work,” in James E. Cronin and Jonathan Schneer, eds., Social Conflict and the Political Order in Modern Britain (New Brunswick, N.J., 1982), 201; Price, “The Labour Process and Labour History,” Social History, VIII (1983), 63.

Rethinking Palmares: Slave Resistance in Colonial Brazil Stuart B. Schwartz Colonial Brazil, based as it was on the coerced labor of Indians and Africans, was continually threatened by various forms of resistance to the fundamental institution of slavery.1 Throughout the Americas wherever slavery was a basic institution, slave resistance, the fear of slave revolt, and the problem of fugitive slaves plagued colonists and colonial administrators. This resistance took a number of forms and was expressed in a variety of ways. Day to day recalcitrance, slow downs, and sabotage were probably the most common forms of resistance, while self-destruction through suicide, infanticide, or overt attempts at vengeance were the most extreme in a personal sense. In Brazil, the most dramatic examples of collective action were a number of slave revolts that took place in Bahia in the early nineteenth century, but actions like the Malê rebellion of 1835 were truly extraordinary events.2 By far, the most common form of slave resistance in colonial Brazil was flight, and a characteristic problem of the Brazilian slave regime was the continual and widespread existence of fugitive communities called variously mocambos, ladeiras, magotes, or quilombos. There was a time when Brazilian historiography ignored this aspect of Brazil’s past, but work during the past half century, especially on the great

Source: Schwartz, Stuart B., “Rethinking Palmares: Slave Resistance in Colonial Brazil,” in Stuart B. Schwartz, Slaves, Peasants, and Rebels: Reconsidering Brazilian Slavery, Urbana: University of Illinois Press, 1992, 103–136. Copyright 1992 Board of Trustees. Used with permission of the University of Illinois Press. 1  This chapter has appeared in Portuguese as “Mocambos, Quilombos, e Palmares: A Resistência escrava no Brasil colonial,” Estudos Econômicos 17 (1987), 61–88. It includes portions of my earlier article, “The Mocambo: Slave Resistance in Colonial Bahia,” Journal of Social History 3 (Summer 1970), 313–33. 2  The series of slave revolts in Bahia between 1807 and 1835 has been studied by João José Reis in “Slave resistance in Brazil, Bahia, 1808–1835,” Luso-Brazilian Review 25:1 (Summer, 1988), 111–44. See also Stuart B. Schwartz, Sugar Plantations in the Formation of Brazilian Society (New York, 1985), especially Chapter 17: “Important Occasions: The War to End Bahian Slavery,” 468–88. On the Malês see João José Reis, Rebelião escrava no Brasil (São Paulo, 1986).

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_041

Rethinking Palmares: Slave Resistance In Colonial Brazil

1295

fugitive community of Palmares, has considerably changed this situation.3 Still, in many ways, the topic of slave flight and resistance in Brazil has been treated as a deceptively simple one, and analyses of it have often been based on a limited set of questions to which common sense answers have been made: Why did slaves flee? To escape slavery. Where were runaway communities located? Far from possible white retaliation. Why did fugitives attack white society? To liberate their fellows and because they hated slavery. Was there class solidarity among slaves? Of course. What kind of societies did fugitives create? More or less egalitarian ones based on African traditions. Noticeably missing from the study of marronage in Brazil has been concern with some of the issues that have preoccupied students of this phenomenon in other American slave societies or solid evidence that would illumine some of the more intractable questions about ethnic solidarities, political goals, and strategies, as well as variations in form. For example, distinctions between the petite marronage of slaves who absented themselves for short periods and those who fled to escape slavery altogether has rarely been made in Brazilian historiography.4 Maroon intentions have preoccupied much writing in Jamaica and Haiti, but 3  For a general overview of the subject see Clovis Moura, Rebeliões da senzala, 3d. ed. (São Paulo, 1981), and his Os quilombos e a rebelião negra, 2d. ed. (São Paulo, 1981). See also, José Alipio Goulart, Da fuga ao suicídio. aspectos de rebeldia dos escravos do Brasil (Rio de Janeiro, 1972). There has been a considerable development of the regional historiography of the quilombos. For example, on Pará there is Vicente Salles, O Negro no Pará (Rio de Janeiro, 1971); on Rio Grande do Sul, Mário José Maestri Filho, Quilombos e quilombolas em terras gauchas, (Porto Alegre, 1979); on Minas Gerais, Waldemar de Almeida Barbosa, Negros e quilombos em Minas Gerais (Belo Horizonte, 1972); on Bahia, aside from the above mentioned article by Schwartz, there is much material contained in Pedro Tomás Pedreira, Os quilombos brasileiros (Salvador, 1973). Many other works dealing with slavery in general at the local or regional level contain information on quilombos. See, for example, Ariosvaldo Figueiredo, O Negro e a violência do Branco (Rio de Janeiro, 1977) on Sergipe. The classic works on Palmares remain Edison Carneiro, O quilombo dos Palmares (São Paulo, 1947) and M. M. de Freitas, O reino negro de Palmares. 2 vols. (Rio de Janeiro, 1954), to which should now be added Décio Freitas, Palmares: a guerra dos escravos (Porto Alegre, 1973). 4  The distinction between mocambo resistance and “petite marronage” was recognized in Brazil. In December 1698 the crown, in response to a petition from the town council of Olinda, ordered that slaves who fled from one plantation to another could not be imprisoned like those who fled to Palmares. (AHU, Conselho Ultramarino, cod. 257, f.1). See the discussion in Gabriel Debien, “Le marronage aux Antilles Françaises au xviiie siécle,” Caribbean Studies 6:3 (1966), 3–44, a portion of which appears in Richard Price, Maroon Societies (New York, 1973), 107–34; and the classic account of Yvan Debbasch, “Le marronage: essai sur la désertion de l’esclave antillais,” L’Année Sociologique (1961), 1–112; (1962), 117–92. Gerald W. [Michael] Mullin, Plight and Rebellion: Slave Resistance in Eighteenth Century Virginia

1296

Schwartz

not in Brazil, except for the case of Palmares. To what extent did escaped slaves organize a resistance that consciously aimed at overthrowing or at least attacking slave society rather than seeking their personal freedom is a question that has remained without answer in Brazil, even though such an answer would provide a measure of the “revolutionary” nature of escaped slave communities.5 To some extent these questions are difficult to answer because of a lack of appropriate documentation, but a close reading of local sources and the use of ethnohistorical techniques may begin to offer a few tentative answers to some of the central questions of fugitive communities in Brazilian slave society. In this chapter, I examine aspects of fugitive communities in three major areas of colonial Brazil: the plantation zone of Bahia; the mining district of Minas Gerais; and the inaccessible frontier of Alagoas, site of Palmares, the largest fugitive community. The goal here is to find patterns in the origins, creation, internal organization, and destruction of these fugitive communities in order to better understand the slave regime and the way in which Africans and Afro-Brazilians responded to it.

Bahia: A Plantation World

Runaway communities flourished in almost all areas of the captaincy of Bahia, although in some regions the problem was unusually acute. The geography and ecology of much of the Bahian littoral aided escape, and the result was a large number of fugitives and mocambos. A report by an unnamed Jesuit written in 1619 outlined the problem and its perception by white society: this people has the custom of fleeing to the woods and joining in hideouts where they live by attacks on the settlers, stealing livestock and ruining crops and cane fields which results in much damage and many losses beyond that of loosing [sic] their daily labor. And many of these (escapees) live for many years in the forest never returning and living in these mocambos which are places or villages that they have made deep in (New York, 1972) is an excellent study of the motivations, backgrounds, and actions of fugitive slaves. 5  This is a major theme addressed on a broad scale in Eugene Genovese, From Rebellion to Revolution: Afro-American Slave Revolts in the Making of the Modern World (Baton Rouge, 1979). For Haiti the subject has become a major issue as is evidenced in Leslie F. Manigat, “The Relationship between Marronage and Slave Revolts and Revolution in St. DomingueHaiti,” CPSNWPS, 420–39.

Rethinking Palmares: Slave Resistance In Colonial Brazil

1297

the forest. And from here they set out to make their assaults, robbing and stealing and often killing many, and in these attacks they seek to carry off their male and female relatives to live with them like gentiles.6 The frequency of mocambo formation and the extent of their geographical location within the captaincy is underscored by the following table that spans over two centuries. Certain characteristics of the captaincy of Bahia contributed to slave flight and the formation of runaway communities. A major terminus of the Atlantic slave trade and a major plantation zone throughout its history, Bahia had always maintained a large servile population, which by the end of the colonial era constituted about one-third of the total population. However, in plantation zones, slaves often made up over 60 percent of the inhabitants. Conditions on the engenhos were physically exhausting and treatment in terms of food and housing was poor. Sometimes slaves had to deal with particularly cruel or sadistic masters, but even beyond these the general concept of slave management disregarded the long-term benefits of “good” treatment and emphasized the extraction of as much labor with as little cost as possible.7 Slaves also lived with limited familial opportunities. Patterns in the Atlantic slave trade and a planter preference for young adult males over women resulted in a chronic sexual imbalance. These problems made for a population that had less to lose by flight or other forms of resistance, at least in the view of observers in nineteenth-century Brazil who advocated stable families and a balanced sex ratio among the slaves as a means of control.8 However, these “progressive” ideas were not shared by the slaveowners of colonial Bahia. Flight and mocambos remained a feature of Bahian slavery throughout its history. While the sugar-growing parishes of the Bahian Recôncavo contained the largest number and highest percentage of slaves, the region of Bahia experiencing the greatest incidence of mocambo formation was the southern towns of Cairú, Camamú, and Ilhéus. These towns and their surrounding districts were devoted for the most part to the production of manioc, the basic 6  The term “gentiles” (gentio) was applied to Indians beyond Portuguese control and thus still pagan. It was a pejorative term and had the implied meaning of “barbarian” (baibaro) to which it was often joined. ARSI, Bras.8 (StL VFL Roll 159). 7  For a particularly sadistic master in eighteenth-century Bahia see Luiz R. B. Mott, “Terror na Casa da Torre: tortura de escravos na Bahia colonial,” in Escravidão e invenção da liberdade, ed. João José Reis, (São Paulo, 1988), 17–32. 8  See my discussion of Bahian slave demography in Sugar Plantations in the Formation of Brazilian Society (Cambridge, 1985), 338–78.

1298 Table 3 Date

Schwartz Partial list of Bahian Mocambos Type

1614 1629 1632 1636 1640 1655 1666 1666–67 1667 1681–91

mocambo mocambo mocambo mocambo

1687

mocambo

mocambo

mocambo mocambo

1722 1723 1726 1733 1734

mocambo mocambo mocambo mocambo mocambo

1745 1789 1791

Irará

Acaranquanha

mocambo

Quiricós Camisão

mocambo

1744–64 quilombo

quilombo

Locale

Size/ comments

sertão Rio Vermelho

mocambo

1692 1699 1705 1706 1713 1714

1735 1736

Name

Buraco de Tatú Santana Matas do Concavo

Itapicuru Rio Real Jeremobão Inhambupe Torre Jaguaripe Serra de Jacobina Rio Real, Inhambupe Camamú Cairú Jacuipe Jaguaripe Maragogipe “campos de Cachoeira” Cairú

Canaveíras Santo Amaro, Nazaré Jacobina Rio das Contãs Itapuã Santo Amaro Ilhéus Jacuipe

40 +

60 +

Sources

1 2 3 4 5 6 7 8 9 10 11

mulatto leader

12 13 14 15 16 17

400 “grande” “old” “grande”

18 19 20 21 22 23 4

61 +

25 26 27 28

1299

Rethinking Palmares: Slave Resistance In Colonial Brazil

Date

Type

1796

quilombo

1801 1807 1807

quilombo mocambo, quilombo quilombo

1809 1825 1826

quilombo mocambo quilombo

Name

Locale

“cabula”

Serra de Orobó Jacobina suburbios

Rio das Contas

Urubá

Sources: 1. Leite, História de companhia de Jesús no Brasil, 10 vols. (Lisbon, 1938–50), 5: 265. 2. Moura, Rebeliões, 75. 3. ACS, 1:213. 4. ACS, 1:310–11, 329. 5. ACS, 1: 6. Pedreira, Os Quilombos, 78. 7. Documentos do Arquivo Nacional, 27, 25. 8. DH, 11:385–86. 9. DH, 8:301–2. 10. ACS, 124.1. 11. ACS, 124.1. 12. Ignácio Accioli de Cerqueira e Silva, Memórias históricas e políticas da Província da Bahia, ed. Braz do Amaral, 6 vols. (Bahia, 1925), 2:142. 13. APB, CdG 150, f. 144v. 14. Pedreira, Os Quilombos, 83–84. 15. Ibid. 16. Ibid., 86–88. 17. Ibid., 91–92. 18. AHU, Bahia, p.a. 27.

Size/ comments

29

“innumerable”

Ilhéus Cachoeira Itaparica suburbias

Sources

30 31 32

“grande” muitos 50 +

19. DH, 45, 20. Pedreira, Os Quilombos, 101–2. 21. DH, 75, 106, 133, 138. 22. DH, 75, 298; 76, 20–21. 23. DH, 76, 81. 24. DH, 76, 335. 25. AHU, Bahia, cat. 6,456. 26. BNRJ, II-34,6,32. 27. APB, CaG, 28. APB, OR 86, f. 242–45. 29. Ibid. 30. Pedreira, Os Quilombos, 123–24. 31. BNRJ, I, 31,27,1: AHU, Bahia, cat. 29.815. 32. APB, Cdo G ao SMgd. 177. 33. APB, Cao G ao, 218. 34. BNRJ, II-33,26,35. 35. Pedreira, Os Quilombos, 141–43.

33 34 35

1300

Schwartz

subsistence crop of Brazil. The work requirements were less than those of the sugar plantations and slaves lived in smaller units in this region. While the predominance of slaves in the population found in the sugar-growing areas was absent, the proportion of slaves in the population of this southern zone still reached between 40 and 60 percent. Relatively good conditions in terms of work requirements, diet, and physical well-being as well as a large proportion of slaves in the population have been postulated in other situations as factors that stimulated slave resistance.9 In this case, however, the frontier nature of this region and its unstable military conditions was the most important contributing factor to successful slave escapes. Cairú and Camamú were constantly threatened by attack from hostile Aimoré Indians. This fact and the distance from military aid coming from Salvador made suppression of slave mocambos difficult. Attacks by Indians or, gentio barbaro (savage gentiles), and mocambo depredations were linked in the mind of the colonists, and various measures were taken to suppress both. Expeditions to eliminate these threats set out in 1663, 1692, 1697, and 1723, but the frequency of repetition indicates a lack of success. Black and mulatto freedmen, “tame” Indians, and black militia units from Bahia were all used in these expeditions, but a major innovation was the use of Indian fighters and backwoodsmen (bandeirantes) from São Paulo.10 This tactic, initiated in the 1670s by Governor Afonso Furtado do Castro do Rio Mendonça, had some success, and “Paulista” contingents were subsequently employed elsewhere in the Northeast for similar operations, the most notable being the destruction of the great quilombo of Palmares in 1684–85.11 While the mocambos of southern Bahia never reached the size or extent of Palmares, the threat they posed was no less real. One mocambo was reported in 1723 to have over four hundred inhabitants, but size alone was not the sole determinant of mocambo danger in this region.12 In another instance in 1692, 9  Marion D. deB. Kilson, “Towards Freedom: An Analysis of Slave Revolts in the United States,” Phylon 25 (1964) 175–87; see also Orlando Patterson, The Sociology of Slavery (London, 1967), 274–80. 10  D. João de Lencastre to Câmara of Cairú (10 Dec. 1697), APB, Cartas do governo 150. In 1667 Gov. Alexandre de Sousa Freire asked the governor of Pernambuco for forty black militiamen to be used with Bahian blacks and Indians against quilombos in Cairú and Camamú. (AHU, Bahia pap. avul. caixa 10, 1st ser. uncatalogued). 11  See the discussion in Stuart B. Schwartz, ed., A Governor and His Image in Baroque Brazil. The Funeral Eulogy of Afonso Furtado de Castro do Rio de Mendonça by Juan Lopes Sierra, trans. Ruth Jones (Minneapolis, 1979), 11–14, 43–49. Sebastião da Rocha Pitta, História da America Portugueza (1724) (Lisbon, 1880), 192–97. 12  King to Gov. Vasco Fernandes Cezar de Meneses (12 Feb. 1723), AHU, Conselho Ultramarino codice 247.

Rethinking Palmares: Slave Resistance In Colonial Brazil

1301

a group of fugitives led by five mulatto “captains” began to sack the farmlands near Camamu and threatened to seize the town itself. Not only was southern Bahia disrupted, but the Recôncavo was also thrown into turmoil as word of these events reached the slave quarters of the engenhos and planters feared a similar outbreak. A Portuguese military expedition in 1692 finally destroyed this mocambo by laying siege to the stockaded village. The final battle cry of the defenders: “Death to the whites and long live liberty.”13 The fear that towns like Cairú and Camamú, far from the centers of governmental authority, might actually be seized was not wholly exaggerated. In 1767, the interim captain of Sergipe de El-Rey reported continual depredations by armed bands of fugitives, adding that in the time of his predecessor an armed band of escaped slaves had marched into town at nine in the morning with flags, drums, and crowns on their heads and demanded that the royal official grant them letters of manumission. The official had sounded the alarm but the absence of troops allowed the fugitives to escape unharmed.14 Such audacity also underlined a basic reality. The majority of Bahian mocambos were located relatively close to population centers or to the surrounding plantations. Whereas Palmares flourished in the remote interior of Alagoas and other fugitive communities also existed in remote regions, the vast majority of the Bahian mocambos and those elsewhere in Brazil remained close to towns and farms, although often in inaccessible locations. In fact, some of the towns within the present-day urban network of Salvador originated as runaway communities. The reasons for this pattern of fugitive settlement are varied. Certainly, until the eighteenth century, hostile Indians constituted an effective barrier to black as well as white penetration of many regions. Most importantly, the internal economy of the mocambos made proximity to settled areas a prerequisite for success. Rather than a return to African pastoral or agricultural pursuits, mocambo economies were often parasitic, based on highway theft, cattle rustling, raiding, and extortion. These activities might be combined with agriculture as well, but rarely did mocambos become wholly self-sufficient and completely

13  Consulta, Conselho Ultramarino (9 Nov. 1692), DHBNR 89 (1950), 206. The quilombo activity in Camamú in this period created fears of a general slave uprising. Governor Antônio Luiz Gonçalves Câmara Coutinho wrote to the Conselho Ultramarino that “em Camamú se levantarem huns mulatos e convocarem asi grande quantidade de Negros querendose fazerse senhores daquella villa.” See BA, 51-IX-30 (Bahia, 23 June 1692). 14  José Lopes da Cruz, capitão interino de Sergipe to Gov. of Bahia (26 Sept. 1767), APB, Cartas ao governo 198.

1302

Schwartz

isolated from the colonial society that generated, and at the same time feared them. The depredations of escaped slaves led colonial officials to consider mocambo fugitives along with highwaymen to be common criminals, thus subject to regular criminal penalties. However, the actions of fugitive slaves were more than simple crimes because they implied, and were sometimes overtly directed as, an attack on the existing social order. In a very real sense mocambo depredations foreshadowed the social banditry, or cangaço, of postcolonial Brazil. The mocambo represented an expression of social protest in a slave society. Anti-Mocambo Measures Colonists and royal officials developed a number of measures to deal with mocambo formation and activity. One tactic was to apprehend fugitives before they could join together in bands. As early as 1612, Alexandre de Moura, captain of Pernambuco, petitioned the crown for the creation of a capitão de campo (bush captain) in each of the eight parishes of that captaincy, who with the aid of twenty Indians would hunt down escaped slaves.15 It is uncertain exactly when these officers were introduced to Bahia but by 1625 the town council of Salvador had set the scale of rewards for these slave hunters. The capitão do campo, or capitão do mato as the office came to be called, worked on a commission basis, receiving a reward for each fugitive captured. This system was formalized in 1676.16 The câmara (town council) of Salvador fixed the price of reward in accordance with the distance involved. By 1637, these rewards were extended to anyone capturing a fugitive, not just to the bush captains. As we

15  B I, Correspondência de Alvaro e Gaspar de Sousa (17 Aug. 1612), f.81. 16  A CB, I,4. The terms capitão do campo and do matto were used interchangeably. By the late seventeenth century letters patent were granted for positions as capitão mor das entradas dos mocambos. See for example, ACS, 124.1 f.126 (10 Nov. 1687) and AHU, Bahia pap. avul. caixa 26, 1st uncat. (1 July 1718). Governor Fernando José de Portugal wrote in 1788 of the necessity of the capitães das entradas and pointed out that the royal treasury spent nothing on their maintenance since it was the slaveowners who paid for their services according to a law of January 28, 1676. [“a fazenda real nada dispende com estes postos pois os senhores dos negros que fogem são os que satisfazem as diligencias em virtude de hum regimento dado aos capitães de assaltos em 28 Jan. 1676”] Portugal to Martinho de Mela e Castro (30 April 1788), Inventários de documentos relativos ao Brasil, ed. Eduardo Castro e Almeida, 8 vols. (Rio de Janeiro, 1914), III, doc. 12.917.

Rethinking Palmares: Slave Resistance In Colonial Brazil

1303

shall see, a similar system was adopted in Minas Gerais and elsewhere in Brazil. The bush captain became a ubiquitous aspect of rural Brazil.17 The system was not without difficulties. To claim the prescribed rewards, overzealous bush captains were not above arresting slaves who were merely on errands. Slaveowners sometimes showed a marked reluctance to pay the fee for old or infirm slaves who were no longer useful. On a number of occasions, a backlog of elderly unclaimed fugitives in the municipal jail of Salvador forced the town council to auction the prisoners to the public in order to pay for expenses. The position of bush captain often attracted somewhat marginal individuals, former slaves and colored freedmen, who were resented by the slaveowners and hated by the slaves.18 Still, the bush captains provided a relatively effective means of apprehending individual fugitives, although controlling the problems of slave revolt or the activities of already formed mocambos was usually beyond their capacities. A second and still unstudied method of slave control and capture in Brazil was the calculated use of Indians as slave catchers and as a counterforce to mocambos and possible slave revolts. In the sixteenth century, sugar planters and absentee donataries sought to bring Indians from the interior to serve as a defense force against possible slave uprisings as well as a buffer against still unreduced tribes of the backlands.19 In the seventeenth century, colonists in Bahia tried unsuccessfully to have Indian villages located near their farms. The Jesuits objected, fearing colonist exploitation of Indians as laborers; the clerics, however, recognized that Indian allies were the “walls and bulwarks” of the colony. As early as 1614, Indians from the Jesuit mission village of São João were used to destroy a mocambo. Probably the most explicit statement of the usefulness of Indian allies against a restive slave population was made 17  B GUC, cod. 706 (7 March 1703); cod. 709 (5 May 1703), f. 140; cod. 711 (5 March 1744), f.123; and ACB, I (13 Feb. 1637), 328–29. 18  A CB, I 326 (27 Jan. 1637). There are many instances of capitães do mato exceeding their rights or causing trouble. The captain major of Sergipe de El-Rey complained in 1806 of a certain Daniel Dias who was trying to purchase an office as a “capitão das entradas” simply to increase his power and who went about the captaincy drunk and who was by his habits and actions “unworthy of the uniform he wore.” [so para ingrosar mais seos despotismos pois he de numero dos valentes daquele e de custumes pessimos e continuamente anda inbriegado, e se fas pelo seu procidimentos e custumes indigno da farda que tras.] See APB, Cartas ao Governo 208 (Sergipe, 16 Nov. 1806). 19  In the sixteenth century the Duke of Aveiro and the Count of Linhares both absentee landowners in Brazil sought to bring Indians in to their plantations for labor and defensive purposes. See AGS, sec. prov. 1487, (7 Oct. 1603); ANTT, CSJ, maço 8 doc. 9 (28 Aug. 1585); mao 16, provisão, 1586.

1304

Schwartz

in 1633 by Duarte Gomes de Silveira, a colonist in Paraiba who wrote: “There is no doubt that without Indians in Brazil there can be no Negroes of Guinea, or better said, there can be no Brazil, for without them [Negroes] nothing can be done and they are ten times more numerous than the whites; and if today it is costly to dominate them with the Indians whom they greatly fear … what will happen without Indians? The next day they will revolt and it is a great risk to resist domestic enemies.”20 Indian irregulars led by Portuguese officers or captains were consistently and successfully employed against mocambos throughout colonial Brazil. The destruction of virtually every mocambo from Palmares to the much smaller hideouts of Bahia, Rio de Janeiro, and Goias depended to a large extent on Indian troops and auxiliaries. Paradoxically, there are also many references to the incorporation of African and Afro-Brazilian slaves into Indian villages and of Indian inhabitants in fugitive communities. Portuguese authorities feared the disruptive and potentially dangerous nature of such contacts. In 1706, the crown ordered that blacks, mixed bloods, and slaves be prevented from penetrating the interior, where they might join with hostile Indian groups. Despite such measures, Afro-Indian cooperation against both the Portuguese and the Dutch in Brazil was common. In Bahia, a famous example of Afro-Indian collaboration is provided by the long-lived syncretic messianic religion called Santidade, which flourished in the southern areas of the captaincy in the late sixteenth century among Indian groups. By 1613, it was reported that escaped slaves had joined the movement and were participating in its raids and even stealing slaves from Salvador. As late as 1627, despite punitive expeditions, Santidade adherents were still launching attacks.21 This leads us to the still-ignored problem of Afro-Indian contacts and social relations. Despite Portuguese attempts to turn the Indians into allies against potential slave resistance, a number of factors drew African slaves and Indians together. For the runaways and unreduced Indian tribes there was the common goal of opposition to the European-imposed slave regime. Within captivity too Indians and Africans were often in intimate and common contact. Indians continued to comprise a large, although decreasing percentage of the plantation labor force in the period 1580–1650, and marriages between blacks and Indians were not uncommon. Some observers, like the Jesuit Father Belchior Cordeiro, 20  “Información q. hize por mandado de VMg. sobre unos capitulos q. Duarte Gomez de Silveira Vezino de Parahiba embio a la Mesa de Consciência,” AGS, sec. prov. lib. 1583, fs.382–89. 21  The santidade movement is discussed in some detail in Sugar Plantations, 47–49.

Rethinking Palmares: Slave Resistance In Colonial Brazil

1305

felt that African slaves became more tractable when brought into contact with Christian Indians, but whatever the policy, Indian-African contacts did take place. Indians remained throughout the colonial era both the best potential allies and the most effective opponents of slave fugitives.22 The primary tactic employed against mocambos was simply to destroy them and to kill or reenslave their inhabitants. Portuguese opposition to the fugitive communities can be easily explained. Mocambo raids and thefts endangered towns, disrupted production, and cut lines of communication and travel.23 Moreover, a mocambo either by its raids or by its attraction drew other slaves from captivity. Many observers noted the effects of mocambos on the slave quarters and one report of 1692 noted that “no settler will have his slaves secure” so long as mocambos persisted.24 Mocambos posed a threat to the economic and social fabric of this slave regime. For most colonial officials accomodation with mocambos and white inhabitants was simply unthinkable. Unlike Jamaica, where a treaty was finally concluded with the runaway maroons, similar tactics were harshly rebuked when suggested in Brazil. In 1640 Viceroy Jorge de Mascarenhas, Marquis of Montalvão, suggested as a wartime measure that a peace mission of a Jesuit linguist and Henrique Dias, leader of a pro-Portuguese black regiment, be sent to a certain mocambo. The mission would offer freedom to runaways if they would serve in the black regiment and if they agreed to harbor no new fugitives. His suggestion met with a stern rebuff from the planter-dominated câmara of Salvador, which stated: “Under no circumstances is it proper to attempt reconciliation nor to give way to slaves who might be conciliated in this matter. That which is proper is only to extinguish them and to conquer them so that those who are still domesticated will not join them and those who are in rebellion will not aspire to greater misdeeds.”25

22  Serafim Leite, “Enformação dalgumas cousas do Brasil por Belchior Cordeiro,” Anais da Academia Portuguesa da História 2d. ser. 15 (1965), 175–202. 23  In a provocative essay Thomas Flory has suggested that a desire for the lands cleared and developed by mocambo fugitives was also a major impulse for colonial society’s attack upon them. His evidence is drawn mainly from the Palmares case. Moreover, the fact that many mocambos were in inaccessible areas and usually not in lands devoted to the principle export crops tends to argue against this hypothesis, although a desire to obtain improved lands is certainly plausible in some cases. See Thomas Flory, “Fugitive Slaves and Free Society: The Case of Brazil, Journal of Negro History 54:2 (Spring, 1979), 116–30. 24  B A, 51-IX-30 (23 June 1692), f.13v.; Diogo de Campos Moreno, “Report of the State of Brazil, 1612,” edited by Engel Sluiter, HAHR 29 (1949), 518–62. 25  A CB I (25 Nov. 1640), 477–78.

1306

Schwartz

These same sentiments were echoed in 1663 in Bahia by the Viceroy, Count of Óbidos, who wanted the great fugitive redoubt of Palmares destroyed as “punishment and example in order to put an end to the hopes of the other slaves.” He wanted no quarter given to those who resisted and the settlement burned to the ground so that nothing remained but “the memory of its destruction for the ultimate “disenchantment of the slaves” of Pernambuco and Bahia.26 Such extermination was usually carried out by military expeditions conducted by private individuals with local backing or by government troops. Private contracts were sometimes made with backwoodsmen with stipulated rewards for each slave captured. Bush captains, Indian auxiliaries, governmentsponsored military columns—all were designed to confront the threat of the fugitive communities to the slave regime. Mocambo Ethnography: The Case of Buraco de Tatú The varied and disparate documents that mention the activities of escaped slaves in Brazil reveal little about the social and political organization within the fugitive communities. For this reason the documents pertaining to the destruction of the quilombo known as the Buraco de Tatú (Armadillo’s Hole) are of singular importance; for although by no means complete they do provide a glimpse into what may have been the history of a typical Bahian mocambo.27 In 1763 a Portuguese-led military expedition destroyed the Buraco de Tatú, located just east-northeast of the city of Salvador, near the present-day bathing beaches of Itapoam. Responding to complaints and disturbed by mocambo activities, Dom Marcos de Noronha, Count of Arcos and Viceroy of Brazil, began in 1760 a campaign to eliminate fugitive communities. In that year, he appointed Joaquim da Costa Cardoso as captain major of the conquest of savage gentiles (capitão mor da conquista do gentio barbaro) and apparently entrusted him with mounting a punitive expedition.28 Although Costa Cardoso’s commission indicated that hostile Indians were his main objective, he also displayed considerable interest in destroying “various quilombos of Negroes

26  Óbidos to Gov. Francisco de Brito Freyre (9 Sept. 1663), BNRJ 8,1,3, fs. 3v–4. 27  Lucinda Coutinho de Mello Coelho, “O quilombo Buraco de Tatú,” Mensário do Arquivo National 10:4 (1979), 4–8. Some further details were added by Pedro Tomás Pedreira, “Sobre o Quilombo Buraco de Tatú,” M.A.N. 10:7 (1979), 7–10. 28  A NTT, Chancelaria D. José I, livro 70, f.257v. (11 Jan. 1762).

Rethinking Palmares: Slave Resistance In Colonial Brazil

1307

in the outskirts of the city.”29 Neither their numbers nor their location at the time can be determined, but in addition to Itapoam there were also mocambos in Cairú and Ipitanga. The Buraco de Tatú had existed for twenty years. Like most Bahian mocambos its economy was basically parasitic, based on theft, extortion, and sporadic raiding. The principal victims, however, were not white sugar planters, but rather the blacks who “came every day to the city [Salvador] to sell the foodstuffs they grow on their plots.”30 The most attractive women were also taken back to the mocambo. The chronic lack of women in the Brazilian slave force was reproduced and exacerbated in the mocambos. Fugitives preferred to take black or mulatto women, and there are few references to the abduction of European women. No such charge was made against the inhabitants of the Buraco de Tatú.31 Despite the actions of the fugitives, there were freedmen and slaves who out of necessity or sympathy cooperated with the Buraco de Tatú. João Baptista, a mulatto farmer, worked with the runaways and supplied them with firewood.32 He was apparently not alone in his practices. Blacks in the city of Salvador aided the quilombo by helping the fugitives to enter the city by night in order to buy powder and shot. Such contact was unsettling to slaveowners and royal officials, who feared increased escapes or a general uprising. As in other instances, whites also cooperated with the quilombo—in their case to avoid harm to life or property. The actions taken against non-guilombo blacks and the cooperation between whites and the escaped slaves, although coerced, indicates that the Buraco de Tatú fugitives had no intention of a total war of liberation against all the slaveowning segments of the population. Quilombos, in fact, might provide focal points in more general slave rebellions as they did in Bahia in the early 29  A HU, Bahia pap. avul. no. 6451. 30  Ibid. n.6449. There is a copy of the document in IHGB, 1.119 (Correspondência do Governador da Bahia 1751–82). The document was also printed in Inventário dos documentos relativos ao Brasil, ed. Eduardo Castro de Almeida, 8 vols. (Rio de Janeiro, 1914), 2 (Bahia 1763–86), 44–45. 31  The rescue of two young white Paulista girls and their brother from a quilombo in Minas Gerais is the subject of a report of 1737 in which the author wrote: “it was a painful act to see the tears and lamentations with which the children were received (by their mother) mixing at the same time joy and sadness.” See ANTT, Ms. do Brasil 11, 153–154v. (9 Jan. 1737); ms. do Brasil 4 (8 March 1737). Sexual fear of slaves seems to play a relatively minor role in the anti-quilombo campaigns, although traces of these sentiments sometimes emerge. 32   “Certidão da sentena condemnatória dos negros do quilombo Buraco de Tatú (12 Jan. 1764),” AHU, Bahia pap. avul. n.6456.

1308

Schwartz

nineteenth century, but in general, the goals of the fugitive communities seem to have been the more immediate and practical ones of survival beyond the control of white society. Slave resistance in all its forms may have been a threat to the slave system, but despite the implications of quilombo existence for the slave “class,” planters and colonial officials perceived enough divisions among the slaves to risk arming engenho slaves in order to combat fugitives, as the Conde do Ponte suggested in 1807.33 The Buraco de Tatú was destroyed on September 2, 1763, and from the military descriptions of that action and a plan drawn by the attackers to illustrate those reports it is possible to infer much about the internal life of this community. The quilombo was a well-organized village laid out in a rectilinear pattern of six rows of houses bisected by a large central street (see figure 4). There were thirty-two rectangular residence units (B) and since there were approximately sixty-five adults in the quilombo, we can assume that these units represent houses rather than compounds. The close correlation of two adults per house suggests a monogamous pattern, but the evidence is unclear since the accompanying documents make no mention of children. When children born in a quilombo were captured they often became the property of the expedition’s leader, and this may explain their absence from the judicial records.34 Taken as a whole, the monogamous marital pattern, the rectangular house shape, and the even rows of houses suggest a reproduction of a plantation senzala (slave quarters) rather than any specific African pattern. Conversely, the large central street equally dividing the rectangular houses and the existence of what may have been a ceremonial, or a “palaver,” house in front of a plaza (H) are all elements found among northwest Bantu groups such as the Koko, Teke (Anzico), and Mabea.35 The surviving documents, in fact, give little indication of the ethnic origins of the inhabitants of the Buraco de Tatú. At least one inhabitant was a crioulo. Another was referred to as a mandingueiro, a term that in the mid-eighteenth century simply meant “sorcerer,” but could also suggest Mandinga origins for at least this fugitive. The most reasonable assumption is that no one African group inhabited this mocambo. Like many fugitive communities in Brazil, the Buraco de Tatú was cleverly fortified. Entry into the mocambo was made difficult by an extensive defensive 33  Conde do Ponte to Visconde de Anadia (27 April 1807), APB, Cartas do Governo 177. 34  A CB 1:119 (24 Jan. 1629). 35  See George P. Murdock, Africa: Its Peoples and Their Culture History (New York, 1959), 276. See also Richard W. Hull, African Cities and Towns before the European Conquest (New York, 1976), 33–49.

Plan of the Buraco de Tatú in 1763.

1309

Figure 4

Rethinking Palmares: Slave Resistance In Colonial Brazil

1310

Schwartz

network. The rear was protected by a swampy dike about the height of a man. The three sides of the village were protected by a maze of sharpened stakes (L) driven into the ground and covered to prevent detection by an unsuspecting intruder. This defense was augmented by a series of twenty-one pits (D) filled with sharpened spikes and disguised by brush and grass. Leading into the mocambo was a false road especially well protected by spikes and camouflaged traps. Only when the watchman (N) placed planks (C, O, M) over some of the obstacles did entry and exit become possible. The Portuguese noted the effectiveness of this defense and took pains to point out the problems it created to the crown. It was a defense quite unlike that of the palisaded Angolan quilombo described by Father Antonio Cavazzi in 1680.36 Still, covered traps and sharpened stakes were used for village protection in Africa from Nigeria southward to the old Kingdom of Kongo and were also used at Palmares and by other fugitive communities.37 In the predatory economy of the Buraco de Tatú agriculture was not a major activity. The plan does show a trellis of maracuja (Q) and a number of small gardens perhaps equivalent to the dawn gardens of the Kongo, but these seem to be devoted to herbs rather than staple crops. No roças or farmlands are indicated in the area around the mocambo. These fugitives probably exacted foodstuffs as tribute from their neighbors and may have supplemented their diet with fish since the village was located near the coast.38 A few aspects of the quilombo’s internal life can be gleaned from the report of its destruction. Politically, the Buraco de Tatú had two chieftains or captains. Antonio de Sousa was a war-captain, and a second leader, Theodoro, controlled the quilombo itself (“tive administração do quilombo”). Each leader had a consort who was called a queen. Nine houses were separated from the main village. This separation may simply indicate latecomers or the divided political leadership, or may even suggest the possibility that this was the residence of a lineage unable to live in the main village or even an age group of young males required to live apart. This latter possibility is doubtful, however, since the Portuguese 36  Antonio Cavazzi de Montecuccolo, Istoria descrizione de tre Regni Congo, Matamba, et Angola (Bologna, 1687), 205–7. 37  Georges Balandier, La vie quotidienne au Royaume de Kongo du xvi au xviii siècle (Paris, 1964); J. Van Wing, Études Bakongo (Louvain, 1921), 148. Raymond Kent, “Palmares: An African Kingdom in Brazil,” Journal of African History 6 (1961), 161–75. 38  It is clear that some mocambos practiced agriculture. This is indicated in various documents such as that on the quilombo of Orobó of 1796 reprinted in Donald Pierson, Negroes in Brazil 2d. ed. (Carbondale, Ill., 1967), 49. Such practices, however, were difficult unless the mocambo was isolated and relatively permanent.

Rethinking Palmares: Slave Resistance In Colonial Brazil

1311

would have found this situation notable enough to mention and the records are silent. The religion of the inhabitants is unknown. Two individuals were mentioned as sorcerers, one of them an old woman (R). Women are traditionally the leaders in the Yoruba cults (candomblé) still practiced in Bahia, but the dates of this mocambo (1743–63) precede the large-scale importation of Yoruba slaves in Brazil. The Buraco de Tatú was destroyed on September 2, 1763. Under the leadership of Joaquim da Costa Cardoso a force of two hundred men including a troop of grenadiers, but made up mostly of Indian auxiliary militia and Indians from a village in Jaguaripe, carried out the attack. Their battle orders were to remain in the field until, “the quilombo has been destroyed, the blacks captured, the resisters killed, the woods searched, the huts and defenses burned, and the trenches filled in.”39 Indian guides were used to scout the quilombo’s defenses before the assault was launched. The attack probably came from the unprotected coastal side of the village. Surprise worked to the attackers advantage until an old woman (T) raised the alarm. The defenders, some of whom were armed with bows (P) were greatly outnumbered and overwhelmed. The hero of the defense was José Lopes, who fired two shots at the attackers and, who shouted defiantly that it would take more than two hundred men to capture him. He was mistaken. Four fugitives were killed and sixty-one taken prisoner. No casualties were reported among the expeditionary troops. Upon their capture the fugitives were incarcerated in Salvador. Thirty-one whose only crime was to escape slavery were, in accordance with a Royal Order of March 3, 1741, branded with the letter F ( fugido).40 After their masters paid the costs of capture to the royal treasury, the slaves were returned to captivity. Some were singled out for exemplary punishment. Antonio de Sousa, captain of the quilombo, was sentenced to public flogging and life in the galleys. His friend, Miguel Cosme, “reported to be a great thief,” received a sentence of flogging and six years at the oars. Theodoro and José Lopes were both whipped in public and sent to the galleys for ten years. José Piahuy, “a great backwoodsman and thief,” received two hundred lashes and a four-year term at the oars while the crioulo Leonardo received a like number of stripes. João Baptista, the mulatto farmer and accomplice of the fugitives, was sentenced to five years of penal exile and a stiff fine.41 The two queens received relatively light sentences. The Buraco de Tatú provides an example that recapitulates many aspects of the history of fugitive communities in Brazil. Relatively small in size (less 39  A HU, Bahia pap. avul. n.6649. 40  B GUC, cod. 707, Livro de registro da Relação. 41  A HU, Bahia pap. avul. n.6649.

1312

Schwartz

than 100 inhabitants) and located close to centers of population, these communities developed syncretic traditions fusing Brazilian and African elements. Their inhabitants seem also to have been of various origins, Brazilian-born and Africans of different ethnic backgrounds. Although they stole from slaves and free people of color as well as whites, some freedmen willingly cooperated with the fugitives. The punitive military expedition and the use of Indians within it represented the usual colonial response to the mocambos. Living by their wits and daring, the fugitives of the Buraco de Tatú maintained their independence for twenty years until their actions and the threat of their very existence caused the colonial authorities to exterminate their community. In many ways, the history of the Buraco de Tatú seems to be typical of the history of fugitive communities in Brazil.

Minas Gerais: A Mining Economy

The patterns of quilombo formation and the responses of colonial society that we have examined thus far in the case of Bahia were to a large extent reproduced in the mining areas of south central Brazil, although with certain differences as might be expected given the different social and economic formation of the region. The discovery of rich deposits of gold in the mountainous region that came to be known as Minas Gerais and the subsequent development there of a society based on slave labor created conditions that particularly favored slave runaways and the formation of mocambos.42 Slaves made up between one-third and one-half of the total population of the captaincy during most of the eighteenth century and free people of color constituted by 1821 another 40 percent of the total.43 Together, then, the Afro-Brazilian population—slave 42  There is now a significant literature on quilombos in Minas Gerais. Aside from Almeida Barbosa, Negros e quilombos em Minas Gerais, noted above, see also A. J. R. RussellWood, The Black Man in Slavery and Freedom in Colonial Brazil, (New York, 1982); Julio Pinto Vallejos, “Slave Control and Resistance in Colonial Minas Gerais,” Journal of Latin American Studies 17 (1985), 1–34. Carlos Magno Guimãres, A negação da ordem escravista (São Paulo, 1988), is particularly valuable for its list of quilombos in Minas Gerais and of letters patent for slave catchers. See also his summary article “Os Quilombos do século do ouro,” Estudos Econômicos 18 (1988), 7–45. Some of his findings and other new materials are presented in Kathleen Joan Higgins, “The Slave Society in Eighteenth-Century Sabara: A Community Study in Colonial Brazil” (Ph.D. diss., Yale University, 1987,) 258–307. Still valuable is C. R. Boxer, The Golden Age of Brazil: 1695–1750 (Berkeley, 1962). 43  A. J. R. Russell-Wood, “Technology and Society: The Impact of Gold Mining on the Institution of Slavery in Portuguese America,” Journal of Economic History 34:1 (March,

Rethinking Palmares: Slave Resistance In Colonial Brazil

1313

and free—eventually made up about three-quarters of the inhabitants. Slaves performed virtually every task, but above all they did most of the mining. They were expensive and highly valued. So long as they were productive and turned over the gold they found to their masters, slaves often had considerable autonomy of movement in the mining district.44 The large sea of slaves and free coloreds provided a potentially friendly environment for runaways. The discontinuous nature of settlement and the mountainous topography provided large inaccessible tracts for hideouts and even in the many urban concentrations, the large free population of color made the detection of fugitives difficult. Moreover, because fugitive slaves could often provide gold that they had stolen or found, some whites were willing to cooperate with the mocambos or to protect fugitives. Finally, in the lawless and restive conditions of early Minas Gerais slaves were often armed by their owners and participated in the various antigovernment movements and in the civil disturbances of the War of the Emboabas.45 All of these conditions contributed to an unstable situation of slave control as well as feelings of insecurity and fear among royal officials, municipal councils, and the white population in general. Rumors of planned slave revolts circulated in 1719, 1725, and 1756, but the main problem continued to be mocambos. Throughout the eighteenth century governors, miners, royal officials, and town councils complained of thefts, murders, abductions, and other crimes committed by calhambolas, the inhabitants of mocambos. The response in Minas replicated that of the coastal plantation zones. In Minas, as in Bahia, attempts to use free Indians as slave catchers and the deployment of Indian villages and later of royal troops for this purpose had little impact. Moreover, Minas had distinct problems. Its free population was generally unruly. In the early years of settlement the area experienced a short civil war and a number

1974), 59–83. The demographic structure of Minas Gerais has been analyzed in a series of studies by Iraci del Nero da Costa such as Populações mineiras (São Paulo, 1981); Vila Rica: População (1719–1826) (São Paulo, 1979) and Francisco Vidal Luna, Minas Gerais: Escravos e senhores (São Paulo, 1981). See also their jointly authored Minas colonial: economia e sociedade (São Paulo, 1982). 44  This point is emphasized by Higgins, “The Slave Society,” 258–307. I believe, however, that the high numbers of fugitives and quilombos may be a characteristic of frontier regions in general as much as a result of the peculiar structure of Minas Gerais. There is some indication that the frequency of quilombo formation began to decline as the slave population became demographically more stable and a shift was made from mining to mixed agriculture. 45  Vallejos, “Slave Control,” 6–8; Russell-Wood, The Black Man, 42.

1314

Schwartz

of antigovernment tax riots. The miners at first also refused to pay a tax for fugitive control.46 Not until 1744 were royal judges in Minas Gerais authorized to raise money (up to 300 oitavas of gold) to pay for anti-guilombo operations.47 The activist and racist Conde de Assumar (1717–21) made mocambo control a central concern of his governorship. Like his predecessor, he suggested the arming of Indians and their use as slave catchers. In 1717, he suggested the creation of capitães do mato and by 1722 these posts had been established and a set of standing orders (regimento) had been issued with a sliding scale of rewards for the return of fugitives depending on the distance the captain had to travel.48 The difficulty of recovery from mocambos is reflected in the fact that a fugitive caught within a league of the residence of the bush captain fetched 4 oitavas of gold while one taken from a quilombo (defined as an encampment of more than four Blacks with houses set up) brought 20 oitavas.49 A modern study of local sources was able to identify 117 known quilombos in the region before 1800 and appointments of almost 500 slave catchers during the eighteenth century.50 Assumar and various municipal councils in Minas Gerais were so preoccupied with the problem of slave control and mocambos that they were willing to suggest or try a variety of stringent and extraordinary measures aimed not only at the fugitives, but at the free colored population as well. Assumar was responsible for an attempt to limit the number of manumissions in the region, claiming that the grants of liberty led to slave thefts and prostitution. He also suggested that the large number of free coloreds who controlled property in the area threatened the social hierarchy, and he ordered that free colored residents be prohibited from owning slaves and that no free black could serve as a godparent for a slave. These measures were impossible to enforce but they demonstrated Assumar’s fear of a social order in which the lines of race and class had become blurred. For the threat of mocambos, he had other remedies. Unlike Bahia, where anti-guilombo operations were left to bush captains or to 46  Consulta, Conselho Ultramarino (22 Dec. 1718), IHGB, Arq. 1.1.25. 47  Provisão (2 Dec. 1744), ANRJ, Cod. 542, f.24–25. 48  On Assumar’s campaign against quilombos see Francisco Antônio Lopes, “Câmara e cadeia de Villa Rica,” Anuario do Museu da Inconfidência (1952), 103–251, which reprints important documents. Assumar’s comment on the failure of Indians to stop quilombo formation under his predecessor is found in AHU, Minas pap. avul. crown to Conde de Assumar (12 Jan. 1719). 49  Regimento dos capitães do mato was issued in 1715 and then reissued in 1722. The version that remained in force was that of 17 December 1724. See ANTT, Mss. do Brasil 28, 307–9v. 50  Guimarães, A negação da ordem escravista, 129–71.

Rethinking Palmares: Slave Resistance In Colonial Brazil

1315

officially sponsored expeditions, under Assumar anyone wishing to attack a quilombo could do so and carry whatever arms were necessary.51 Such measures indicate a level of insecurity and fear in the mining zones seemingly exceeding that of the plantation areas. Assumar was perhaps an extreme example of such concern, but he was not alone. Local town councils in Minas also attempted to come to grips with threats to the social order by various ordinances aimed at controlling the free colored population and by actions against fugitives. A number of anti-mocambo operations were organized by town councils. In 1735, the câmaia (town council) of Vila Rica called for the cutting off of a hand as a punishment for fugitive slaves.52 Perhaps the most infamous action was the barbarous suggestion of the town council of Mariana that fugitives when caught have their Achilles tendon severed, allowing them to hobble to work, but making flight almost impossible. This suggestion met with a rebuff by the Viceroy, The Count of Arcos, as an unchristian measure but, of course, he did not live in Minas Gerais. It should be noted that this suggestion by the câmaia of Mariana often cited by historians was part of a more general appeal arguing that the free colored of the captaincy aided the escaped slaves in their crimes and that measures should be imposed limiting manumissions, curtailing the movement of the free colored population, and prohibiting them from bearing or even owning arms.53 These attitudes and fears sometimes had dire consequences. In 1716, a quilombo with 80 to 100 blacks who were raiding the roads near Vila Real and Vila Nova da Rainha was unsuccessfully attacked by a punitive expedition. A second force of 150 men was organized by the two towns; it destroyed the quilombo, killing a number of the defenders in a rage after these had surrendered. The royal judge refused to prosecute the guilty for fear that men would not join such expeditions in the future.54 The Overseas Council in Lisbon lamented theses “excesses,” but supported the judge, although in the regimento for bush captains it warned against the excessive use of force.55 Still, for the most part quilombos and fugitive slaves were beyond the norms of civil society. In 1738, a petition from residents of Vila Rica to be exempted from prosecution for the death of calhambolas met with a favorable response from the governor.56 51  See the pertinent documentation in Lopes, “Câmara e cadeia,” and also Russell-Wood, The Black Man, 42. 52  Russell-Wood, The Black Man, 42–43. 53  A PB, Ordens rgias 55, f.99–99v. 54  “Ouvidor do Rio das velhas da conta …,” IHGB, Arq. 1.1.24. 55  Regimento dos capites do mato, ANTT, Mss. do Brasil 28, fs.307–309v. 56  Laura de Mello e Souza, Desclasificados do Ouro, (Rio de Janeiro, 1982), 113.

1316

Schwartz

All of the traditional methods of mocambo and fugitive control were tried in Minas Gerais, along with a few that were extraordinary. The large population of free coloreds in the captaincy and the many poor “vagrants” presented in themselves a threat to the social order, but one that might be effectively mobilized against the quilombos. Efforts to enlist them in anti-quilombo activities were made on various occasions.57 But these efforts did little to stem the problem. The free poor and the fugitive slaves were both manifestations of the inherent conditions of this society. Quilombos, then, were an endemic problem in Minas Gerais. They were numerous and sometimes attained large size, although here too it is difficult to know much about their internal organization since we must depend on the descriptions of quilombo destruction. For example, an account of an expedition that rescued some white children from a quilombo noted that the “mulatto entitled king, a concubine, and four slaves remained at large.”58 Governor Gomes Freire de Andrada described an attack against a “small” quilombo of over a hundred blacks made in 1746. He noted that three assaults were necessary to overcome their resistance and that twenty or so were killed, over seventy men as well as a large number of women were captured. This “small” quilombo must have contained between one and two hundred people.59 The two largest quilombos of Minas—that of Ambrôsio, destroyed in 1746 and the Quilombo Grande, attacked and eliminated in 1759—held large numbers of fugitives, the latter perhaps containing over a thousand inhabitants.60 These, however, were exceptionally large in size. Minas Gerais, therefore, despite its differing economic basis and social and racial configuration, reproduced and intensified many of the conditions that led to mocambo formation in the zones of export agriculture. The early frontier conditions, the considerable freedom of movement allowed to slaves in mining, the urban network, and the racial composition of the region all contributed to the formation of fugitive communities in the region. The response of colonial government to this problem in Minas Gerais and the techniques used to combat it were similar to those applied elsewhere in the colony. It is striking that the term quilombo was far more frequently used in Minas Gerais than in Bahia, where mocambo was preferred, although by the mid-eighteenth century both terms were in use. The word quilombo, in fact, came to mean an 57  Mello e Souza, Desclasificados, 72.84. 58  A NTT, Mss. do Brasil 11, fs.153–54. 59  Conselho Ultramarino to Gomes Freire (6 May 1747), AHU, Rio de Janeiro pap. avul. caixa 22. 60  Almeida Barbosa, Negros e quilombos, 31–53.

Rethinking Palmares: Slave Resistance In Colonial Brazil

1317

encampment of any group of outlaws, so that in 1737 an official in Vila Rica could report preparations to destroy a “quilombo of whites, in flight for various atrocious crimes (brancos omiziados por crimes atroces).”61 However, the term was used primarily to describe fugitive slave communities; and it has become a symbol of slave resistance in Brazil and, in more modern times, of a movement for equality for blacks in Brazil. The linguistic difference between Bahia and Minas Gerais is to some extent chronological and it is related to the history of the great fugitive community of Palmares, which for almost a century had resisted all efforts to destroy it. Palmares became a symbol for royal administrators of how any fugitive community might become a real threat to civil society in a society so based on slavery. When the Count of Assumar wrote in 1719 that: “the blacks (of Minas) may be tempted to repeat the acts of the Palmares of Pernambuco, emboldened by their multitude,” he was voicing a real fear.62 While Palmares was atypical in its size and its duration, its history can not be separated from that of the other fugitive communities, if only because of its influence on how slaveowners and royal officials viewed the problem. Moreover, because of its longevity and size and the long contact of colonial society with it, Palmares offers some opportunities to penetrate the internal dynamic of a fugitive community.

Rethinking Palmares

In dealing with the question of fugitive communities in Brazil it is necessary to keep the quilombo of Palmares in mind. Palmares, located in the interior of Alagoas, was by far the longest-lived and largest fugitive community. For almost the whole seventeenth century (1605?–1694) it persisted despite determined attempts to eliminate it by the Dutch and Portuguese colonial governments and by local residents of the neighboring captaincies. Because of its reputed size (over twenty thousand inhabitants), longevity, and the continual colonial contact with it, we know more about its internal structure than we do about most of the mocambos. Still, the documentation on Palmares is not extensive and it tends to concentrate on the last decade of its existence and its final destruction.63 Therefore much remains unknown about it. That fact has 61  Assumar to crown (20 April 1719) cited in Vallejos “Slave Control,” 15. 62  See the discussion in Higgins, “The Slave Society in Eighteenth-Century Sabara,” 258–62. 63  The classic accounts of Palmares are Edison Carneiro, O quilombo dos Palmares, 3d. ed. (Rio de Janeiro, 1966), originally published in Spanish as Guerras de los Palmares (Mexico City, 1946); Décio Freitas, Palmares: A guerra dos escravos (Porto Alegre, 1971), now in

1318

Schwartz

not deterred authors from attempting to write its history or for romanticizing it into a “Black Troy,” or a “republic.” More recently, it has taken on a symbolic importance for Afro-Brazilians in their struggle for racial and social equality.64 It has long been recognized that Palmares was based on a number of traditional African forms of political and social organization, although like most fugitive communities it combined these with aspects of European culture and specifically local adaptations.65 Palmares was not a single community but a number of mocambos united to form a neo-African kingdom.66 Various eyewitness accounts reveal much about Palmares’ internal organization, although we must recognize that Palmares also had a history and the organization and institutions noted at the end of the seventeenth century were not necessarily those of the earlier period. Moreover, the size of Palmares also changed over time. A mid-seventeenth century account described Palmares as divided into two main settlements with many smaller ones, placing the population of the various settlements in Palmares at about eleven thousand. A later and often repeated estimate raised that figure to twenty thousand. This latter number seems exaggerated. During most of the seventeenth century Pernambuco and its adjacent captaincies had about two hundred engenhos with an average of a hundred slaves per mill. In other words, the estimate for Palmares of twenty thousand inhabitants would equal the number of all slaves in the sugar

a 4th edition of 1982 and originally published in Spanish as Palmares-la guerra negra (Montevideo, 1971). There is the older but still useful work of M. M. de Freitas, O reino negro de Palmares 2 vols. (Rio de Janeiro, 1954). 64  Indications of the symbolic meaning of Palmares can be seen in works such as Clovis Moura, Brasil: as raizes do protesto negro (São Paulo, 1983); Abdias do Nascimento, O negro revoltado (Rio de Janeiro, 1968), and his more recent Quilombismo (Petrópolis, 1980). Two motion pictures, “Ganga Zumba” (1963) and “Quilombo,” (1984), both directed by Carlos Diegues have taken up this theme. In 1984, a conference was held in Alagoas to commemorate the history and significance of Palmares. 65  The African aspects of Palmares have fascinated scholars since Nina Rodrigues wrote his still useful description of that community in 1906. See Os africanos no Brasil (São Paulo, 1933). The most consistent attempt to identify the African aspects of Palmares was made by Raymond Kent, “Palmares: An African State in Brazil,” Journal of African History 6 (1965), 161–75, but while Kent touches many important issues, his translations and ethnographic discussions can not always be trusted. 66  German anthropologist Stephan Palmié has pointed out that the structure of the “kingdom” of Palmares paralleled that of contemporaneous Bantu states such as the kingdom of Kongo. Stephan Palmié, “African States in the New World? Remarks on the Tradition of Transatlantic Resistance” (ms., 1988).

Rethinking Palmares: Slave Resistance In Colonial Brazil

1319

economy of the region.67 Although this number seems unlikely, Palmares was undoubtedly the largest fugitive community to have existed in Brazil. During its long history, Palmares was constantly under attack. The Dutch mounted three expeditions against it and after Portugal regained control of the Northeast in 1654 the war continued. Between 1672 and 1680 there was a military expedition almost every year. The fugitives resisted valiantly, but this constant pressure caused them to sue for peace with a recently arrived governor of Pernambuco in 1678. The “king” of Palmares, Ganga Zumba, had in fact tried this policy whenever a new governor arrived. Like the maroons of Jamaica, he had promised loyalty to the Portuguese crown and return of any new fugitives in exchange for recognition of the quilombo’s freedom. The Portuguese accepted these terms but soon violated them, and in Palmares itself a revolt took place in which the accomodationist Ganga Zumba was overthrown and killed by his nephew Zumbi.68 The war continued. Expeditions went almost annually against the fugitives in the 1680s but with little success. The defenders of Palmares became masters of guerilla warfare, adept at the use of camouflage and ambush. Frustrated, Portuguese colonial administrators adopted a new tactic. Hardened Indian fighters and slavers from São Paulo who had been used in Bahia to open the interior were now contracted to eliminate Palmares. Their assault began in 1692, and for two years with the aid of local troops and the indispensable Indian allies they slowly reduced the perimeter of the main mocambo’s defenses. The final battle was fought in February 1694. Two hundred fugitives were killed, five hundred captured, and another two hundred reportedly committed suicide rather than surrender. Zumbi, wounded and in flight, was betrayed, captured, and decapitated. Palmares was no more, but as late as 1746, slaves were still fleeing to the site of Palmares and once again forming into fugitive groups. European observers did not always understand what they saw, but from their descriptions it is clear that Palmares was an organized state under the control of a king with subordinate chiefs in outlying settlements. While some accounts speak of an election process, the leadership of one village by Ganga Zumba’s mother and the succession of Zumbi, Ganga Zumba’s nephew to the 67  Francisco de Brito Freyre, Nova Lusitania. História da guerra brasilica (Lisbon, 1675), 280– 82. (I have used the facsimile second edition, Recife, 1977). 68  Décio Freitas, Palmares presents a fascinating biography of Zumbi as a remarkable man, captured as a child in a raid on Palmares and then raised and educated in Latin and Portuguese by a priest in Porto Calvo. In 1670, at the age of fifteen, the youth fled back to Palmares and later rose to leadership. The precise sources of this account are not made clear by Freitas.

1320

Schwartz

throne suggests the existence of a royal lineage.69 The ceremonial postures and demonstrations of obedience required in the king’s presence all point to forms of African kingship. The Palmares’s fugitives lived by agriculture, although like other mocambos they also traded for arms and other commodities with those whites who lived on the borders. Like most of the mocambos they also raided for women, cattle, and food. As in many African societies, slavery existed in Palmares. Those who came to Palmares by choice were considered free, but those taken in raids were enslaved. The villages of Palmares were protected by palisades, walls, or by a network of hidden traps much like those described and shown in the description of the Bahian mocambo of Buraco de Tatú. Religion in the encampments was a fusion of Christian and African elements, although here too there may have been far more African features than observers realized. In many ways Palmares seems to have been an adaption of African cultural forms to the Brazilian colonial situation in which slaves of various origins, African and crioulo came together in their common opposition to slavery. Within Palmares people called each other malungo, or comrade, a term of adoptive kinship also used among slaves who had arrived together on the same slave ship. In Palmares we can see the attempt to form a community out of peoples of disparate origins. Such an attempt had to be made by all fugitive communities, but in the case of Palmares there are some specific features that help to explain its particular history as well as the history of slave resistance in colonial Brazil as a whole. The search for “African” elements at Palmares and in the cultural “survivals” of slaves or fugitives as a whole has too often focused on specific cultural or ethnic identities. In fact, much of what passed for African “ethnicity” in Brazil were colonial creations. Categories or groupings such as “Congo” or “Angola” had no ethnic content in themselves and often combined peoples drawn from broad areas of Africa who before enslavement had shared little sense of relationship or identity. That these categories were sometimes adopted by the slaves themselves indicates not only the slaves’ adaptability, but also the fact African societies had considerable experience with, and a variety of institutions for the integration of disparate peoples and the creation of solidarities across ethnic lines.70 69  There are also accounts of “elections” in Angolan quilombos. See António de Oliveira Cadornega, História das guerras angolanas, 3 vols. (Lisbon, 1940), II, 221. 70  Palmié, “African States,” 10–11. See also his “Ethnogenetic Processes and Cultural Transfer in Afro-American Slave Populations” (ms., 1989). Palmié drew considerable inspiration from Igor Kopytoff, “The Internal African Frontier: The Making of African Political Culture,” in The African Frontier, ed. Igor Kopytoff (Bloomington, Ind., 1987), 3–84.

Rethinking Palmares: Slave Resistance In Colonial Brazil

1321

There is, I believe, a deeper story in Palmares and one with broad implications for the subsequent history of slave resistance in Brazil. A key to the problem lies in the etymology of the word quilombo. This term came to mean in Brazil any community of escaped slaves, and its usual meaning and origin is given as the Mbundu word for war-camp. By the eighteenth century the term was in general use in Brazil, but it always remained secondary to the older term mocambo, a Mbundu word meaning hideout. In fact, the word quilombo does not appear in any contemporaneous document until the end of the seventeenth century except for its midcentury use by the poet Gregório de Mattos, who employed it with the meaning of any place where blacks congregated. The first document I have seen with the term quilombo used for a fugitive community is dated 1691 and it deals specifically with Palmares.71 The chronology and the connection with Palmares are not accidental. Within the term quilombo is encoded an unwritten history that only now because of recent research in African history can be at least partially understood. While Palmares combined a number of African cultural traditions and included among its inhabitants crioulos, mulattoes, Indians, and even some renegade whites, or mestiços, as well as Africans, clearly the traditions of Angola predominated. Its residents referred to Palmares as angola janga (little Angola) in recognition of that fact, and in a complaint of 1672 the municipal council of Salvador referred to the “oppression we all suffer from the gentiles of Angola who live in Palmares.”72 But, within the context of Angolan history, what is the significance of that connection for the history of Palmares? The kingdom of Ndongo, which the Portuguese came to call Angola in the late sixteenth century was a land in turmoil, invaded from the coast by the Portuguese and from the interior by bands of marauding warriors from central Africa.73 The dissolution of the old Kingdom of Kongo and the Lunda state in Kitanga created a period of military struggle and disruption that destroyed 71  The absence of the word quilombo in reference to early fugitive communities is noted in Kent, “Palmares,” 162–63. See also Mario José Maestri Filho, “Em torno ao quilombo,” História em Cadernos, (Mestrado de História-Universidade Federal do Rio de Janeiro), 2:2 (1984), 9–19. 72  Livro das Atas da Câmara do Salvador (1669–84), 23, cited in Pedro Tomás Pedreira, “Os quilombos dos Palmares e o senado da câmara da cidade do Salvador,” M.A.N. 11:3 (1980), 14–17. 73  David Birmingham, Trade and Conflict in Angola (Oxford, 1966), presents a detailed description of these events. See also his synthesis in “Central Africa from Cameroun to the Zambezi,” in Cambridge History of Africa, eds. J. D. Fage and Roland Oliver 5 vols. to date (Cambridge, 1975–), 4, (1975), 325–83. See also Jan Vansina, Kingdoms of the Savanna, (Madison, 1968), 64–70, 124–38.

1322

Schwartz

villages and uprooted peoples. Powerful groups of uprooted warriors calling themselves “Imbangala” or “Yaka” and called by the Portuguese, “Jaga,” swept into present-day Angola, disrupting existing states and eventually creating a series of new polities.74 The precise origins and cultural traditions of the Imbangala and even the relationship of the designations Jaga, Imbangala, and Yaka have been a matter of debate among Africanists for some time, but some aspects of Imbangala/ Jaga society were noted by contemporary observers that are of direct interest to historians of slave resistance in Brazil, and especially to those interested in Palmares.75 First, the Imbangala raiders lived on a permanent war-footing. Reportedly, they killed the babies born to their women, but integrated adopted children into their ranks so that over time they came to be a composite force of large numbers of people of various ethnic backgrounds united by an organized military structure. That organization and a reputed military ferocity made them the scourge of the region, highly effective and greatly feared. Imbangala-Portuguese relations were alternately hostile and friendly. Between 1611 and 1619, Imbangala lords served as mercenaries for the Portuguese governors and supplied a flow of captives to the slave traders at Luanda.76 New states were formed by an Imbangala fusion with the indigenous lineages as the Imbangala conquered or created a number of kingdoms among the Mbundu peoples of the Congo-Angola region. Two of these states were the kingdom of Kasange and the kingdom of Matamba, ruled by Queen Nzinga, with whom the Portuguese first fought until the mid-seventeenth century when an alliance was formed. These states battled each other for control of the Kwango river basin, a struggle that opened up this region to increased slaving.77 As the Imbangala moved southward into Angola in the early seventeenth century they encountered among the Mbundu people an institution that they adopted to their purposes.78 This was the ki-lombo, a male initiation society or 74  Joseph C.  Miller, Kings and Kinsmen: Early Mbundu States in Angola, (Oxford, 1976), contains an excellent description of the Imbangala and their institutions. 75  On the debate, see Joseph C. Miller, “The Imbangala and the Chronology of Early Central African History,” Journal of African History, 13:4 (1972), 549–74; “Requiem for the Jaga,” Cahiers d’etudes africaines 13:1 (1973), 121–49; John Thornton, “A Resurrection for the Jaga,” Cahiers d’études africaines 18 (1978), 223–28. 76  Beatrix Hintze, “Angola nas garras do tráfico de escravos,” Revista Internacional de Estudos Africanos 1:1 (1984), 11–60. 77  See Roy Glasgow, Nzinga (São Paulo, 1982); Joseph C. Miller, “Kings, Lists, and History in Kisanje,” History in Africa 6 (1979), 51–94. 78  Miller, Kings and Kinsmen, 151–75, 224–64, provides an intense analysis. The classic sources on the Jaga ki-lombo are Cadornega and Cavazzi de Montecuccolo.

Rethinking Palmares: Slave Resistance In Colonial Brazil

1323

circumcision camp where young men were prepared for adulthood and warrior status. The Imbangala molded this institution to their own purposes. Torn from ancestral lands and gods, sharing no common lineage, living by conquest and according to European observers rejecting agriculture, the traditional basis of societies in this region, the Imbangala needed an institution that provided cohesion to the disparate ethnic elements comprising their bands. The ki-lombo, a military society to which any man by training and initiation could belong, served that purpose. Designed for war, this institution created a powerful warrior cult by incorporating large numbers of strangers who lacked a common ancestry. The Imbangala ki-lombo was distinctive because of its ritual laws. Lineage and kinship, so important to the other basically matrilineal peoples of the region, were denied within its confines and although European observers spoke of infanticide, strictly speaking, women could leave the confines of the ki-lombo itself to bear their children. What was prohibited was a legal matrilineal link within the ki-lombo that might challenge the concept of a society structured by initiation rather than by kinship. Historian Joseph Miller believes that the Imbangala “killing” of their own children was a metaphor for the ceremonial elimination of kinship ties and their replacement with the rules and proscriptions of the ki-lombo. The creation of a social organization based on association created risks. The inhabitants of the ki-lombo stood in a special spiritual danger since they lacked the normal lineage ancestors who might intercede with the gods on their behalf. Thus a chief figure in the ki-lombo was the nganga a nzumbi, a priest whose responsibility was to deal with the spirits of the dead. The Ganga Zumba of Palmares was probably the holder of this office, which was in effect not a personal name but a title. There are other echoes from the descriptions of Angola that seem suggestive. In the Imbangala, quilombo leadership depended on some kind of popular acclaim or election just as some of the Brazilian accounts suggest.79 Most curious is the observation of Andrew Battell, who lived among the Imbangala and who noted that their chief luxury was palm wine and that their routes and camps were influenced by the availability of palm trees. His comment makes the association of the maroon community with a region of Palmares (the word means palm trees) seem more than coincidental.80 If the founders of Palmares had used the Imbangala ki-lombo as the basis for their society, their version of it was incomplete or at least a variation on the basic model. A number of features associated with the Imbangala ki-lombos had 79  Cadornega, História, II, 221. 80  E. G. Ravenstein, The Strange Adventures of Andrew Battel of Leigh in Angola and Adjoining Regions, (London, 1901).

1324

Schwartz

no parallel in Brazil. The Imbangala were always referred to as cannibals who practiced cannibalism and human sacrifice to terrorize their enemies. These practices were strictly controlled as was the preparation of magi a samba, a paste made from human fat and other substances which supposedly made the ki-lombo warriors invincible. A strict set of ritual laws (kijila) surrounded the ki-lombo. Women were prohibited from the interior compound of the ki-lombo and there were strict ritual proscriptions against menstruating women. None of these customs are mentioned in the surviving documentation on Palmares. The use of the term quilombo in reference to Palmares does not necessarily mean that all the ritual aspects of that institution as they were practiced in Angola were present in Brazil or that the founders or the subsequent leaders of Palmares were necessarily Imbangala.81 Many aspects of the Imbangala kilombo could be found in other Central African institutions like the kimpasi, secret initiation camps of the Kongo, which also created new social bonds by association.82 Much of what was inherent in the ki-lombo would have been understood by non-Imbangala. As noted, Imbangala dynasties and institutions were incorporated in a number of Mbundu states, and the quilombo came to symbolize the sovereignty of these states. Our best source in this regard is Antônio de Oliveira de Cadornega, the principal chronicler of seventeenthcentury Angola. Cadornega used the term quilombo to describe Jaga troops, quilombos de Jagas, or gente e quilombos de Jagas, but also as a descriptive term for the kingdoms of Matamba and Kasange.83 The use of the phrase “kingdom and quilombo” of Matamba was a general descriptive use of quilombo that referred to these Imbangala influenced polities but did not necessarily suggest the full existence of the original institution nor its ritual practices. Quilombo was becoming a synonym for a kingdom of a particular type in Angola. Given the poor documentary record of Palmares much of the above hypothesis is admittedly tenuous, but I believe there is enough evidence to suggest that the introduction of the term quilombo into Brazil in the late seventeenth century was not accidental and that it represented more than simply a linguistic 81  The Jaga origins of Palmares have concerned scholars for many years. Nina Rodrigues was content to point out the Bantu origin of the titles, personal names, and toponymy of Palmares. M. M. de Freitas argued that since the Palmarinos were such inveterate warriors they must have been Jagas. Freitas argued “that the first runaway was of the sacred caste of the Jagas and the founder of the Palmares’ dynasty, (o primeiro quilombola era da casta sagrada dos jagas e o fundador da dinastia palmarina), (1,278). Kent, “Palmares,” argued that the Jagas were probably not the originators of Palmares. 82  John K. Thornton, The Kingdom of Kongo (Madison, 1983), 61, 107. 83  Cadornega, História, 1, 89; 2, 222.

Rethinking Palmares: Slave Resistance In Colonial Brazil

1325

borrowing. If true, then we must deal with the African aspects of Palmares not as “survivals” disembodied from their original cultural milieu, but a far more dynamic and perhaps intentional use of an African institution that had been specifically designed to create a community among peoples of disparate origins and to provide an effective military organization. Surely, the fugitive slaves of Brazil fitted such a description, and the attacks made upon them by colonial governments made the military organization of the quilombo essential for survival. The success of quilombos varied as greatly as the quilombos themselves in size, leadership, longevity, and internal organization. Taken together, Palmares and the smaller fugitive communities constituted a continuous commentary on the Brazilian slave regime.

The “Second Slavery”: Bonded Labor and the Transformation of the Nineteenth-Century World Economy Dale Tomich The abolition of the slave trade and slavery in the Western Hemisphere are certainly among the most significant and dramatic occurrences of the nineteenth century. By means as diverse as legislation, revolution, and civil war, slavery and the slave trade were eradicated in a sequence of events beginning with the Haitian Revolution in 1791 and extending to slave emancipation in Brazil in 1888. Indeed, the strength and effectiveness of anti-slavery thought and action contributed importantly to the nineteenth century’s self-consciousness as a period of the growth of human freedom and moral and material progress (see, for example, Davis 1984). During this period, slavery came to be understood as the antithesis of the emergent forms of polity, moral sensibility, and economic activity: it formed the negative standard against which the new forms of freedom were defined. In the elaboration of political economy, for example, the stark contrast between the African slave and the free worker of Europe and North America was used to illuminate what was distinctive about capitalist property relations, wage labor, and industrial production (Smith 1976, 90, 413; Marx 1976, 1031–1034). Slavery was not treated as simply one form of human labor among others; rather, it came to be conceived as the polar opposite of free (wage) labor. It was viewed as the epitome of archaic, backward, and inefficient production and was generally presumed to be incompatible with the emerging modern world, while free (wage) labor was regarded as the universal outcome of the historical processes of capitalist development. The presumption that slavery is incompatible with the modern world has persisted into the twentieth century. Within this framework, scholarly debate has focused on whether material or moral factors were more important in its demise. Whichever interpretation is favored, the abolition of slavery generally has been understood in one of two ways. One view, which emphasizes the role of Britain as harbinger of a modern political, economic, and ideological order, is akin to a type of domino theory. Once Great Britain abolished the S ource: Tomich, Dale, “The ‘Second Slavery’: Bonded Labor and the Transformation of the Nineteenth-Century World Economy,” in Dale Tomich, Through the Prism of Slavery: Labor, Capital, and World Economy, Lanham, Md.: Rowmand & Littlefield, 2003, 56–74. Republished with permission of Rowmand & Littlefield. © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_042

The “ Second Slavery ”

1327

slave trade, the fate of African slavery in the Americas was sealed, and forces were set in motion that slowly, unevenly, but inevitably dismantled the peculiar institution throughout the hemisphere. (It should be noted that in such accounts the revolution in Saint Domingue and the founding of Haiti are apparently aberrations that cannot be assimilated into the narrative and are ignored [see Trouillot 1995, 70–107]). The other view emphasizes the national histories of the various slave societies of the Americas. In this perspective, the “internal” contradictions of slavery become heightened, and in case after case, slave relations give way to a higher form of economic rationality. Both interpretations assume the singularity of slavery. Slavery is seen or presumed to be essentially the same phenomenon everywhere, and different slave systems are distinguished from one another only by their economic, cultural, and political contexts. As a result, the abolition of slavery, whether considered in its international connection or in its various national arenas, is regarded as a unilinear transition from archaic to modern forms of economy. It is this assumption of the singularity of slavery that I wish to examine here. My purpose in this chapter is to call attention to the changing character of slavery in the nineteenth century world economy. (It should be noted that I do not propose to avoid explaining the causes of slave emancipation which remain diverse and conjunctural.) This chapter demonstrates the formation and reformation of slave relations within historical processes of the capitalist world economy. If slavery was ultimately abolished everywhere in the hemisphere, the “anti-slavery century” was nonetheless the apogee of its development. Beneath the apparent uniformity of nineteenth-century slave emancipation, we find complex and differentiated trajectories and outcomes that are traceable to the position of particular slave systems within the world economy. At the end of the eighteenth century, the Caribbean sugar industry, particularly the British and French colonial empires, was at the center of slave production in the world economy. Slavery in non-sugar areas was moribund. Rice, indigo, and tobacco were of secondary importance in the North American slave colonies, Brazil was at the end of its “gold cycle,” and in the Spanish Empire, with the exception of Cuba, slavery was marginal. Yet, during the course of the nineteenth century, slavery expanded on a massive scale precisely in these relatively backward areas in order to supply the growing world demand for cotton, coffee, and sugar. At the same time, the former centers of slave production declined. This second cycle of slavery was initiated by the rise of British hegemony and declined with the challenges to it as U.S. economic and political preeminence grew in the Western Hemisphere and the depressions of the 1870s and 1880s triggered crises of colonial commodity markets. Examination of these concomitant processes of decline and expansion at once emphasizes

1328

Tomich

the complex ways that slavery is imbricated in world economic processes and the intrinsic and irreducible unevenness of capitalist development.

British Hegemony and the New International Division of Labor

The changes in the world economy during the first half of the nineteenth century fundamentally altered the parameters and conditions of production in the periphery. These changes included not only the spatial redistribution and quantitative increase in the production of tropical products but also the qualitative restructuring of the social relations and processes organizing the world market. Previously, the relations between core and periphery had been constituted by competing colonial empires. Each metropolitan power maintained an exclusive sphere of production in its colonies. The division of labor between metropolis and colony and the nature and direction of commodity flows were defined through politically enforced monopolies, privileges, and restrictions determined in the metropolis. Each metropolis reserved for itself the produce of its own colonies, monopolized colonial shipping, and used the colonies as a sheltered market for its industry. By means of these mercantilist policies, rival nation-states forcibly expanded their markets, stimulated production, and promoted the accumulation of national wealth. This system expressed not only the limits of commodity production and exchange but also the weak integration of the world market during this period. Within the framework of this market structure, world production of tropical staples grew steadily but slowly, and colonial producers were relatively insulated from direct competition with one another by their reliance on the political conditions of their monopoly of the metropolitan market. However, this form of the organization of the world economy broke down between 1780 and 1815, and the emergence of British economic and political hegemony signaled the beginning of the structural transformation of the world market. The market was no longer constituted through the direct political domination over the sources of colonial production. Rather, the key to power under the emerging conditions of world economy was economic control over the flow of commodities. The nexus of direct colonial control was broken, and the system of imperial preference broke down. Increasingly a more or less self-regulating market, the political conditions for which were maintained and established by the power of the British state, became the mediation between producers and consumers, and supply, demand, and price appeared as the determinants of the division of labor and flow of commodities on the world market (McMichael 1984, 1–31).

The “ Second Slavery ”

1329

This restructuring of the world market was underpinned by processes of industrialization, urbanization, and population growth. While the rate of economic advance during the first half of the nineteenth century was slow compared with that of the second half, the advance of industrialization in Europe and North America changed the pattern of demand in the world market over the course of this entire period. Modern industry required new raw materials on an unprecedented scope and scale, while the growth of population and development of predominantly urban middle and working classes in Europe was associated with new patterns of consumption that increased Europe’s dependency on peripheral producers for foodstuffs. While its relative importance declined, in absolute terms sugar remained a key item of world trade. Alongside it, cotton and coffee assumed new commercial importance. The production and consumption of these articles increased on a massive scale over the course of the century and their importance grew. European and North American capital extended the area of supply of these materials and established a global system of transport based on the railroad and steamship to carry them, while, increasingly, peripheral regions specialized in their production. New poles of economic attraction were created between core and periphery that did not coincide with the old colonial boundaries. World trade and crop cultivation increased dramatically though unevenly throughout the century, and world rather than local prices increasingly dominated the trade in agricultural staples and raw materials (Hobsbawm 1968, 128; Woodruff 1971, 657–658; McMichael 1984, 12–13). The chief agent and beneficiary of this transformation was Great Britain. With the collapse of France and its colonial empire after 1815, there was no power that could rival Britain in the international arena, and a process of reintegration of the world market began under the hegemony of British capital. Britain’s position was not due simply to technological superiority. Rather, British commercial, financial, and maritime supremacy sustained its industrial development; and, in turn, Britain’s productive advantage over its rivals widened, and its control over the market was strengthened. The control of international finance by London and the creation of the new financial institutions of the City represented new levels of integration of the world economy and new channels for economic domination. Britain emerged as the keystone of international trade. World production and consumption were progressively shaped around the conditions imposed by the requirements of British capital accumulation and integrated into its rhythms and cycles. Although the thrust of British economic activity toward the periphery became more pronounced after 1857, British economic development became increasingly dependent upon trade with the periphery, especially Latin America

1330

Tomich

and India, for industrial raw materials, foodstuffs, and, to a lesser extent, as an outlet for manufactured goods and the investment of accumulated surpluses during the first half of the century as well. British imports increased fourfold between 1780 and 1850. The amount of raw cotton used by British industry increased nearly fivefold between 1800 and 1830, and by 1831 it supplanted sugar as the country’s leading import. On the other hand, the growth of cotton manufactures in Britain depended not on domestic consumption but on markets in the periphery. Particularly important in this regard was Latin America. Brazil constituted the largest single market for British cotton exports during the first half of the century (Hobsbawm 1968, 58, 135, 138, 146–148; Woodruff 1971, 662–663). To the extent that Britain came to control commerce outside the bounds of its own empire, it became less committed to formal colonialism as the means of defining the nature and direction of commodity flows and the division of labor between core and periphery. Instead, British commercial and industrial superiority enabled it to penetrate the markets of the other colonizing powers, and to establish trade with the periphery on the basis of complementarity— British manufactured goods (and other services such as capital, shipping, banking, and insurance) for peripheral raw materials and agricultural products. It was the world’s largest consumer of products from the periphery and was the only purchaser that could absorb increased peripheral production. On the other side of the coin, it was the only country that could supply the credit, machinery, and manufactured goods required to support this expansion. The advance of British industry accentuated the relative differential between industrial and agricultural prices in the world economy and pushed Britain to develop cheap new sources of supply in the periphery in order to redress its unfavorable balance of trade. The result was growing European demand for tropical and sub-tropical food products, among them sugar, cotton, and coffee, and increased international specialization in production of food and raw materials. The dramatic fall in price of these commodities benefited Britain more than any other country (Hobsbawm 1968, 135, 138, 146–148). The establishment of this division of labor between core and periphery was organized by the City of London whose position as the center of world trade was both instrument and expression of British hegemony. The extension of commodity production in the periphery and the expansion not only of British trade with the periphery but also of its rivals relied upon the financial power of London banks. As McMichael has argued, British loan capital extended the scope of the world market for all states. A system of multilateral trading emerged that depended upon sterling balances and the credit of London banks as well as the City’s ability to settle trade balances among states

The “ Second Slavery ”

1331

indirectly. Bills of exchange drawn on London banks replaced transfer of precious metals in organizing international exchanges, and sterling balances were used to adjust the status of national currencies in world trade. The centralization of banking enabled Britain to maintain and extend world exchange and to achieve financial supremacy beyond its commercial and industrial supremacy (McMichael 1984, 12, 21–23, 26–27). The creation of these global exchange relations centered on Britain established a world division of labor dependent upon and responsive to an integrated world market. Within this new configuration the conditions of slave labor in the world economy were altered.

The Transformation of American Slavery

The effect of these developments was not to destroy archaic forms of social organization and establish the general mobility of capital and labor in a universal free market. Rather, previously existing social relations were recast within the new constellation of political and economic forces. The prior interdependence of colonialism and slavery was broken up, and the conditions of existence, function, and significance of each were modified. The rise of British hegemony and the industrial revolution in Britain restructured the world division of labor and stimulated the material expansion of the world economy. These developments not only created the conditions for the destruction of slavery within the British Empire but also encouraged the expansion and intensification of slavery outside of it. This “second slavery” developed not as a historical premise of productive capital, but presupposing its existence and as a condition for its reproduction. The systemic meaning and character of slavery was transformed. The emerging centers of slave production were now increasingly integrated into industrial production and driven by capital’s “boundless thirst for wealth.” The British West Indies were the initial beneficiaries of the rise of British hegemony and the destruction of the French colonial sugar industry. Between 1791 and 1815, sugar production in the British Caribbean rose more rapidly than at any other time in its history. (See chapter 4.) The old colonies increased their output and new sugar territories were added to the empire. However, the impact of the transformation of the sugar market was felt differently among the various British colonies. The small islands of the British Lesser Antilles, exploited intensively since the seventeenth and eighteenth centuries expanded rapidly during these years. However, by the late 1820s, they reached the physical and technical limits of expansion and went into a period of decline. In Jamaica, on the other hand, there was room for territorial expansion and new investment.

1332

Tomich

However, this expansion was into the inland valleys of the island, and expensive overland transportation raised the price of the Jamaican product. Slave production in the British West Indies had been geared to the pre-industrial organization of the world economy and was dependent upon now outmoded monopolies. The productivity of slave labor could not be increased, and new supplies of slaves and land were not available. Despite the gains made during this period, the expansion and intensification of sugar production pushed the older colonies to their limit, and they were surpassed by newer producers in an expanding market. The alternative was to start slave production in new areas. Britain further expanded her sugar cultivation by bringing the islands conquered in the Seven Year’s War—Grenada and the Grenadines, Saint Vincent, Trinidad, and Tobago—under cultivation as well as the territories of Berbice, Demerara, and Essequibo newly acquired from the Netherlands. However, if these new areas of slave production were maintained within the old colonial nexus of monopoly and restriction, surplus production in the less productive older colonies would be subsidized within anomalous social relations. Whereas, if a labor force could be guaranteed by other means in the new production areas they would still prosper, and the market mechanism of accumulation and expanded reproduction of the world economy would be freed. With the centralization of world trade in its hands, there was no need for Britain to secure its own labor supply for tropical and semi-tropical goods. Its concern was increasingly for cheap commodities without regard to the form of labor that produced them. The slave as productive labor took precedence over the slave as commodity. These structural shifts in the world economy contributed to the efficacy of the anti-slavery movement in Britain. However, the movement against the slave trade was not simply a function of economic factors, but added another dimension to the processes leading to the destruction of slavery and forced different paths of development on British and non-British colonies. The abolition of the slave trade not only cut off the labor supply to the British slave colonies but, as Paula Beiguelman has indicated, destroyed the commodity market most intimately linked to the slave form (Beiguelman 1978, 71–80). British efforts to suppress the slave trade internationally formed a counterpoint to the expansion of slavery in Cuba, the United States, and Brazil during the nineteenth century. Slavery in these areas developed in the face of the restriction and ultimately the elimination of the slave trade as well as against a mature, widespread, and successful abolitionist movement. The interplay of market forces and the anti-slavery movement pushed Britain toward a policy of free trade and undercut the competitive position of its West Indian colonies. By 1815, British expansion in the Caribbean came

The “ Second Slavery ”

1333

to an end. With the exception of Jamaica and Guiana, British Caribbean sugar production had reached its saturation point. British West Indian planters were faced with rising costs and the inability to expand production. The abolition of the slave trade had cut off their labor supply, while British slave emancipation disrupted the production in both the decadent and the vital slave colonies alike. The high price of British West Indian sugar and the protective tariffs it required restricted British domestic consumption and seriously weakened Britain’s position in the re-export market. As cheaper American sugar invaded European markets, re-exports of British West Indian sugar had to be subsidized by the government by means of drawbacks and export bounties in order to make them competitive with sugar from countries that were often substantial customers for British manufactures. The British West Indian sugar colonies found themselves unable to compete in an expanding world economy. Their position was undercut not by British industry but by the more efficient production in the new American slave zones and elsewhere in the periphery. The very processes that contributed to the destruction of slavery within the British Empire resulted in the intensification of slave production elsewhere in the hemisphere. The demand for cotton, coffee, and sugar reached unprecedented proportions during the nineteenth century, and the production of these crops revitalized slavery in Cuba, the United States, and Brazil as part of this emergent capitalist international division of labor. This is reflected in the scale and nature of slave production itself. Vast expanses of land were opened up, and millions of slaves set to work producing these crops. New industrial technology—notably the railroad, steamship, and steam mills—transformed the labor process in the new slave frontiers. Behind this expansion was the power of British capital and state organizing the world market and international division of labor. In the first half of the nineteenth century, London provided what Jenks has aptly described as a “credit bridge” to the United States, Brazil, and the Spanish colonies in order to stimulate production and trade (Jenks 1971, 67; McMichael 1984, 22–23). By themselves or in conjunction with foreign, especially North American, banks and merchant houses, London financial institutions provided capital through credit and direct investment for the development of plantations and the railroads and banks supporting them. The development of these new plantation zones lowered costs and increased the scale of production as well as providing outlets for surplus British capital in one form or another. British credit also expanded world trade and increased the demand for new crops outside of Britain. For example, merchants and bankers in New York, Boston, and Philadelphia could use the trade surplus from cotton exports to draw bills on London banks in order to finance purchases of not only British manufactures but also significant amounts of

1334

Tomich

Brazilian coffee and West Indian sugar. While these activities helped build up American finance, commerce, and industry, such multilateral trading extended British financial power, increased the volume of commodities in circulation, developed cheap sources of supply, and secured markets for British manufactures throughout the world economy. It was within this complex of world production and exchange that the “second slavery” emerged. Cuban sugar production increased sharply in scale after the British occupation of Havana (1762), which brought with it increased imports of slaves. However, it catapulted forward from the 1790s onward aided by the destruction of export production in Saint Domingue and the decline of the British West Indian sugar industry. By 1830, Cuba emerged as the world’s largest producer with an output of 104,971 metric tons. World demand continued to grow at an accelerating rate, and Cuban production more than kept pace with it. By 1848, the 260,463 metric tons produced accounted for nearly one-quarter of the world’s supply (Moreno Fraginals 1978, I, 46–47, 67–71, 95–102, 167–255; II, 93–97, 106–174; III, 35–36; Scott 1985, 10; Guerra y Sánchez 1964, 40–45, 52–53; Knight 1970, 14–18, 40–41). This expansion of production in Cuba was accompanied by and dependent upon a dramatic increase in the slave labor force. In Cuba, slave imports rose drastically, and the demographic composition of the island was transformed despite British attempts to put an end to the slave trade. Hubert Aimes estimates that 400,000 slaves were imported into Cuba between 1762 and 1838. Cuba’s slave population went from 85,900 in 1792 to 199,100 in 1817, and in the latter year over 32,000 slaves were imported. By 1827 the slave population in Cuba reached 286,900, and in 1841 it stood at 436,500 and accounted for more than 43 percent of the total population. The slave trade continued to supply Cuba with fresh manpower until the middle 1860s, and the slave population in 1862 was some 368,550 (Curtin 1969, 34; Knight 1970, 10–11, 22–23, 41; Scott 1985, 87). The development of the sugar industry was centered in the western part of the island. Sugar cultivation spread south and west of Havana displacing coffee and tobacco producers and spreading onto new lands. New and everlarger plantations were established at a frenetic pace and old ones increased their capacity. The number of ingenios increased almost fourfold between 1800 and 1857. During the initial stages of expansion the multiplication of traditional production units accounted for much of the increase in total production. However, steam power made an early appearance, and the methods of sugar manufacture in Cuba were transformed by the application of modern industrial techniques. According to Knight, in 1827 only 2.5 percent of the 1,000 ingenios in Cuba were steam powered. But by 1860, 70.8 percent

The “ Second Slavery ”

1335

of the 1,365 ingenios used steam. Railroads opened new lands and, in contrast to the situation in Jamaica, permitted profitable exploitation of the interior by the sugar industry. In 1837, thirteen years after the first steam-driven railway began to operate in England, the first railroad in Latin America or the Caribbean was completed between Havana and Güines. This line, created to serve the sugar industry, was only 51 miles long, but it was successful and soon railroads were in operation in all sugar growing areas of the island. Shipping costs were reduced drastically, and land use was maximized (Guerra y Sánchez 1964, 66; Knight 1970, 32–33, 35–36, 38; Scott 1985, 21–24). Cubans enjoyed the technological edge of latecomers. Though they were few in number, the appearance of mechanized sugar mills represented a qualitative transformation in the conditions of sugar production. The Cuban sugar mill developed on a giant scale, and the technology of sugar production there attained the most advanced level known under slavery. Steam-powered mills and the vacuum pan increased the capacity of the more advanced plantations and produced more and higher-quality sugar. On large estates small rail lines were introduced, often using animal-drawn equipment, to transport cane to the mills from the fields, for transportation within the factories and to the wharves. These developments broke the fixed ratio between land, labor, and mill capacity that had limited the development of the old ingenio. With the introduction of steam-driven machinery, it was no longer necessary to limit the acreage under cane. The use of rail transport within estates allowed a greater area to be planted which provided the increased supply of cane required by modern refining techniques. The scale of production increased, and sugar mills could grow in size. The capital requirements for founding an ingenio increased enormously. With the introduction of estate railways, there was bitter competition for land and labor. Small producers were squeezed out, and a monocultural economy emerged that was dominated by large planters who could afford the increased costs of the new mechanized mills. The foundations were established for Cuba’s position as the world’s sugar bowl (Knight 1970, 18–19, 30–40; Guerra y Sánchez 1964, 54, 66; Moreno Fraginals 1978, I, 167–255; II, 93–97, 106–174; III, 34–36; Scott 1985, 20–21). Similarly the growing industrial demand for cotton, the new industrial raw material par excellence, revived slavery in the United States. Between 1780 and the mid-nineteenth century, annual cotton consumption by British mills increased from 2,000 to 250,000 tons. This demand was met through the development of a zone of supply in the American South. Cotton had played a relatively unimportant role in colonial North American agriculture, and supplies of cotton to Britain were of little commercial significance before the 1790s. In the year 1793, the first year of the cotton gin, 487,000 pounds

1336

Tomich

were sent to England. Cotton exports reached 62,186,081 pounds by 1811, the eve of the war with England. From the close of the War of 1812 to 1859 the amount of cotton produced in the United States increase by over thirtyfold from less than 150,000 bales to 4,541,000 bales. Cotton became the leading American export commodity, By 1860, it accounted for more than half the nation’s total exports of $400 million, and had attained an unprecedented position in the American economy (Cohn 1956, 88; Gray 1958, II, 691; Scherer 1969, 125–126, 142–146; Furtado 1968, 112–113; Crawford 1924 126, 129–131, 137–139). The growing demand for cotton opened a vast territory of virgin land to commercial agriculture and the profitable use of slaves. The invention of the cotton gin by Whitney enabled short staple cotton that could be grown in frost areas to be grown on a commercial scale. Up until the Civil War cotton cultivation pushed rapidly into the fertile lands of Alabama, Mississippi, Louisiana, Arkansas, Texas, and Florida. To keep pace with growing world demand, less fertile lands in the Piedmont were abandoned in favor of the ideal climate and soil conditions of the lower South. This migration was less an escape from soil exhaustion than, as Wright argues, “a rational process of geographical expansion and relocation.” Indian lands were occupied and the plantation system revived by vast tracts of new land as the cotton kingdom shifted its center from the Upper South to the new, fertile lands of the Gulf States. The steamboat and railroad facilitated the development of cotton culture in the Lower South. The steamboat brought modern transportation methods to the navigable rivers and coastal waters, and from 1845 to 1860 the South built more miles of railroad than the New England and the Middle Atlantic States combined. This extensive water and rail transportation network opened new lands to cotton cultivation, linked the South to the markets for its goods, and reduced the cost of Southern living by providing a plentiful supply of Western foodstuffs. By 1850, the “Cotton Belt” extended for more than a thousand miles from South Carolina to the region near San Antonio and dominated world production (Cohn 1956, 90, 95, 108–109; Wright 1978, 13, 15–17, 325; Gray 1958, II, 691; Scherer 1969, 202–204, 336–339). Slave labor provided sufficient cheap manpower to support this expansion. Slavery and cotton marched hand in hand across the map of the South. The slave population of the South increased from about one million at the beginning of the century to four million on the eve of the Civil War. Gray estimates that slave imports into the United States between 1800 and 1860 amounted to about 320,000 slaves, of whom some 270,000 were smuggled in after the U.S. abolition of the slave trade in 1808. The majority of the labor supply was thus provided by the remarkable natural increase of North American slaves unique among New World slave populations and the regional shift of slaves from

The “ Second Slavery ”

1337

the North and Upper South to the Cotton Belt. Nevertheless, the availability of slave labor curbed the expansion of the cotton industry. The demand for slave labor drove the price of a good field hand from $200 in 1790 to $2,000 in 1860. The cotton plantation did not require the heavy capitalization of the sugar estate, but these high slave prices favored the larger productive units. Big planters of the Gulf States could absorb these high prices. They not only had larger, more productive estates on more fertile land and easier access to credit on better terms, but they worked large gangs of slaves and could more easily reproduce their labor force without recourse to market. They were thus able to increase their advantage at the expense of small planters. The cotton industry absorbed almost the entire increase in the slave population, and the vast majority of slaves were concentrated in the Cotton Belt by the eve of the Civil War (Cohn 1956, 90–91, 116–117; Scherer 1969, 149–150, 199; Gray 1958, II, 691; Furtado 1968, 124, 127–128n). Coffee was also transformed into an article of mass consumption during the nineteenth century, and Brazil emerged as the new center of world production. The progress of coffee cultivation in Brazil was slow until it began to be exported as a result of short supply and high prices caused by the Haitian revolution. But by the decade 1821–1831, coffee was already Brazil’s third leading export after sugar and cotton and accounted for 18 percent of exports by value. Between 1837 and 1878, Brazil’s annual coffee exports climbed dramatically from one million to four million sacks, and by 1881, coffee represented more than 60 percent of Brazil’s exports by value (Prado 1981, 157–159; Furtado 1963, 121–124). The development of the coffee plantation in Brazil relied upon the intense utilization of slave labor. The equipment used was simpler than in sugar manufacture and, more often than not, of local manufacture. Costs were low, and if there was sufficient land, the availability of labor was the only obstacle to its growth. Curtin estimates slave imports to Brazil during the nineteenth century at 1,145,000. However, the high rate of slave mortality reduced the impact of new Africans on the economy. Beyond this, the coffee economy had available to it, first, the underutilized slave manpower reservoir in the former mining region, and, later, transfers of manpower from the sugar industry in the Northeast. In 1823, the provinces of Rio de Janeiro, Minas Gerais, and São Paulo had a slave population of about 386,000 while Bahia, Pernambuco, and Maranhao, sugar and cotton regions of the Northeast and North, had approximately 484,000. In the next fifty years the development of the coffee industry reversed this situation. Bahia, Pernambuco, and Maranhao had about 346,000 slaves while the coffee provinces of Rio de Janeiro, Minas Gerais, and São Paulo counted over 800,000 (Curtin 1969, 234–237; Furtado 1963, 124, 127–128n).

1338

Tomich

The initial expansion of coffee cultivation in Brazil was concentrated in the Paraiba Valley, which remained its center until the third quarter of the nineteenth century. Close to Rio de Janeiro, this area enjoyed a relatively abundant supply of manpower because of the decline of the nearby mining economy, while abundant mule convoys provided easy transport to the nearby port. However, as the boom continued, the railroad aided in pushing coffee cultivation further inland. The Estrada de Ferro Central do Brasil (Dom Pedro Segundo), built by the Brazilian government, but with the aid of British loans and by a British contractor, opened the further reaches of the Paraiba Valley, parts of Minas Gerais, and finally the Oeste Paulista to coffee cultivation. The effect of this line on the economy of the Paraiba Valley was not long lasting because the valley had almost reached its peak of productivity by the time the railway was constructed. But it served to prolong its prosperity by lessening the cost of transport, and also it contributed to the intensification of coffee production in the upper Paraiba Valley in the province of São Paulo (Furtado 1963, 123–124; Prado Junior 1981, 161–162; Graham 1968, 52–54). Land speculation and speculative agricultural practices pushed coffee cultivation south and west into the state of São Paulo. The Oeste Paulista (geographically the east and northeast of the province of São Paulo, but in terms of the spread of coffee to the west of the Paraiba Valley) enjoys natural advantages and greater ease of transport and communication over Paraiba Valley. The irregular and varied terrain of the Paraiba Valley and the dispersion of hillsides with proper exposure to sun and shelter from wind scattered the coffee estates in small nuclei separate from one another. Credit and manpower were in relatively short supply as well, and plantations did not exceed a few tens of thousands of plants at most. In the new region of São Paulo, the unbroken terrain was covered with a uniform and uninterrupted “sea of coffee” that covered the landscape as far as the eye could see. Large estates began to form with greater frequency in these new zones of the Oeste Paulista. From early on plantations were formed with hundreds of thousands of plants, and by the end of the century, estates with more than a million plants began to appear. The scale of production and domination of resources by the big estates and the size of the investment required made these new lands less accessible to small and medium producers. In contrast to the backward techniques that characterized the earlier coffee production in the Paraiba Valley, the Oeste Paulista produced on a greater scale and with greater technical sophistication. These large coffee plantations utilized elaborate modern equipment for processing the coffee and even smaller planters sent their coffee to the towns to be processed mechanically. These large plantations could extend the area under cultivation and use privately owned railroads instead of oxcarts to haul the coffee

The “ Second Slavery ”

1339

to the center of the plantation for processing. The railroad made the exploitation of Oeste Paulista possible, and the railroad in turn depended on coffee. A new network of rail lines centering on the São Paulo-Santos axis lowered freight costs and linked producers with the rapidly growing world demand for coffee. Further, the railroad opened up the vast coffee lands to the west of Campinas and thus ensured the region’s prosperity into the twentieth century (Graham 1968, 45–46, 66–67, 71–72; Prado Junior 1981, 164–166). Conclusion This chapter argues that slave labor and its abolition cannot be seen as a linear evolutionary process, but as complex, multiple, and qualitatively different relations within the global processes of accumulation and division of labor. For the purposes of this discussion, two qualitatively distinct relations of slavery and processes of slave development—each with different roles and meanings— in the nineteenth century world economy can be distinguished. The first was constituted by a specific set of socio-historical processes and played a particular role in the formation of the world economy from the sixteenth through the eighteenth centuries. These relations were either destroyed or radically reconstituted by the nineteenth century transformation of the world economy. The second was created by and within the historical processes and ensemble of social relations specific to the nineteenth century world economy itself. The second slavery consolidated a new international division of labor and provided important industrial raw materials and foodstuffs for industrializing core powers. Far from being a moribund institution during the nineteenth century, slavery demonstrated its adaptability and vitality. Implicit in these two processes of the development of slavery are two processes of emancipation. A full account of the history of the destruction of slavery during the nineteenth century would have to take into consideration a variety of political, social, and ideological factors, not least of which would be the actions of the enslaved themselves. Nevertheless, the transformation of the world economy made the conditions of the existence of slave labor more vulnerable and volatile than previously. Price competition in an expanding world market and the growth of wage labor made the productivity of labor more important. The new zones of slave production no longer monopolized the production of particular commodities but had to compete with other forms of labor organization elsewhere in the world economy as the spectrum of forms of labor control was expanded and a global hierarchy of labor was created. Slave producers had to compete with one another and

1340

Tomich

with other peripheral producers, and their position in the international relations of production were determined by the price of staples. At the same time, price differentials were evened out by industrial production and the integrated world market, and world prices were established. The relations between core and periphery were determined by the opposition of industrial prices versus primary goods prices, and high cost versus cheap labor. Further, the systemic meaning of slavery was transformed with the emergence of the capital-wage labor relation during the nineteenth century. The products of slave labor entered directly into the consumption of the European wage working class on an increasing scale. They were important as a means of maintaining the exchange relation between wage labor and capital and also contributed directly to lowering the cost of reproducing the labor power of wage labor. As the capital-wage labor relation became widely established, a systemic imperative to increase surplus value by reducing the value of labor power emerged which required slave producers to provide cheaper and cheaper goods for working class consumption. At the same time as British hegemony created an integrated world market, the conditions for the production and reproduction of the social relations of capital became “national”: The conditions imposed by the international division of labor occasioned a variety of political responses on the part of the planter classes of Cuba, the American South, and Brazil. Unlike their colonial predecessors, they developed varying degrees and modes of national self-consciousness through which they attempted to consolidate their position in the world economy. The conditions imposed by the international division of labor occasioned a variety of political responses on their part, including at times anti-colonialism, anti-slavery, and even attempts to establish an independent national state. Further, slaveholders had to be entrepreneurs concerned with the productivity of labor. While the Civil War and emancipation truncated the “natural evolution” of slavery in the United States, the search for greater productivity and shortage of labor following suppression of slave trade led the Cuban and Brazilian planter classes to experiment with new forms of labor organization and new sources of labor. Slavery in these areas was complemented by other forms of labor control—indentured labor, wage labor, and peasant labor. The development of these other forms of labor control are conventionally seen as evidence of the dissolution of slavery, but slaves remained the strategic fulcrum of the labor process and other forms were complementary to it. These mixed forms of plantation labor are a testimony to the resiliency and adaptability of slave labor. Such experiences were characteristic of the second slavery and aided planters in Cuba and Brazil in negotiating the transition to post-slave production more successfully than their predecessors.

The “ Second Slavery ”

1341

Bibliography

Primary Sources Manuscript Sources



Fonds Généralités



Fonds Martinique

France. Archives Nationales. Section Outre-Mer (hereafter ANSOM).

Généralités, Carton 43, Dossier 350. Procès-Verbaux des deliberations de la Commission Schoelcher. Généralités, Carton 9, Dossier 99. De Moges, Mémoire a son successeur (1840). Généralités, Carton 144 Dossier 1221. Condition des esclaves. Publication d‘une expose sommaire de mesures prises dans les differents colonies pour l’éxécution de l’ordonnance royale du 4 janvier 1840. 1841–1843.

Martinique, État des cultures, 1831–1850. Martinique, Carton 7, Dossier 83. Tournées du gouverneur. 1829–1851, 1870. Martinique, Carton 11, Dossier 108. Correspondance politique. 1848–1851. Martinique, Carton 11, Dossier 109. Correspondance générale. C. A. Bruat, Gouverneur. Octobre 1848–juin 1851. Martinique, Carton 56, Dossier 464. Correspondance générale entre le Ministre de la Marine et le Gouverneur Rostoland. 1848–1849.



Printed Sources



Secondary Sources

France. Ministère de la Marine et des Colonies. 1840, 1841, 1842, 1843. Commission instituée par décision royale du 26 mai 1840, pour l‘examen des questions relatives à l’esclavage et à la constitution politique des colonies. 2 vols. Paris: Imprimerie Royale. France. Ministère de la Marine et des Colonies. 1844. Exposé général des résultats du patronage des esclaves dans les colonies françaises. Paris: Imprimerie Royale. France. Ministère du Commerce et des manufactures. 1829. Commission formée avec l’aprobation du Roi … pour l’examen de certaines questions de législation commerciale: Enquête sur les sucres. Paris: Imprimerie Royale.

Abbott, A. 1992. “From Causes to Events: Notes on Narrative Positivism.” Sociological Methods and Research 22: 428–455. Adamson, Alan. 1972. Sugar without Slaves: The Political Economy of British Guiana, 1838–1904. New Haven, Conn.: Yale University Press.

1342

Tomich

Aimes, Hubert. 1967. A History of Slavery in Cuba, 1511 to 1868. New York: Octagon Books. Aminzade, R. 1992. “Historical Sociology and Time.” Sociological Methods and Research 22: 456–580. Anonymous (Dr. Collins). 1811. Practical Rules for the Management and Medical Treatment of Negro Slaves in the Sugar Colonies by a Professional Planter. London: Vernor, Hood, and Sharp, Hatchard. Arrighi, Giovanni. 1994. The Long Twentieth Century: Money, Power, and the Origin of Our Times. London: Verso Books. Arrighi, Giovanni. 1998. “Capitalism and the Modern World-System: Rethinking the Non-Debates of the 1970s.” Review XXI, 1: 113–129. Aston, T. H., and C. H. E. Philpin, eds. 1985. The Brenner Debate. Agrarian Class Structure and Economic Development in Pre-industrial Europe. Cambridge: Cambridge University Press. Bastide, Roger. 1978. The African Religions of Brazil: Toward a Sociology of the Interpenetration of Civilizations. Trans. Helen Seba. Baltimore: Johns Hopkins University Press. Bauer, John. 1970. “International Repercussions of the Haitian Revolution.” The Americas XXVI, 4: 394–418. Beckles, Hilary McD. 1987. “‘The Williams Effect’: Eric William’s Capitalism and Slavery and the Growth of West Indian Political Economy.” In Barbara Solow and Stanley L. Engerman, eds., British Capitalism and Caribbean Slavery. The Legacy of Eric Williams. Cambridge: Cambridge University Press, 303–316. Beiguelman, Paula. 1973. Pequenos Estudos de Ciência Politica. São Paulo: Biblioteca de Ciências Sociais. Beiguelman, Paula. 1978. “The Destruction of Modern Slavery: A Theoretical Issue.” Review II, 1: 71–80. Bergad, Laird W. 1989. “The Economic Viability of Sugar Production Based on Slave Labor in Cuba, 1859–1878.” Latin American Research Review XXIV, 1: 95–113. Besson, Jean. 1984a. “Family Land and Caribbean Society: Toward an Ethnography of Afro-Caribbean Peasantries.” In Elizabeth Thomas-Hope, ed., Perspectives on Caribbean Regional Identity. University of Liverpool, Center for Latin American Studies, Monograph Series, no. 11. Liverpool: University of Liverpool, 57–83. Besson, Jean. 1984b. “Land Tenure in the Free Villages of Trelawny, Jamaica: A Case Study in the Caribbean Response to Emancipation.” Slavery and Abolition 5: 3–23. Braithwaite, Edward. 1971. The Development of Creole Society in Jamaica, 1770–1820. Oxford: Oxford University Press. Braudel, Fernand. 1972. “History and the Social Sciences.” In Peter Burke, ed., Economy and Society in Early Modern Europe. New York: Harper Torchbooks, 11–42. Brenner, Robert. 1977. “The Origins of Capitalist Development: A Critique of NeoSmithian Marxism.” New Left Review 104: 25–92.

The “ Second Slavery ”

1343

Cain, P. J., and A. G. Hopkins. 1993. British Imperialism: Innovation and Expansion, 1688– 1914. London: Longman. Cardoso, Ciro Flammarion S. 1979. Agricultura, Escravidão, Capitalismo. Petropolis: Vozes. Charroppin, A. 1848. Du travail libre dans les colonies françaises. Bordeaux: n.p. Cohen, G. A. 1978. Karl Marx’s Theory of History: A Defence. Princeton, N.J.: Princeton University Press. Cohn, D. L. 1956. The Life and Times of King Cotton. New York: Oxford University Press. Conrad, John H., and John R. Meyer. 1964. The Economics of Slavery and Other Studies in Economic History. Chicago: Aldine. Corrigan, Philip. 1977. “Feudal Relics or Capitalist Monuments? Notes on the Sociology of Unfree Labor.” Sociology 11, 3: 435–463. Crawford, M. D. C. 1924. The Heritage of Cotton. The Fibre of Two Worlds and Many Ages. New York: Grosset & Dunlap. Curtin, Philip D. 1969. The Atlantic Slave Trade: A Census. Madison: University of Wisconsin Press. Daubrée, Paul. 1841. Question coloniale sous le rapport industriel. Paris: n.p. David, Paul A., Herbert G. Gutman, Richard Sutch, Peter Temin, and Gavin Wright. 1976. Reckoning with Slavery. New York: Oxford University Press. Davis, David Brion. 1984. Slavery and Human Progress. New York: Oxford University Press. Debbasch, Yvan. 1961, 1962. “Le marronage. Essai sur la désertion de l’esclave antil-lais.” Année Sociologique: 1–112; 117–195. Debien, Gabriel. 1974. Les esclaves aux Antilles françaises (XVIIe–XVIIIe siecles). BasseTerre: Société d’histoire de la Guadeloupe. Fort-de-France: Société d’histoire de la Martinique. De Cassagnac, Adolphe Granier. 1842. Voyage aux Antilles. Paris: n.p. Deerr, Noel. 1945. The History of Sugar. 2 vols. London: Chapman and Hall. De la Cornillère, M. le Conte E. 1843. La Martinique en 1842: Intérêts coloniaux, souvenirs du voyage. Paris: n.p. Derosne, Charles. 1824. Mémoire sur la fabrication du sucre dans les colonies par de nouveaux procédés. Paris: n.p. Derosne, Charles. 1833. Notice on the New Process for Making Sugar, Lately Introduced in the French and English Colonies. Paris: n.p. Derosne, Charles, and Jean François Cail. 1844. De la elaboración del azúcar en las colonias y de los nuevos aparatos destinados a mejorarla. Havana: n.p. Dicken, Peter. 1998. Global Shift: Transforming the World Economy. New York: Guilford Press. Dobb, Maurice. 1947. Studies in the Development of Capitalism. New York: International Publishers.

1344

Tomich

Drescher, Seymour M. 1977. Econocide: British Slavery in the Era of Abolition. Pittsburgh: University of Pittsburgh Press. Dunayevskaya, Raya. 1958. Marxism and Freedom. New York: Twayne Publishers. Dutrône de la Couture, J. E. 1791. Précis sur la canne et sur les moyens d’en extraire le sel essential, suivi de plusieurs mémoires sur le sucre, sur le vin de canne, sur l’indigo, sur les habitations et sur l’état actuel de Saint-Domingue. Paris: Debure, De Seine. Engerman, Stanley. 1973. “Some Considerations Relating to Property Rights in Man.” Journal of Economic History XXXIII, 1: 43–65. Engerman, Stanley. 1976. “Some Economic and Demographic Comparisons of Slavery in the United States and the British West Indies.” The Economic History Review, second ser. XXIX, 2: 258–275. Fogel, Robert W. 1989. Without Consent or Contract. New York: W.W. Norton & Co. Fogel, Robert W., and Stanley L. Engerman. 1974. Time on the Cross. The Economics of American Slavery. 2 vols. Boston: Little, Brown and Company. Franco, Maria Sylvia de Carvalho. 1976. Homens Livres no Ordem Escravocrata. Sao Paulo: Editora Atica. Frank, André Gunder. 1967. Capitalism and Underdevelopment in Latin America. New York: Monthly Review Press. Friedlaender, Heinrich. 1978. Historia Economica de Cuba. Havana: Editorial de Ciencias Sociales. Fukuyama, Francis. 1992. The End of History and the Last Man. London: Hamish Hamilton. Furtado, Celso. 1963. The Economic Growth of Brazil. A Survey from Colonial to Modern Times. Berkeley: University of California Press. Genovese, Eugene D. 1967. The Political Economy of Slavery. New York: Random House. Gilroy, Paul. 1993. The Black Atlantic: Modernity and Double Consciousness. London: Verso. Gisler, Antoine. 1965. L’esclavage aux Antilles françaises (XVIIe–XIXe siecle): Contribution au problème de l’esclavage. Fribourg: Editions universitaires Fribourg Suisse. Goveia, E. V. 1965. Slave Society in the British Leeward Islands at the End of the Eighteenth Century. New Haven, Conn.: Yale University Press. Graham, Richard. 1968. Britain and the Onset of Modernization in Brazil. 1850–1914. Cambridge: Cambridge University Press. Gray, L. C. 1958. History of Agriculture in the Southern United States. 2 vols. Gloucester, Mass.: Peter Smith. [Originally Carnegie Institution of Washington, n.d.] Greenberg, Michael. 1977. “The New Economic History and the Understanding of Slavery: A Methodological Critique.” Dialectical Anthropology 2: 131–141. Griffin, L. J. 1992. “Temporality, Events, and Explanation in Historical Sociology. An Introduction,” Sociological Methods and Research 22: 403–427.

The “ Second Slavery ”

1345

Guerra y Sánchez, Ramiro. 1964. Sugar and Society in the Caribbean. An Economic History of Cuban Agriculture. New Haven, Conn.: Yale University Press. Gutman, Herbert G. 1975. Slavery and the Numbers Game. Urbana: University of Illinois Press. Hall, Douglas. 1961. “Incalculability as a Feature of Sugar Production during the Eighteenth Century.” Social and Economic Studies 10, 3: 305–318. Hall, Douglas. 1978. “The Flight from the Plantations Reconsidered: The British West Indies, 1838–1842.” The Journal of Caribbean History 10–11: 7–23. Herskovits, Melville J. 1937. Life in a Haitian Valley. New York: Knopf. Higman, Barry W. 1976. Slave Population and Economy in Jamaica, 1801–1834. Cambridge: Cambridge University Press. Higman, Barry W. 1988. Jamaica Surveyed. Kingston: Institute of Jamaica Publications, Ltd. Hilton, Rodney, ed. 1976. The Transition from Feudalism to Capitalism. London: New Left Books. Hobsbawm, Eric. 1968. Industry and Empire. Harmondsworth: Penguin Books. Hoogvelt, Ankie. 1997. Globalization and the Postcolonial World: The New Political Economy of Development. Baltimore: The Johns Hopkins University Press. Hopkins, Terence K. 1982a. “The Study of the Capitalist World-Economy. Some Introductory Considerations.” In Terence K. Hopkins and Immanuel Wallerstein et al., eds., World-Systems Analysis. Theory and Methodology. Beverly Hills: Sage, 9–38. Hopkins, Terence K. 1982b. “World-System Analysis: Methodological Considerations.” In Terence K. Hopkins and Immanuel Wallerstein, World-Systems Analysis. Theory and Methodology. Beverly Hills: Sage Publications, 145–158. Ingham, Geoffrey. 1984. Capitalism Divided? The City and Industry in British Social Development. New York: Schocken Books. James, C. L. R. 1963. The Black Jacobins: Toussaint L’Ouverture and the San Domingo Revolution. New York: Vintage Books. Jenks, Leland H. 1973. The Migration of British Capital to 1875. New York: Barnes & Noble. Knight, Franklin W. 1970. Slave Society in Cuba during the Nineteenth Century. Madison: University of Wisconsin Press. Korsch, Karl. 1963. Karl Marx. New York: Russell and Russell. Koselleck, Reinhart. 1985. Futures Past: On the Semantics of Historical Time. Cambridge, Mass.: MIT Press. Kosik, Karel. 1976. Dialectics of the Concrete. A Study of Problems of Man and World. Trans. Karel Kovanda with James Schmidt. Boston: D. Reidel. Kuczynski, J. 1945. A Short History of Labor Conditions under Industrial Capitalism, vol. IV London: F. Muller.

1346

Tomich

Labrousse, C. E. 1954. Aspects de l’évolution économique et sociale de la France et du Royaume-Uni de 1815 à 1880. Paris: Centre de Documentation Universitaire. Laclau, Ernesto. 1971. “Imperialism in Latin America.” New Left Review 67 (May–June): 19–38. Lavollée, P. 1841. Notes sur les cultures et la production de la Martinique et de la Guadeloupe. Paris: Imprimerie Royale. Le Goff, J. 1980. Time, Work, and Culture in the Middle Ages. Chicago: University of Chicago Press. Lepkowski, Tadeusz. 1968. Haiti. 2 vols. Havana: Estudios del centro de documentación sobre America Latina Juan F. Noyala, Casa de las Americas. Lewis, M. G. 1929. Journal of a West India Proprietor. Boston: Houghton Mifflin. Linebaugh, Peter. 1992. The London Hanged: Crime and Civil Society in the Eighteenth Century. Cambridge: Cambridge University Press. Lukàcs, Georg. 1960. Histoire et conscience de classe. Paris: Éditions de Minuit. Malowist, Marian. 1969. “Les débuts du système de plantations dans la période des grandes découvertes.” Africana Bulletin 10: 9–30. Marrero, Levi. 1983–1986. Cuba: Economía y Sociedad. Azucar, Ilustracion y Conciencia 1763–1868. 10 vols. Madrid: Editorial Playor, S.A. Marx, Karl. 1973. Grundrisse. Harmondsworth: Penguin. Marx, Karl. 1976. Capital. Vol. I. Harmondsworth: Penguin. McMichael, Philip. 1984. Settlers and the Agrarian Question: Capitalism in Colonial Australia. Cambridge: The Cambridge University Press. McMichael, Philip. 1990. “Incorporating Comparison within a World-Historical Perspective: An Alternative Comparative Method.” American Sociological Review 55: 385–397. McMichael, Philip. 1991. “Slavery in the Regime of Wage Labor: Beyond Paternalism in the U.S. Cotton Culture.” Social Concept 6, 1 (December): 10–28. McMichael, Philip. 2000. Development and Social Change: A Global Perspective. Thousand Oaks, Calif.: Pine Forge Press. Merrington, John. 1968. “Theory and Practice in Gramsci’s Marxism.” In Ralph Miliband and John Saville, eds., The Socialist Register 1968. London: The Merlin Press, 145–176. Merrington, John. 1976. “Town and Country in the Transition to Capitalism.” In Rodney Hilton, ed., The Transition from Feudalism to Capitalism. London: New Left Books, 170–195. Mintz, Sidney W. 1964. “Currency Problems in Eighteenth-Century Jamaica and Gresham’s Law.” In Robert A. Manners, ed., Process and Pattern in Culture. Chicago: Aldine, 248–265. Mintz, Sidney W. 1974. Caribbean Transformations. Chicago: Aldine. Mintz, Sidney W. 1977. “The So-Called World System: Local Initiative and Local Response.” Dialectical Anthropology 2 : 253–270.

The “ Second Slavery ”

1347

Mintz, Sidney W. 1979. “Slavery and the Rise of Peasantries.” Historical Reflections VI: 213–242. Mintz, Sidney W. 1982. “Descrying the Peasantry.” Review VI, 2: 209–225. Mintz, Sidney W. 1985. Sweetness and Power: The Place of Sugar in Modern History. New York: Viking Penguin. Mintz, Sidney W., and Douglas Hall. 1960. The Origins of the Jamaican Internal Marketing System. Yale University Publications in Anthropology, 57. New Haven, Conn.: Department of Anthropology, Yale University. Moreno Fraginals, Manuel. 1976. The Sugarmill: The Socioeconomic Complex of Sugar in Cuba, 1760–1860. Trans. Cedric Belfrage. New York: Monthly Review. Moreno Fraginals, Manuel. 1978. El ingenio. Complejo económico social cubano del azúcar. 3 vols. Havana: Editorial de ciencias sociales. Ortiz, Fernando. 1970. Cuban Counterpoint: Tobacco and Sugar. New York: Vintage Books. Ortiz, Fernando. 1991. Contrapunteo Cubano del tabaco y el azucar. Havana: Editorial de Ciencias Sociales. Oxaal, Ivar. 1968. Black Intellectuals Come to Power. Boston: Schenkman. Paquette, Robert L. 1988. Sugar Is Made with Blood: The Conspiracy of La Escalera and the Conflict between Empires over Slavery in Cuba. Middleton, Conn.: Wesleyan University Press. Peytraud, Lucien. 1973. L‘esclavage aux Antilles françaises avant 1789 d’après des documents inédits des Archives Coloniales. Pointe-à-Pitre, Guadeloupe: Désormaux. Polanyi, Karl. 1957. The Great Transformation. The Political and Economic Origins of Our Time. Boston: Beacon Press. Prado Jùnior, Caio. 1981. Historia económica do Brasil. São Paulo: Brasiliense. Quadagno, J., and S. J. Knapp. 1992. “Have Historical Sociologists Foresaken Theory? Thoughts on the History/Theory Relationship.” Sociological Methods and Research 22: 481–507. Quijano, Aníbal. 2000. “Coloniality of Power, Eurocentrism, and Latin America.” Nepantla I, 3: 533–580. Ragatz, Lowell. 1971. The Fall of the Planter Class in the British West Indies, 1763–1833. New York: Octagon Books. Renard, Raymond. 1973. La Martinique de 1848 à 1870. Pointe-à-Pitre, Guadeloupe, Groupe Universitaire de Recheches Inter-Caraïbes. Rodney, Walter. 1981. History of the Guyanese Working People, 1881–1905. Baltimore: The Johns Hopkins University Press. Rosdolsky, Roman. 1974. “Comments on the Method of Marx’s Capital.” New German Critique 3: 62–72. Rosdolsky, Roman. 1977. The Making of Marx’s “Capital.” London: Pluto Press.

1348

Tomich

Sainte-Croix, F. Renouard, Marquis de. 1822. Statistique de la Martinique. 2 vols. Paris: Chaumerot, Librairie Palais Royale. Samuel, Raphael. 1977. “The Workshop of the World: Steam Power and Hand Technology in Mid-Victorian Britain.” History Workshop 3: 6–72. Sartre, J.-P. 1982. Critique of Dialectical Reason. London: Verso. Sayer, Derek. 1987. The Violence of Abstraction. The Analytical Foundations of Historical Materialism. Oxford: Basil Blackwell. Scherer, J. A. B. 1969. Cotton as a World Power. A Study in the Economic Interpretation of History. New York: Greenwood Publishing Corp. Schmidt, Alfred. 1983. History and Structure: An Essay on Hegelian-Marxist and Structuralist Theories of History. Cambridge, Mass.: MIT Press. Schoelcher, Victor. 1976. Des Colonies françaises. Abolition immédiate de l’esclavage. Basse-Terre: Société d’histoire de la Guadeloupe. Fort-de-France: Société d’histoire de la Martinique. Schoelcher, Victor. 1847. Histoire de l’esclavage pendant les deux dernières années. 2 vols. Paris: n.p. Schumpeter, Joseph A. 1934. The Theory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest, and the Business Cycle. Cambridge, Mass.: Harvard University Press. Scott, Rebecca J. 1985. Slave Emancipation in Cuba. The Transition to Free Labor, 1860– 1899. Princeton, N.J.: Princeton University Press. Sewell, William H., Jr. 1997. “Three Temporalities: Toward an Eventful Sociology.” In Terrence J. McDonald, ed., The Historical Turn in the Social Sciences. Ann Arbor: University of Michigan Press, 245–280. Soleau, A. 1835. Notes sur les Guyanes française, hollandaise, anglaise, et sur les Antilles françaises (Cayenne, Surinam, Demerary, la Martinique, la Guadeloupe). Paris: n.p. Smith, Adam. 1976. An Inquiry into the Nature and Causes of the Wealth of Nation. Chicago: The University of Chicago Press. Stiglitz, Joseph E. 2002. Globalization and Its Discontents. New York: W.W. Norton & Co. Taylor, P. J. 1987. “The Poverty of International Comparisons: Some Methodological Lessons from World-Systems Analysis.” Studies in Comparative International Development (Spring): 12–39. Tilly, Charles. 1984. Big Structures, Large Processes, Huge Comparisons. New York: Russell Sage Foundation. Tilly, Charles. 1995. “Macrosociology, Past and Future.” Newsletter of the Comparative and Historical Sociology Section of the American Sociological Association 8, 1 and 2: 1, 3–4. Tomich, Dale W. 1980. “Rapporti Sociali di Produzione e Mercato Mondiale nel Dibattito Recente sulla Transizione dal Feudalismo al Capitalismo.” Studi Storici 21, 3: 539–564.

The “ Second Slavery ”

1349

Tomich, Dale W. 1988. “The ‘Second Slavery’: Bonded Labor and the Transformation of the Nineteenth Century World Economy.” In Francisco O. Ramirez, ed., Rethinking the Nineteenth Century. Westport, Conn.: Greenwood Press, 103–117. Tomich, Dale W. 1990. Slavery in the Circuit of Sugar: Martinique in the World Economy, 1830–1848. Baltimore: The Johns Hopkins University Press. Tomich, Dale W. 1991. “World Slavery and Caribbean Capitalism: The Cuban Sugar Industry, 1760–1868.” Theory and Society 20, 3: 297–320. Tomich, Dale W. 1994. “Small Islands and Huge Comparisons: Caribbean Plantations, Historical Unevenness, and Capitalist Modernity.” Social Science History 18, 3: 340–358. Torras, Jaume. 1993. “The Building of a Market.” Els Espais del Mercat. 2on Coloqui Internacional d’Història Local. València, Spain: Diputació de Valencia. Trouillot, Rolph. 1988. Peasants and Capital: Dominica in the World Economy. Baltimore: The Johns Hopkins University Press. Trouillot, Rolph. 1995. Silencing the Past: Power and the Production of History. Boston: Beacon Pressn. Wallerstein, Immanuel. 1974. The Modern World System: Capitalist Agriculture and the Origins of the European World-Economy in the Sixteenth Century. Vol. 1. New York: Academic Press. Wallerstein, Immanuel. 1979. The Capitalist World-Economy. New York: Cambridge University Press. Wallerstein, Immanuel. 1980. The Modern World-System Mercantilism and the Consolidation of the European World-Economy, 1600–1750. Vol. 2. New York: Academic Press. Weber, Max. 1978. Economy and Society. Berkeley: University of California Press. Williams, Eric. 1966. Capitalism and Slavery. New York: Capricorn Books. Wallerstein, Immanuel. 1971. Inward Hunger: The Education of a Prime Minister. Chicago: University of Chicago Press. Wolf, Eric R. 1982. Europe and the People without History. Berkeley: University of California Press. Woodman, H. D. 1968. King Cotton and His Retainers: Financing and Marketing the Cotton Crop of the South. Lexington: University of Kentucky Press. Woodruff, William. 1971. “The Emergence of an International Economy, 1700–1914.” In Carlo M. Cipolla, ed., The Fontana Economic History of Europe. Vol. 4, part 2, The Emergence of Industrial Societies. London: Fontana Books, 656–716. Wright, Gavin. 1978. The Political Economy of the Cotton South. New York: W.W. Norton & Company. Zanetti Lecuona, Oscar, and Alejandro García Alvarez. 1987. Caminos Para el Azúcar. Havana: Editorial de Ciencias Sociales.

Antislavery Debates: Tides of Historiography in Slavery and Antislavery Seymour Drescher Half a century ago, the historiography of antislavery remained comfortably within the frame of reference set by the first abolitionist account written immediately after the abolition of the British slave trade in 1807. For a century and half thereafter the history of antislavery was largely the history of the dramatic triumph and expansion of the movement launched by small bands of abolitionists in America and Britain in the 1770s and 1780s. After millennia of the acceptance of slavery as a global institution, Britain was among the first nations to permanently ban its own slave trade (1807), and to emancipate its colonial slaves (1834). It was thereafter the principal agent in securing the ending of the transoceanic slave trades of the world. As late as the centennial of British emancipation in 1933 the British role in the process was celebrated as a story of humanity’s ascent to a higher moral plane. It was evoked as a selfless action that legitimized Britain’s imperial stewardship over millions of people in the Caribbean, Africa and Asia, who were still deemed to be too weak to speak for or protect themselves. In the economically depressed world of 1933 and in the face of expanding despotism and brutalization the story of abolition offered dramatic evidence of the reality of moral progress in history. Amidst the ominous news of intensively coercive regimes emanating from Germany and the Soviet Union, the long march of antislavery was hailed as the supreme example of groups with religious ideals and motivations who were able to mobilize public opinion and political will to outlaw an institution accepted throughout the world for millennia. To explain the origins of this triumphal process, historiographical attention naturally focused on the initiative of gallant individuals or small bands of heroes in Britain, the Americas, and Europe who galvanized the moral conscience of the West. Historians continued to draw upon Thomas Clarkson’s classical history of the convergence of individuals whose voices finally produced a great public tide in the affairs of men and

Source: Drescher, Seymour, “Antislavery Debates: Tides of Historiography in Slavery and Antislavery,” European Review, 19(1) (2011): 131–148. © Cambridge University Press, reproduced with permission.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_043

Antislavery Debates

1351

brothers that swept abolition on to victory in the name of providence and humanity.1,2 There were, of course, counter-currents to this tradition throughout the 19th and early 20th centuries. Initially, the dominant driving force was tied to the economic and imperial power of Pax Britannica. In the 19th-century, the United States’ South, continental Europe, and Latin America critics of the abolitionist crusade questioned Britain’s motives. They insisted that claims of humanitarians and reformers were fronts for British policies with economic or even revolutionary aims. They lacked any motivation to investigate the history of antislavery on a continental or global scale. They invariably interpreted British abolition cynically, emphasizing economic or imperialistic motives rather than humanitarian or idealistic ones.3 The moral meta-narrative remained dominant for a century and a half after the abolition of the AngloAmerican transatlantic slave trades in 1807. By the time of the sesquicentennial of British slave emancipation in 1983, however, the old interpretative frame of reference was in retreat. The mass annihilations and coerced labor systems of the Second World War had eroded the meta-narrative of Western history as a secular civilizing process. Decolonization struggles and ‘third world’ migrants in Britain called for increased attention to the experience of enslaved Africans as victims, martyrs and resisters. A new generation of African and West Indian historians emerged. They viewed the abolitionist process with far more cynicism than their British predecessors. The change of mood was epitomized in the sesquicentennial conference at the University of Hull in 1983. The opening speaker observed that even in the birthplace of William Wilberforce, the old leading actors in the historiography of abolitionism had virtually vanished from the agenda. There were no panels devoted specifically to British abolitionism, nor to its statesmen, nor to its ‘Saints’—not even a single paper on Hull’s hometown hero, William Wilberforce himself. The speaker was reminded of an English radical’s observation when he was forced to leave England in 1817 at the height of

1  See, J. R. Oldfield (2007) Chords of Freedom: Commemoration, Ritual and British Transatlantic Slavery (Manchester: Manchester University Press) ch. 4. 2  S. Drescher (1985) The historical context of British abolition. In: D. Richardson (ed.) Abolition and its Aftermath: The Historical Context (London: Frank Cass, 1985) ch. 1. 3  See, for example, F. Hochstetter (1905) Die wirtschaftlichen und politischen Motive für die Abschaffung des Britischen Sklavenhandels im Jahre 1806–1807 (Leipzig).

1352

Drescher

Wilberforce’s sanctification. Highest on Cobbett’s little list of things that he would not miss was: ‘No Wilberforces. Think of that! No Wilberforces!’4 As though to symbolize the new trend, the keynote speaker of the conference, was C. L. R. James, a radical Trinidadian writer, famous for having made the Saint-Domingue Revolution, not British abolition, the pivot of Africa’s long march from chattel slavery to political independence. James was appropriately introduced by a descendant of Wilberforce.5 It was Eric Williams, however, another Trinidadian of slave descent, who in 1944 framed the postwar historiographical challenge to the traditional narrative. A student at Oxford in the 1930s and by then a professor at Howard University in Washington, DC, Williams launched a frontal attack on the earlier view that the British had transcended self-interest in abolishing the transoceanic slave trades and colonial slavery. His Capitalism and Slavery was republished (with far greater impact) in the mid-1960s, after he had become the first Prime Minister of Trinidad and Tobago. Williams set out to devalue the entire idealistic pedigree of British antislavery. When one looked for causality in history, politics and morals were merely rationalizations for or against the thing attacked or defended. The real object of the struggle was always measurable ‘in pounds sterling or pounds avoirdupois, in dollars and cents, yards, feet and inches.’6 The history of slavery flowed directly from the ebb and flow of capitalism, the principal economic system of modern times. In phase one (mercantilism) the profits from the expanding slave trade and slavery provided the wealth that financed the British industrial revolution. In a second phase, beginning with the American Revolution and the shift to free trade capitalism, irreversible, continuous decline of the British slave system made it an encumbrance to the emergent British economy. The erstwhile ‘disinterested’ forces of abolition were thus revealed as the ideological cover for the more powerful force of industrial capitalism. The implications of this rise and decline theory of slavery were vast. Britain was not only the world’s leading industrial power during the 4  S. Drescher (1985) The historical context of British abolition. In: Abolition and its Aftermath (London: Frank Cass), pp. 3–26; also quoted in K. Charles Belmonte (2002) William Wilberforce: A Hero for Humanity (Colorado Springs: Nav Press), p. 166. 5  C. L. R. James (2001) The Black Jacobins: Toussaint L’Ouverture and the San Domingo Revolution (London: Penguin, 2001), originally published in London by Secker and Warburg (1938). 6  E. Williams (1944 [1994, 1966]) Capitalism and Slavery (Chapel Hill: University of North Carolina Press), p. 211. Williams relied primarily on Lowell Joseph Ragatz’s work of 1928, The Fall of the Planter Class in the British Caribbean, 1763–1833 (reprinted in 1966 and 1971) (New York: Octagon Books).

Antislavery Debates

1353

19th century but the hub of antislavery and the lynchpin of global emancipation in the age of abolition. The rise of antislavery depended upon the prior and irreversible decline of slavery. By the early 1970s, what one scholar coined Williams’s ‘decline thesis’ of abolition had gained currency even among scholars who objected to his uncompromising historical materialism in principle, and his dismissal of the abolitionists’ significance in particular.7 In a 1977 monograph, Econocide, I systematically challenged both the decline thesis of British slavery and that the ending of the slave trade should be viewed as a response to that development. Systematically analyzing both Williams’s favored data set and others that had not yet been examined, Econocide showed that the rise of British abolitionism in the 1780s and the subsequent slave trade abolition act of 1807 came when the British Afro-Caribbean slave system was expanding, not declining. Abolition came precisely when Britain led the world in the production of sugar and coffee. It also came at a moment when naval supremacy and colonial expansion enabled Britain to exponentially expand its tropical slave frontier. With the seizure of new territories in the Caribbean, in South America and in Africa the British were in a position to increase their transatlantic slave trade indefinitely and to dominate that trade for generations to come.8 One historian has continued to challenge Econocide’s refutation of the economic decline thesis and another has argued for a truncated temporal version of Williams’s argument. On the whole, however, ‘most other historians take the opposite view, discounting any overall case for serious economic decline in the British Caribbean by 1807.’9 Summing up the overall consensus David Richardson concludes: ‘Contrary to the assumptions of some historians, therefore, British abolitionism came against a background of economic growth 7  S. Drescher (1986) The decline thesis of British slavery since Econocide. Slavery and Abolition, 7 (May), 21. 8  S. Drescher (1977) Econocide: British Slavery in the Era of Abolition (Pittsburgh: University of Pittsburgh Press) reprinted, with a New Preface, and a Foreword by David Brion Davis (Chapel Hill: University of North Carolina Press, 2010), pp. xiii–xxx. 9  For the dissenters, see S. H. H. Carrington (2002) The Sugar Industry and the Abolition of the Slave Trade, 1775–1810 (Gainesville: University Press of Florida) who identifies Econocide as ‘one of the most polemical books since Adam Smith’s Wealth of Nations’ (p. 4); and D. B. Ryden (2009) West Indian Slavery and British Abolition, 1783–1807 (New York: Cambridge University Press). Alvin O. Thompson, a Caribbean historian, notes that ‘within the last few decades more polemical literature in the form of monographs and journal articles, has been written on this subject than perhaps any other aspect of slavery.’ A. O. Thompson (2002) Unprofitable Servants: Crown Slaves in Berbice, Guyana 1803–1831 (Barbados: University of the West Indies), p. 3.

1354

Drescher

and prosperity, not stagnation, decline or crisis in Britain’s slave colonies, with slave owners enjoying lifestyles unimaginable to the huge majority of Britons.’10 A decade after Econocide’s publication, David Eltis made another major contribution to the debate. His Economic Growth and the Ending of the Transatlantic Slave Trade (1987) extended the story to the termination of the transatlantic branch of that massive forced migration of Africans in the 1860s. Eltis mobilized elegant analytic arguments and massive empirical evidence to demonstrate that the British sacrificed substantial economic, fiscal, diplomatic and consumer benefits, not only in abolishing their own slave trade, but in acting as the major catalyst in suppressing the entire transatlantic system. The tension between economic growth and abolition remained as patent at the end of the abolitionist process as at the beginning. Eltis’s leadership in the compilation of the Transatlantic Slave Trade Database during the two decades following the publication of Economic Growth further confirmed the finding that the transatlantic slave trade remained a massive, viable and profitable enterprise for nearly half a century after British abolition. Only British-led naval and diplomatic pressures brought it to an end.11 Ironically, even historians who focus on validating Williams’s claims for decisive contributions of ‘rising’ slavery to the British Industrial Revolution before 1775 sometimes inadvertently lend support to the Drescher/Eltis refutation of the decline theory of abolition. The most massive recent compilation of data in favor of the decisive contribution of Atlantic slavery to British economic development is Joseph E. Inikori’s Africans and the Industrial Revolution in England (2002). Inikori’s own tables indicated that the total value of commodities imported by Britain from the slave-labor zones of the Americas increased from the pre-abolitionist era to the outbreak of the American Civil

10  Quotes from K. Morgan (2000) Slavery, Atlantic Trade and the British Economy, 1660–1800 (Cambridge: Cambridge University Press), p. 53; D. Richardson (2007) The ending of the British slave trade in 1807: the economic context. In: S. Farrell et al. (eds) The British Slave Trade: Abolition, Parliament and People (Edinburgh: Edinburgh University Press), pp. 127–140 (quotation on p. 133). D. B. Davis (2010), Foreword In: S. Drescher Econocide: British Slavery in the Era of Abolition, 2nd edn (Pittsburgh: University of Pittsburgh Press); Drescher; and S. Drescher (1997) Capitalism and slavery after fifty years. Slavery and Abolition, 18(3), pp. 212–227; also in Cateau and Carrington, Capitalism, 81–98. 11  D. Eltis (1987) Economic Growth (New York: Oxford University Press) esp. chs 1, 12, 13; and D. Eltis, S. D. Behrendt, D. Richardson and H. S. Klein. Voyages: the trans-atlantic slave trade database, http://www.slavevoyages.org/tast/index.faces.

Antislavery Debates

1355

War. His own sources also indicate that the British Caribbean share of this trade dropped only after the abolition of the British slave trade and slavery.12 The reassessment of the vitality of British slavery at the moment of the abolitionist turn had implications for the relation of economic development to the ending of slavery far beyond Britain and the transatlantic slave trade. The authors of Econocide and Economic Growth offered evidence against another deeply embedded concept in the historiography of slavery and abolition. Both Williams and the historians whom he challenged all accepted the proposition that slavery, a backward and inefficient mode of production, was rendered uncompetitive by free labor. Given this premise it could be argued that the obsolescence of the British slave system had actually been postponed by imperial protection. Slave owners, Williams had concluded, became Rip Van Winkles. ‘Drugged by the potion of mercantilism … [they] had gone to sleep for a hundred years’ before 1776 (Ref. 6, p. 125). The rediscovery of British slavery’s continued vitality now converged with a major historiographical breakthrough on the other side of the Atlantic. Three years before Econocide the idea of slave labor’s inefficiency was dramatically challenged by Robert W. Fogel and Stanley L. Engerman’s Time on the Cross (1974).13 Using systematic econometric methods and an abundance of new 12   J. E. Inikori (2002) Africans and the Industrial Revolution in England: A Study in International Trade and Economic Development (New York: Cambridge University Press), p. 176, Table 4.2. Inikori’s data for 1784–1856, are mainly derived from R. Davis (1979) The Industrial Revolution and British Overseas Trade (Leicester: Leicester University Press) appendix, Tables 57–64. It should be noted that Inikori’s book is primarily a defense of another of Williams’s arguments. Capitalism and Slavery asserted that the profits of the slave trade and plantation slavery were the principal sources of capital for the British Industrial Revolution. For a succinct response to this other thesis, see the article by D. Eltis and S. Engerman (2000) The importance of slavery and the slave trade to industrializing Britain. Journal of Economic History (March). In his Introduction, Inikori states that the main arguments concerning the economic basis of abolition have stood the test of time ‘in spite of voluminous opposition,’ (J. E. Inikori (2002) Africans and the Industrial Revolution in England : A Study in International Trade and Economic Development (New York: Cambridge University Press), pp. 6–7. Although Inikori does not refer to abolition in the text, he emphasizes (p. 118) that the most dynamic part of English overseas trade was with the slave-based Atlantic economies right down to 1850 (i.e. the year of Britain’s most decisive naval intervention against the slave trade outside its own empire (in Brazil). The views of the main participants in this debate are collected in B. L. Solow and S. L. Engerman (eds) (1987) British Capitalism and Caribbean Slavery: The Legacy of Eric Williams (Cambridge: Cambridge University Press). 13  R. W. Fogel and S. L. Engerman (1974) Time on the Cross: The Economics of American Negro Slavery, 2 vols. (Boston: Little, Brown).

1356

Drescher

data they demonstrated the remarkable efficiency, profitability and competitiveness of antebellum slavery in the United States’ South. The authors’ principal focus was the economics of slavery, but they ultimately confronted the long dominant ‘free labor ideology’ in American historiography. That historiography rested on the implicit assumption that slavery’s inherent immorality, brutality and hyper-exploitation bred attitudes that entailed excessive soil exhaustion, general indebtedness, disincentives to industry, high levels of poverty, illiteracy, and lethargy in all facets of civil society. Such arguments relied upon Adam Smith’s dictum declaring the inherent superiority of free labor, on enlightenment or socialist theories of historical progress, and on the abolitionist axiom that no good could come of evil. Slavery’s putative inefficiency was often equally attractive to anti-abolitionist historians. They could argue that slaveholders had actually sacrificed efficiency to the mission of civilizing slaves. In this perspective, slaveholders had assumed a ‘white man’s burden,’ sustaining an unprofitable economic system that Christianized and educated African savages. In this scenario abolitionist fanaticism was the catalyst for the most blood-drenched path to emancipation in the history of the Americas.14 The discovery of slavery’s dynamism and profitability therefore opened a whole new perspective on the fate of the institution. What was violently or non-violently ended was not an anachronistic institution somehow briefly surviving amidst modernizing capitalism. It was an integral part of Western economic growth and processes. It had to be destroyed despite its continued economic contribution to capitalism. Once historians entertained this new paradigm, they could more easily explain the economic incentives that induced one Atlantic European society after another to ship masses of slaves for three centuries to areas of low population and to produce highly desired commodities for European production and consumption. Historians working on a wide range of New World societies could now consider the Atlantic world in a new light. For nearly 300 years before 1830 four out of every five migrants brought to the New World came as involuntary African slaves. Robert Fogel, Stanley Engerman, David Eltis, Rebecca Scott, Herbert Klein, Laird Bergad and many others illustrated the ways in which colonial slave societies continued to produce consumer goods and raw materials for a rapidly expanding European economy into the late 19th century.15 Slaveholders 14  R. W. Fogel (2003) The Slavery Debates: A Retrospective, 1952–1990 (Baton Rouge: Louisiana State University Press). 15  D. Eltis (2000) The Rise of African Slavery in the Americas (New York: Cambridge University Press).

Antislavery Debates

1357

also produced considerable wealth for themselves as well. A succession of plantation societies ranked among the most productive and, per capita, the wealthiest zones on the planet: Brazil, Barbados, Jamaica, Saint-Domingue and Cuba. On the eve of the American Civil War, the monetary value of Southern US slaves was equal to 80% of the annual Gross National Product of the United States—equivalent to what today would be $9.75 trillion. As David Brion Davis notes, ‘there were good reasons why, in 1860, two-thirds of the richest Americans lived in the slaveholding south.16 One result of these two paradigm shifts—the challenges to the decline theory of abolition and to the free labor ideology—seems clear. There is now a greater appreciation of the profound incompatibility between economic self-interest and antislavery policy [see above footnote 16]. This was true not only for Great Britain and the United States but for all of the major New World Slave systems and some of those in the Old World as well. In the French empire the first abolitionist attack was launched at the very moment that France possessed the most productive and valuable slave colony in the Americas. Similarly, in the Spanish and Portuguese Americas, both internal agitation and external pressure for slave trade abolition and slave emancipation began immediately after British slave trade abolition. The universal perception by both rulers and merchants was that ‘the south Atlantic system and colonial political economies did depend on the survival of slavery.’17 Moreover, when the slave systems of the Old World began to come under pressure, in the late 19th century, emancipation began in many areas that had just undergone a surge of growth in their slave and slave trade systems. Indeed, the caution exercised by European imperialists in dismantling slave labor systems in Afro-Asia often involved a conviction of the failure of the free labor ideology in the New World. This was only reinforced by imperialist difficulties in recruiting Africans for voluntary wage labor. Increasingly, historians of slavery and abolition have realized that explanations for the turn against slavery have to come to terms with the notion that modern capitalism proved to be ‘supremely agnostic and pluralistic in its ability to coexist, and to thrive, with a whole range of labor systems right through the abolitionist century after 1780.’ And, as David Davis has added: ‘One should now extend that period to the very present’ (Ref. 11, p. 15; Ref. 8, p. 186) 16  D. B. Davis (2006) Inhuman Bondage: The Rise and Fall of Slavery in the New World (New York: Oxford University Press), p. 298. Davis reminds us that, of course, the GNP was small by today’s standards. 17  J. Adelman (2006) Sovereignty and Revolution in the Iberian Atlantic (Princeton: Princeton University Press), pp. 56–58, and 90–100.

1358

Drescher

Thus, the assumption that free labor was inherently superior to slave labor was never a consensual ideology in the West at any time during the age of abolition and emancipation. Abolitionists, of course, invoked a series of brief quotations from very eminent writers of the 18th and 19th centuries in order to bolster this perspective. They quoted Montesquieu, Franklin, Adam Smith, Tocqueville and others. Some historians also treated these disparate quotes as a consensual Western perspective throughout the age of abolition. Further research has since demonstrated that no such consensus existed regarding overseas slavery. Indeed, during the course of their respective abolitions, policy makers in most countries were deeply skeptical about the applicability of the free labor ideology to the problem of emancipation. This applied as much to Old World imperial emancipations at the beginning of the 20th century as to the New World colonies at the beginning of the 19th.18 Alternatives If systematic investigation of economic development and the political economy seems to reveal as many obstacles as encouragement for explanations of anti-slavery and the ending of slavery, how can one account for the near globalization and victories of antislavery during the course of the 19th and early 20th centuries? In the 1960s, David Brion Davis already posed the question in his Problem of Slavery in Western Culture (1965). He explored Western writings on slavery stretching over two millennia, and brought the story to the brink of the emergence of abolitionism in the 1770s.19 The key issue, however, still remained to be addressed. How could one explain the transformation of isolated examples of writers’ unease or even occasional denunciations of its barbarity into full-fledged political attack, and into 18  On the widespread American historiographical assumption of antebellum subscription to free labor superiority, see R. W. Fogel and S. L. Engerman (1989) Time on the Cross: The Economics of American Negro Slavery, 2nd edn (Boston: Little, Brown), pp. 286–287. On European skepticism towards the proposition of free labor’s inherently productive superiority, see S. Drescher (2002) The Mighty Experiment: Free Labor versus Slavery in British Emancipation (New York: Oxford University Press). For Western skepticism of its applicability in the European dominions of the Old World see, inter alia, F. Cooper (1996) Decolonization and African Society: The Labor Question in French and British Africa (New York: Cambridge University Press) ch. 2; and P. C. Emmer (1996) The ideology of free labor and Dutch colonial policy. In: G. Oostendie (ed.) Fifty Years Later: Antislavery, Capitalism and Modernity in the Dutch Orbit (Pittsburgh: University of Pittsburgh Press), pp. 207–222. 19  D. B. Davis (1965) Problem of Slavery in Western Culture (Ithaca: Cornell University Press).

Antislavery Debates

1359

their triumph of abolition over entrenched and rapidly expanding overseas slave systems?20 How could one explain the paradox that New World slavery was complementary to an expanding Western free labor capitalist system? One way to attack the problem was to give a new twist to the old actors in the humanitarian narrative. In 1975, Roger Anstey rooted the emergence of abolition in an extension of late 18th-century British religious doctrine to overseas slaves. Nevertheless he attributed the first major triumph in 1806 to deft abolitionist political manipulation of ‘national interest.’21 In 1806, an abolitionist Bill was presented to Parliament as a blow against foreign plantations conquered by the British in the conflict against Napoleon. It prevented British slavers from supplying Africans to these colonies in wartime. The abolitionist’s strategy was to effect a humanitarian outcome in a Machiavellian disguise. Abolition’s triumph was to dress a humanitarian sheep in wolf’s clothing.22 In the same year (1975) David Brion Davis offered another explanation of British abolitionism’s victory as another kind of triumph of class morality in disguise. In Davis’s scenario the emergence of British abolition was reconfigured as an elite displacement of the intensification of class exploitation and dislocation created by British capitalist industrialization. In this sense, abolition functioned as it had for many early 19th-century British radicals, as a diversion from domestic exploitation. In Davis’s subtle psychological perspective, however, the abolitionists were no more engaging in a conscious Machiavellian maneuver for metropolitan capitalism in 1807 than were Anstey’s abolitionists in 1805, prohibiting British capitalists from investing in non-British plantations. In short, for Davis, the abolitionist impulse was an unconscious reaction to disturbing discontents closer to home [see above footnote 22]. What both Anstey and Davis shared was an acknowledgement of the limits of offering moral idealism alone in explaining the victory of abolitionism over a deeply entrenched institution and its related economic interests. Many historians continued to invoke Davis’s early interpretation of abolition as a ruling class diversion after he himself eased away from it. Others dissented. The first challenger, Thomas L. Haskell, relocated the origins of abolition 20  S. Drescher (1997) Capitalism and slavery after fifty years. Slavery and Abolition, 18(3), pp. 213–221; D. B. Davis (2010) Foreword In: S. Drescher Econocide: British Slavery in the Era of Abolition, 2nd edn (Pittsburgh: University of Pittsburgh Press), p. xix. 21  R. Anstey (1975) The Atlantic Slave Trade and British Abolition, 1760–1810 (London: Macmillan). 22  Compare D. B. Davis (1975) Problem of Slavery in the Age of Revolution (Ithaca, NY: Cornell University Press) ch. 5, 8, 9, with D. B. Davis (2006) Inhuman Bondage: The Rise and Fall of Slavery in the New World (New York: Oxford University Press) ch. 11.

1360

Drescher

not in the emergence of unconscious class tensions generated by the Industrial Revolution, but in the activities and ‘cognitive style’ generated among long distance vendors. For Haskell, immersion in the international market accounted for the unprecedented ‘wave of humanitarianism reform sentiment’ [and movements that] ‘swept through the societies of Western Europe, England and North America in the hundred years following 1750. Haskell’s hypothesis dismissed any need for an ‘unconscious dimension,’ or even an unconscious capitalist displacement of class interest in abolitionism. The market itself was the medium of expanding moral perception and political action.23 This was the opening salvo in a series of exchanges about class consciousness and antislavery between Haskell, David Davis and John Ashworth, in what was ultimately published in a volume entitled, The Antislavery Debate: Capitalism and Abolitionism as a Problem in Historical Interpretation (1992) (Ref 25, see also Ref. 4, p. 11). There it became clear that one of the major difficulties in Haskell’s argument derives from the fact that it posits for the entire West what could hardly be found anywhere on the European continent in the century after 1750. There was no wave of abolitionist reform sentiment and virtually no abolitionist movement in most Western societies, least of all any that could be connected to international commercial capitalism. In other words what happened in one part of Anglo-America was simply ascribed to Europe as a whole. As David Eltis later showed (1999), Haskell’s correlation between the market and antislavery could be better used to invert Haskell’s own equation: The same traits designated by Haskell to be ‘embodied in market behavior, could have had almost exactly the opposite effect … [leading] to moral restrictions [against] enslaving one’s … fellow Europeans but not [against] the purchase of a non-European who was already a slave …’.24 The other, perhaps more significant, change from Davis’s initial focus on abolitionist elites in his Age of Revolution was a turn towards investigating the behavior of broader publics in the abolition process. One thing that Williams, Anstey, Davis and Haskell all shared was their attraction to the most articulate members of metropolitan societies, whether in business, politics, religion, or culture.25 An upsurge of historiographical interest in alternative agents in the 23  T. L. Haskell, D. Davis and J. Ashworth (1992) in what was ultimately published in a volume entitled, The Antislavery Debate: Capitalism and Abolitionism as a Problem in Historical Interpretation (Berkeley CA: University of California Press, 1992) Haskell’s quotation is found on p. 107. 24  Ibid., p. 107. 25  Both D. B. Davis (2006) Inhuman Bondage: The Rise and Fall of Slavery in the New World (New York: Oxford University Press) and Williams devoted space to the slave uprisings in

Antislavery Debates

1361

ending of slavery also began in the 1970s. It looked more closely at the mass mobilizations that led to the containment or expansion of slavery after the 1780s. One set of historians turned their attention towards slaves themselves as the key agents in the overthrow of slavery. The model for this historiographical reorientation had already appeared in 1938. As is evident from its title, C. L. R. James’s The Black Jacobins [see above footnote 5] aimed at integrating the story of the largest and most successful slave uprising in world history into the larger history of the Age of Revolution after 1776. Only in the late 1970s, however, did historians of slavery begin to argue systematically in favor of the Saint-Domingue Revolution’s central position in the history of antislavery. Robin Blackburn’s history The Overthrow of Colonial Slavery, 1776–1848 (1988), pivots on his chapter, ‘Revolutionary emancipation and the birth of Haiti.’ His narrative weaves the history of antislavery through the series of political upheavals and military conflicts that crisscrossed the Atlantic world, from the successive revolutions in the British, French and Spanish empires. Even British abolition in 1807 is reinterpreted primarily as a response to a class crisis posing a revolutionary threat.26 Haiti’s was the largest and the only successful slave revolution in history. Before and after it, however, the revolutionary wars of independence in North and South America offered ample recruitment opportunities for slaves desiring to attain freedom and for stimulating emancipation legislation that incorporated slaves into free civil status either gradually or immediately. Later New World emancipations, in the French (1848) and Spanish (1878) empires and in the United States (1865), all came in the wake of revolutions or armed conflicts.

the French or British colonies but they were not at the center of their argument. Williams’s final chapter, on ‘The slaves and slavery’ was a coda, added on to the manuscript at a late stage in the process of publication. For Davis’s own shift of attention to mass mobilizations of both slaves and citizens in the ending of New World slavery see, his five chapters on slave resistance and Anglo-American abolitions in D. B. Davis (2006) Inhuman Bondage: The Rise and Fall of Slavery in the New World (New York: Oxford University Press), pp. 205–322. 26  See R. Blackburn (1988) Overthrow of Colonial Slavery (London: Verso); and R. Blackburn (2006) Haiti, slavery and the age of the democratic revolution. William and Mary Quarterly, 63(4), pp. 643–674; For varying assessments of Haiti’s pivotal position in the history of emancipation, see E. D. (1979) Genovese. In: From Rebellion to Revolution Afro-American Slave Revolts in the Making of the New World (Baton Rouge: Louisiana State University Press); L. Dubois (2004) Avengers of the New World: the Story of the Haitian Revolution (Cambridge MA: Harvard University Press); and D. P. Geggus (2001) The Impact of the Haitian Revolution in the Atlantic World (Columbia, SC: University of South Carolina Press).

1362

Drescher

Those in the Old World’s European empires usually originated in imperial decrees from the metropole. They came in the wake of Europe’s transformation from empires of slavery to empires of antislavery. Some of these Old World emancipations were accompanied by mass flights of enslaved populations in the wake of such decrees.27 Renewed interest in the agency of slaves in the history of emancipation was also driven by a concern to diligently accumulate the evidence of slave resistance, whether successful or unsuccessful, as a means of registering the broadening consciousness of the enslaved. Some historians have recently suggested that, by resisting slavery, slaves were the primary agents in achieving their own freedom. This of course, not only diminishes the role of the heroic leaders. Most historians of antislavery imagine its long history from below as a more complex process than slave self-liberation.28 This brings us to the historiography of that other rank-and-file of abolitionism—the free populations of republics and empires. Significantly, these were the last cohorts of antislavery actors to receive systematic and comparative attention in the 1980s. One reason was that in both the Old and New World’s societies with slaves there was little collective civil activism against slavery 27  For an overview see, S. Drescher (2009) Abolition: A History of Slavery and Antislavery (New York: Cambridge University Press); For the most mobilizing decrees in the early 20th century, see M. A. Klein (1998) Slavery and Colonial rule in French West Africa (New York: Cambridge University Press) ch. 10; P. E. Lovejoy and J. S. Hogendorn Slow Death for Slavery: The course of Abolition in Northern Nigeria, 1897–1936; On the formal transition to empires of antislavery, see S. Miers (2003) Slavery in the Twentieth Century (New York: Rowman and Littlefield) chs 2 and 3. 28  On the Atlantic slave trade, see, inter alia, S. Behrendt, D. Eltis and D. Richardson (2001) The costs of coercion: African agency in the premodern Atlantic world. Economic History Review, 54(3), pp. 454–476; On a version of the Haskellian expansion of increasing identification with slaves as victims, via reports of shipboard revolts, see D. Eltis (2009) Abolition and Identity in the very long run. In: W. Klooster (ed.) Migration, Trade, and Slavery in an Expanding World (Leiden: Brill), pp. 227–257. In some slave revolts, in Demerara (1823) and Jamaica (1831) the leaders attempted to minimize violence in favor of negotiation, and succeeded in strengthening metropolitan abolitionism and accelerating the pressure for British emancipation. See D. B. Davis (2006) Inhuman Bondage: The Rise and Fall of Slavery in the New World (New York: Oxford University Press), ch. 11; and S. Drescher Civilizing insurgency: two variants of slave revolts in the age of revolution. In: E. R. Toledano (2010) Who Abolished Slavery? Slave Revolts and Abolitionism: A Debate with João Pedro Marques, eds S. Drescher and P. C. Emmer (New York: Berghahn Books), pp. 120–132. There are now abundant histories of slave resistance in the Western hemisphere and a less developed scholarship on slave resistance in the East. E. R. Toledano (2007) As if Silent and Absent: Bonds of Enslavement in the Islamic Middle East (New Haven: Yale University Press).

Antislavery Debates

1363

to discuss. Most abolitionist legislation came in the wake of revolutionary decrees or external economic or military pressures. From the first attempts at comparative studies of abolitionist movements, however, it was apparent that there were two contrasting variants of abolitionism. Mass abolitionism, best exemplified in the Anglo-American and Brazilian cases, developed distinguishing characteristics of mass social movements. They used the public sphere to bring pressure to bear upon (often reluctant) politicians and economic interests. They used mass propaganda, public gatherings, lawsuits, boycotts and mass petitions. They usually aimed at expanding participation welcoming groups otherwise excluded (by sex, religion, class, race, legal status or locality) from the existing political process. In Europe, Britain was the outstanding 19th-century example. France and Spain briefly attempted to imitate this model. Only toward the end of the 19th century was the British example replicated on the European continent. In the Western hemisphere, the northern United States and Brazil most closely approached the model first inaugurated in British society.29 The fact that late-eighteenth-century Anglo-Americans possessed the world’s most highly developed and autonomous public spheres allowed the British to become the pioneer organizers of a long-term national social movement against slavery. Thus, between 1788 and 1815 an unprecedented proportion of Britain’s people not only took decisive steps in forming a national opinion against the slave trade, but decisive steps towards creating national lecture tours, giant petition drives, vast coordinated public petition meetings and the formation of special-purpose associations of all kinds. Thereafter, Britain became an example for civil society abolitionism in the United States (1830s) and Brazil (1880s). In the United States, northern abolitionist popular mobilization provoked a counter-mobilization that resulted in the bloodiest and most costly of the New World emancipations. In Brazil, popular mobilization 29  For the theory and practice of the two variants of abolitionism, see, S. Drescher (1999) From Slavery to Freedom: Comparative Studies in the Rise and Fall of Atlantic Slavery (New York: NYU Press), chs 2, 3, 5 and 6; and S. Drescher, Abolition and its Aftermath (London: Frank Cass) chs 8, 9 and 11. On the evolution of a new civil/political society, see C. Tilly (1995) Popular Contention in Great Britain 1758–1834 (Cambridge, MA: Harvard University Press), pp. 197–204. For British abolitionism other relevant studies are, inter alia, S. Drescher (1987) Capitalism and Antislavery: British Mobilization in Comparative Perspective (New York: Oxford); S. Drescher (2009) History’s engines: British mobilization in the age of revolution. William and Mary Quarterly, 66(4), pp. 737–756; J. R. Oldfield (1997) Popular Politics and British Antislavery (Manchester: Manchester University Press); and C. Midgley (1995) Women Against Slavery: The British Campaigns, 1780–1870 (1992) (New York).

1364

Drescher

produced an abolitionism that achieved emancipation almost despite the national government.30 In another great surge of slavery scholarship during the past generation, societies with lesser abolitionist movements or societies without slavery have found their historians. They never generated mass abolitionist movements. These variants of abolitionism were usually confined to local or cultural elites. They were often imitators or even satellites of the British movement, limited both in duration and ambition. They rarely lasted beyond the demise of their own society’s or empire’s slave system. Historians identify a second surge of anti-slavery mobilization toward the end of the 19th century. In 1889, after a century of relative quiescence, Catholic societies were invited to become institutionally involved in a crusade against the slave trade and Muslim slavers in newly imperialized Africa. These societies have not yet elicited systematic and comparative historical analysis. Nor have the religiously-based societies in Afro-Asia. To the extent that this second wave was active it appears to have played far less of a role in the global emancipation process than its early 19thcentury predecessor.31

30  For the significance of popular mobilization in the United States, see inter alia, R. W. Fogel (1989) Without Consent or Contract: The Rise and Fall of American Slavery (New York: Norton); D. N. Gellman (2006), Emancipating New York: The Politics of Slavery and Freedom 1777–1827 (Baton Rouge: Louisiana State University Press); in the Spanish empire, see C. Schmidt-Nowara (1999) Empire and Antislavery : Spain, Cuba, and Puerto Rico, 1833–1874 (Pittsburgh: University of Pittsburgh Press); R. Scott (1985) Explaining abolition: contradiction, adaptation and challenge in Cuban slave society. In: M. M. Fraginals, F. M. Pons and S. L. Engerman (eds) Between Slavery and Free Labor: The Spanish-Speaking Caribbean in the Nineteenth Century (Baltimore: The Johns Hopkins Press), pp. 25–53; and for Brazil, C. T. Castilho (2008) Abolitionism matters: the policy of antislavery in Pernambuco, Brazil, 1869–1888 (PhD dissertation, University of California Berkeley). On France’s less robust second metropolitan abolitionist movement and second slave emancipation, see L. C. Jennings (2000) French Anti-Slavery: the Movement for the Abolition of Slavery in France, 1802–1848 (New York: Cambridge University Press). 31  See O. Pétré-Grenouilleau, (ed.) (2008) Abolir l’esclavage: Un réformisme à l’epreuve (France, Portugal, Suisse), xviiie–xixe siecles (Rennes: Presses Universitaires de Rennes); J.-P. Marques (2006) The Sounds of Silence: Nineteenth-Century Portugal and the Abolition of the Slave Trade trans. Richard Wall (New York: Berghahn Books); T. David, B. Etemad and J. M. Schaufelbuchl (2005) La Suisse et l’esclavage des noirs (Laussanne: Editions Antipodes).

Antislavery Debates



1365

New Paths and Reconsiderations

What is the state of the historiography of slavery and antislavery after half a century of dynamic development? One must first note the quantitative explosion in research and scholarship, whose annual bibliographical listing alone now requires 200 pages of the journal Slavery and Abolition. The historiography has become far more global with a dramatic surge of research on slavery and emancipation beyond the Atlantic world.32 Increased attention is being devoted to integrating the history of coerced labor and servitude in Europe. The advent of the Gulag in the 20th-century Soviet Union and the vast system of slave labor in Nazi Europe offer new areas of slavery for research and comparison. They amply demonstrate that the history of slavery and antislavery is not a linear one from bondage to freedom. There were, in fact, more slaves toiling in the heart of Europe in the 1940s than there had been in all of the Americas a century earlier. Nor does the history of servitude end with the 20th century. The search for ‘slave-like’ conditions in many parts of the world continues to remind us that success is never final.33 New fields of inquiry have been expanded on themes of race, class and gender, of representation, memorialization and reparation.34 The major debates on slavery and abolition during the last half century now seem to have produced a second shift. At the beginning of this essay I remarked that there was a withdrawal from the moral dimension of slavery 32  I will not here attempt an extensive overview of the recent historical works on abolition in Africa and Asia. Its historiography is less systematically embedded in or contrasted to the Western historiographical tradition. But for an introduction, see O. Pétré-Grenouilleau (2004) Les Traites négrières: Essai d’histoire global (Paris: Gallimard); G. Campbell, S. Miers and J. C. Miller (eds) (2008) Women and Slavery, 2 vols (Athens, OH); S. Miers (2003) Slavery in the Twentieth Century (Walnut Creek, CA: AltaMira Press); M. Klein (ed.) (1993) Breaking the Chains: Slavery, Bondage, and Emancipation in Modern Africa and Asia (Madison: University of Wisconsin Press); E. R. Toledano (2007) Introduction. In: As if Silent and Absent: Bonds of Enslavement in the Islamic Middle East (New Haven: Yale University Press); W. G. Clarence-Smith (2006) Islam and the Abolition of Slavery (New York: Oxford University Press); and E. Keren (2009) The transatlantic slave trade in Ghanian academic historiography: history, memory, and power. William and Mary Quarterly 66(4), pp. 975–1000. 33  S. Drescher (1985) In: D. Richardson (ed.) Abolition and its Aftermath: The Historical Context (London: Frank Cass, 1985) ch. 14. 34  See S. L. Engerman (2009) Apologies, regrets and reparations. European Review, 17(3,4), pp. 593–610; and G. Oostendie (ed.) (1996) Public memories of the Atlantic slave trade in contemporary Europe. In: Fifty Years Later: Antislavery, Capitalism and Modernity in the Dutch Orbit (Pittsburgh: University of Pittsburgh Press), pp. 611–626.

1366

Drescher

and abolition in the historiography of the 1960s and 1970s. In that age of decolonization, historians’ wariness of ascribing change to morality was reflected in Eric Williams’s devaluation of the abolitionist’s role. It was also reflected in historians’ propensity to ground the motives of actors within the boundaries of class and class conflict. If we compare the bicentennial commemoration triggered by the AngloAmerican slave trade abolitions of 2007 with the sesquicentennial of British slave emancipation in 1983, a change of tone seems evident. In 2007, there was certainly no return to the humanitarian triumphalism of 1933. Eric Williams had seen to that. However, there was a clear sense that the materialistic overtones that had so much intuitive assent in the 1960s and 1970s had simply left too many questions unaddressed. In contrast to the 1970s and early 1980s, the approach of the commemoration saw the publication of a book on the origins of British Abolitionism, Christopher Brown’s Moral Capital (2006).35 At the same moment, David Brion Davis’s most recent work, Inhuman Bondage (2006) [see above footnote 16], traced the rise and fall of New World slavery. Davis boldly opened his chapter on ‘Explanations of British Abolitionism’ with the observation that this baffling subject raised the issue of ‘moral progress in history.’ Historian Philip Morgan began an essay on abolition noting that his aim was ‘emphatically not to devalue the abolitionist achievement’ or to reduce the significance of the ‘moral revolution’ identified by Davis four decades earlier.36

35  C. L. Brown (2006) Moral Capital: The foundations of British Abolitionism (Chapel Hill: University of North Carolina Press); D. B. Davis (2006) Inhuman Bondage: The Rise and Fall of Slavery in the New World (New York: Oxford University Press) ch. 12 which now emphasizes the moral dimensions of free labor. 36  P. D. Morgan (2010) Ending the slave trade: a Caribbean and Atlantic context. In: R. Peterson (ed.) Abolitionism and Imperialism in Britain, Africa, and the Atlantic (Athens, OH: Ohio University Press), pp. 101–128. Concluding his study of the economic context of British slave trade abolition David Richardson emphasizes ‘the more enlightened moral content of an emergent political economy and their own humanitarian belief.’ D. Richardson (2007) The ending of the British slave trade in 1807: the economic context. In: S. Farrell et al. (eds) The British Slave Trade: Abolition, Parliament and People (Edinburgh: Edinburgh University Press), pp. 127–140, quotation on p. 140. For geographic extensions of the use of the moral dimension of antislavery see Peterson’s Introduction and Jonathon Glassman’s essay in the same volume: ‘Racial violence, universal history, and echoes of abolition in twentieth-century Zanzibar,’ ibid pp. 175–206.

Antislavery Debates

1367

Robert Fogel, looking back over his own decades of intense involvement in the ‘slavery debates’ in the United States37 noted that his subsequent major venture into the rise and fall of American slavery had attempted to rectify a significant lacuna in the earlier book that he authored with Stanley Engerman. That lacuna was a systematic consideration of the ‘Moral Problem of Slavery,’ and of its significance in the ideological and political campaign against slavery. None of these distinguished historians would have ventured to return to the traditional historiography of pure idealism, encapsulated in Lecky’s epitome of the British crusade ‘as probably … among the three or four perfectly virtuous acts recorded in the history of nations.’ Yet all of these historians explicitly acknowledged the role of moral arguments in explaining abolition. They all referred to the language of morality rather than of ideology. The discourse of morality was treated as an appeal to inclusiveness and universal norms. Even Robin Blackburn, writing in the radical tradition and focused on explaining the revolutionary significance of Haitian slaves’ violent destruction of the institution, noted that historians have begun to study that great revolution ‘as an event in the history of the moral imagination as well as a dramatic episode …’.38 However different their approaches, the work of these historians hints at yet another shift in historiographical perspective. In any event, the tendency to ground that momentous change within a single overarching stage of capitalist development appears to have considerably diminished. Acknowledgements I extend my thanks to Stanley L. Engerman and Pieter Emmer for their helpful suggestions and comments.

37  R. W. Fogel (2003) The Slavery Debates: A Retrospective 1952–1990 (Baton Rouge: Louisiana State University Press), pp. 45–48; R. W. Fogel and S. L. Engerman (1989) Afterword. In: Time on the Cross: The Economics of American Negro Slavery (Boston: Little, Brown), p. 289. 38  R. Blackburn (2006) Haiti, slavery, and the age of the democratic revolution. William and Mary Quarterly, 63(4), pp. 643–674. For a fine synthesis of the British and Haitian paths to abolition see A. Hochschild (2005) Bury the Chains: Prophets and Rebels in the Fight to Free an Empire’s Slaves (New York: Houghton Mifflin). The revalorization morality in accounting for the outlawing of slavery does not preclude studies of shifts and ambiguities in the understanding of ‘morality’ during the past century. See also K. A. Appiah (2010) The Honor Code: How Moral Revolutions Happen (New York: W. W. Norton), ch. 3.

Indian Ocean Slaves in Cape Town, 1695–1807 Nigel Worden The most immediate and visible connection of early colonial Cape Town to the Indian Ocean world lay in its sizeable slave population. This article will examine new data from household inventories to enumerate in more precise ways than has previously been possible the Indian Ocean origins of slaves imported into the town in the 18th and early 19th centuries. In particular, it identifies the importance of south Asian sources in the earlier parts of the 18th century and the shift to the African coast in the later 18th and 19th centuries. Both of these regions are neglected in current academic and popular perceptions of the origins of slaves in the city. Comparisons will be made throughout with slavery in other colonial Indian Ocean port cities of the period. It is only since about 2007 that South Africa’s place in the historical networks of the Indian Ocean world has begun to be recognised by South African and Indian Ocean scholars influenced by transnational approaches.1 The process S ource: Worden, Nigel, “Indian Ocean Slaves in Cape Town, 1695-1807,” Journal of Southern African Studies, 42(3) (2016): 389–408. © The Editorial Board of the Journal of Southern African Studies, reprinted by permission of Taylor & Francis Ltd, http://www.tandfonline.com on behalf of The Editorial Board of the Journal of Southern African Studies. 1  For South Africa’s Indian Ocean connections, see, most notably, I. Hofmeyr, ‘South Africa’s Indian Ocean’, History Compass, 11, 7 (2013), pp. 508–12, and P. Gupta, I. Hofmeyr and M. Pearson (eds), Eyes Across the Water: Navigating the Indian Ocean (Pretoria, UNISA Press, 2010). Cape Town’s early colonial links to the Indian Ocean are highlighted especially in K. Ward, ‘“Tavern of the Seas”?: The Cape of Good Hope as an Oceanic Crossroads during the Seventeenth and Eighteenth Centuries’, in J. Bentley, R. Bridenthal and K. Wigen (eds), Seascapes: Maritime Histories, Littoral Cultures, and Transoceanic Exchanges (Honolulu, Hawaii University Press, 2007), pp. 137–52; K. Ward, Networks of Empire: Forced Migration in the Dutch East India Company (Cambridge, Cambridge University Press, 2009); N. Worden, ‘VOC Cape Town as an Indian Ocean Port’, in H. P. Ray and E. A. Alpers (eds), Cross Currents and Community Networks: The History of the Indian Ocean World (Delhi, Oxford University Press, 2007), pp. 142–62. Recent examples of the growing awareness of Cape Town by Indian Ocean specialists include P. Larson, Ocean of Letters: Language and Creolization in an Indian Ocean Diaspora (Cambridge, Cambridge University Press, 2009); C. Anderson, Subaltern Lives: Biographies of Colonialism in the Indian Ocean World, 1790–1920 (Cambridge, Cambridge University Press, 2012). Meg Samuelson also made important initiatives with a series of workshops—‘Indian Ocean Africa’ and ‘Thinking Africa from the Cape’—held at Stellenbosch University, South Africa, in 2011. © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_044

Indian Ocean Slaves In Cape Town, 1695–1807

1369

is far from complete, but has been especially influential among historians of early colonial Cape Town. For in the 17th and 18th centuries Cape Town’s administrative, trading, demographic and cultural connections were at least as strongly linked to the Indian Ocean as to the Atlantic and European worlds, or indeed to its continental African hinterland.2 Nowhere is this more evident than in the origins of Cape Town’s slave population. This has of course long been recognised not only in academic scholarship but also in the memory of slave descendants and in wider popular awareness, through the identification of Cape Islam and ‘Cape Malays’, with a heritage from south-east Asia.3 Yet this is only part of the story, since Cape Town’s slave population came from a much wider geographical range than south-east Asia. Existing accounts of slavery in the town during the period of Vereenigde Oost-Indische Compagnie (Dutch East India Company—VOC) rule in the 17th and 18th centuries have not matched the detail and focus of Andrew Bank’s magisterial analysis of urban slavery in the period of British rule between 1806 and 1834.4 There is a good reason for this. Bank was able to draw on detailed slave registers that were kept in the decades of amelioration leading up to slave emancipation, initiated by the British authorities to check on the illegal importation of slaves into the colony after the abolition of the slave trade in 1807. No such registers exist for the preceding centuries. The only relatively complete set of data for slaves in the VOC period are the opgaaf tax rolls, but these do not distinguish urban residents in Cape Town from farmers in the surrounding grain lands of the Cape district, nor do they list slaves by name and toponym

2  For the most recent scholarship on these themes, see N. Worden (ed.), Cape Town between East and West: Social Identities in a Dutch Colonial Town (Johannesburg and Hilversum, Jacana and Verloren, 2012). 3  The literature on this is now extensive. Notable examples are R. Ross, Cape of Torments: Slavery and Resistance in South Africa (London, Routledge 1983); N. Worden, Slavery in Dutch South Africa (Cambridge, Cambridge University Press, 1985); R. Shell, Children of Bondage: A Social History of Slave Society at the Cape of Good Hope, 1652–1838 (Hanover and Johannesburg, Wesleyan University Press and Wits University Press, 1994). For the more recent growth of public and popular awareness of this heritage, see N. Worden, ‘The Changing Politics of Slave Heritage in the Western Cape, South Africa’, Journal of African History, 50, 1 (2009), pp. 23–40. 4  A. Bank, The Decline of Urban Slavery at the Cape, 1806–1843, Communications, No. 22 (Cape Town, Centre for African Studies, UCT, 1991). For work on urban slavery in the VOC period, which includes slave total numbers extracted from the few available sources, see R. Ross, ‘The Occupations of Slaves in Eighteenth Century Cape Town’, in C. Saunders and H. Phillips (eds), Studies in the History of Cape Town (Cape Town, Centre for African Studies, UCT, 1984), pp. 1–14; and parts of N. Worden, E. van Heyningen and V. Bickford-Smith, Cape Town: The Making of a City (Cape Town, David Philip, 1998).

1370

Worden

(a geographical identifier, such as Philida van Mozambique or Januarij van Timor), which is necessary to derive places of origin.5 A few random lists of Cape Town slave numbers exist from the 18th century, but they also do not provide names or toponyms.6 A pioneering account of slave origins was published in 1978, which drew on samples of slave names from official transfers, Council of Policy records and estate inventories, but these were small in size and none was able to distinguish urban slaves from those in the rest of the colony, a limitation which was equally the case in later and more comprehensive published quantitative analyses of Cape slave origins.7 This article therefore offers a more extensive analysis of the origins of Cape Town’s slave population than has hitherto been possible. It uses the database of Cape household inventories recently compiled in a four-year project by a team of seven full-time professional transcribers.8 These list slaves usually with their names and toponyms. The advantages of such a resource are manifold. The database includes household inventories in an extensive series that runs from 1695 to 1825, and so covers a much larger demographic range than any of the other samples hitherto used by researchers. Moreover, the inclusion of data about the location of properties means that for the first time it is possible to identify slaves who were living in Cape Town, rather than the surrounding farms and villages, thus giving an urban precision that even Bank was unable to do from the sources available to him. This identification of urban slave origins and their changing pattern over the course of the 18th and early 19th centuries enables us to consider the ways in which the town was connected to a wide range of Indian Ocean regions. It also invites comparisons with other VOC ports in the Indian Ocean, in all of which, 5  For a discussion of the nature of archival sources on Cape slaves, see N. Worden, ‘Cape Slaves in the Paper Empire of the VOC’, Kronos, 40 (2014), pp. 23–44. 6  For example, Nationaal Archief, The Hague, Collectie Radermacher 507, for a record of the town’s inhabitants in 1731; Cape Archives Depot, C 2764, Taxatielijst van de slaven, 19 July 1762, pp. 8–29. 7  F. Bradlow and M. Cairns, The Early Cape Muslims (Cape Town, Balkema, 1978). Later analyses include Shell, Children of Bondage and N. Worden, ‘Indian Ocean Slavery and its Demise in the Cape Colony’, in G. Campbell (ed.), Abolition and its Aftermath in Indian Ocean Africa and Asia (Abingdon, Routledge, 2005), pp. 29–49. 8  Transcribed from Cape Archives Depot, MOOC 8/1–8/40, Master of the Orphan Chamber, Inventories: General Series, 1673–1825, available at http://databases.tanap.net/mooc/, retrieved 1 August 2008. These data have already provided a major stimulus to a new quantitative approach to the 18th-century Cape, led by Stellenbosch economic historians: see J. Fourie, ‘The Quantitative Cape: A Review of the New Historiography of the Dutch Cape Colony’, South African Historical Journal, 66, 1 (2014), pp. 142–68.

Indian Ocean Slaves In Cape Town, 1695–1807

1371

unlike Cape Town, slaves were more numerically significant than in their surrounding areas.9 As Remco Raben has commented of early modern south-east Asia, ‘cities were allegedly the most greedy consumers of human slaves’, and his study of VOC Batavia and Colombo show that it was these European trading ports that were true ‘slave societies’, drawing on a range of trading networks in Indian Ocean and south-east Asian waters, rather than supplies from their agrarian hinterlands.10 Hendrik Niemeijer has similarly noted that slavery was the ‘dominant characteristic’ of 17th-century Dutch Batavia, although he adds that bonded labour in the town not only included chattel slaves owned by Europeans but also a wide range of debt-bondsmen and indentured people under the control of European, Chinese and indigenous Javanese residents.11 As Kate Ekama’s account of slavery in Dutch Colombo demonstrates, VOC slaves in urban Ceylon were as closely connected to south-east Asia and the islands of the Indian Ocean as they were to the south Asian subcontinent.12 These studies thus all suggest that slavery in VOC Asian towns was part of a complex network and that its character was shaped by forces that lay far beyond its own territorial boundaries. Analysis of slavery in Cape Town can learn much from these approaches, and not only in a comparative manner. As was the case in VOC city ports in Asia, the composition of Cape Town’s slave population was profoundly influenced by the shifting patterns of trade and exchange that took place throughout the Indian Ocean world. Kerry Ward has shown, through an analysis of the experience of convicts and exiles, how VOC Cape Town was embedded in complex demographic and cultural networks. Similar processes are evident in the ways in which slavery developed in the port. It did not exist or develop in isolation but continued to be part of a wider system of Indian Ocean bonded labour. These connections explain some major shifts that took place in the character of urban slavery in the course of the 18th and early 19th centuries. 9  R. van Welie, ‘Patterns of Slave Trading and Slavery in the Dutch Colonial World, 1596– 1863’, in G. Oostindie (ed.), Dutch Colonialism, Migration and Cultural Heritage (Leiden, KITLV Press, 2008), pp. 217–18. 10  R. Raben, ‘Cities and the Slave Trade in Early-Modern Southeast Asia’, in P. Boomgaard, D. Kooiman and H. Schulte Nordholt (eds), Linking Destinies: Trade, Towns and Kin in Asian History (Leiden, KITLV Press, 2008), p. 119; and R. Raben, ‘Batavia and Colombo: The Ethnic and Spatial Order of Two Colonial Cities, 1600–1800’, doctoral thesis, University of Leiden, 1996, p. 119. 11  H. Niemeijer, Batavia: Een Koloniale Samenleving in de 17de Eeuw (Amsterdam, Uitgeverij Balans, 2005), pp. 50–64. 12  Kate Ekama, ‘Slavery in Dutch Colombo: A Social History’, MA dissertation, Department of History, University of Leiden, 2012.

1372

Worden

Cape Town’s establishment as a VOC garrison and refreshment station in the mid 17th century came several decades after the founding of Batavia in 1619, and its population of free burghers, Company employees and officials was always much less than that of the Company’s headquarters in Java. Nevertheless the total of its Company employees (which included the administration, garrison and workshop artisans) exceeded that of the VOC’s second-largest settlement in Ceylon by the 1750s, while the number of its free burgher households increased markedly during the course of the 18th century.13 However, unlike the VOC’s Asian establishments, Cape Town did not possess a significant indigenous population. Although some Khoe were always present in the town, they were few and far between.14 There was a small Chinese community of convicts and free exiles, stemming from the Chinese diasporic trading population of Batavia, some 60 of whom, James Armstrong has shown, were slave owners, as well as other south and south-east Asian exiles, many of whom owned slaves, but this was no equivalent to the sizeable indigenous free populations of VOC south and south-east Asian port cities.15 As a result, Cape Town was always a relatively small port town within the VOC empire, albeit one whose strategic and geographical position as a halfway house between Europe and Asia gave it a particular significance. The overwhelming predominance of colonial employees and settlers in the town, a large proportion of whom owned at least one slave, meant that Cape Town was every bit as much a slave society as its Indian Ocean counterparts. For example, in 1731, one of the few years for which we have a complete demographic profile of the town, the slaves formed 42.2 per cent of its population, with VOC employees making up 30.3 per cent, free burghers 18.5 per cent and freed slaves and convicts 9 per cent.16 By 1806, slaves formed 55 per cent of the town’s inhabitants, although these population figures excluded government employees.17 Slaves were either owned directly by the Company itself or else by Company employees and free burghers in a private capacity. The 13  For accounts of Cape Town in this period, see Worden et al., Cape Town: The Making, and Worden (ed.), Cape Town Between East and West. For comparisons of VOC personnel with other VOC establishments, F. Lequin, Het Personeel van de Verenigde Oost-Indische Compagnie in Azië in de 18e Eeuw, Meer in het Bijzonder in de Vestiging Bengalen (2nd edn., Alphen aan het Rijn, Canaletto/Repro Holland, 2005), table A.26, pp. 238–43. 14  Worden et al., Cape Town: The Making, pp. 66–7. 15  J. Armstrong, ‘The Chinese Exiles’, in Worden (ed.), Cape Town between East and West, pp. 115–17. 16  Worden et al., Cape Town: The Making, p. 50 17  Bank, Decline of Urban Slavery, p. 236. This includes both privately owned and government slaves.

Indian Ocean Slaves In Cape Town, 1695–1807

1373

absence of a significant indigenous slave-owning population also ensured that slavery in the town entailed chattel bondage as enshrined in Roman Dutch law, with no evidence of the range of forms of unfree labour and debt peonage of the kind that existed in the VOC’s south and south-east Asian settlements.18 As Table 1 indicates, the total number of privately owned Cape Town slaves increased steadily throughout the period to 1806, and especially after 1749, while Table 2 shows that the number of Company slaves remained lower and more constant. This situation differed markedly from the decades after 1806. Andrew Bank has characterised urban slavery in these years as an institution in decline, marked by the ending of the slave trade in 1807 and an increasing number of manumissions and slave sales into the farming regions, to be replaced by indentured and free labour.19 The number of Company- (now government-) owned slaves dwindled sharply. None of this is apparent in the VOC period, least of all in its closing decades. But this was not the only change. There were also significant shifts in the regions of the Indian Ocean with which Cape Town’s slaves were connected. Table 1

Total number of privately owned slaves in Cape Town

year

no.

1727 1731 1749 1774 1806 1822 1827

742 767 1,038 2,373 9,367 7,160 6,222

Source: Figures from Ross, ‘Occupation of Slaves’, Appendix I, p. 14; Worden et al., Cape Town: The Making, p. 50; Bank, Decline of Urban Slavery, p. 236.

18  For example, Niemeijer, Batavia, pp. 63–4, and, for a broader discussion of the impact of European colonial use of slaves in south-east Asia, K. Ward, ‘Slavery in Southeast Asia, 1420–1804’, in D. Eltis and S. Engerman (eds), The Cambridge World History of Slavery, Volume 3, AD 1420–AD 1804 (Cambridge, Cambridge University Press, 2011), pp. 179–81. 19  Bank, Decline of Urban Slavery.

1374 Table 2

Worden Total number of company-owned (government-owned) slaves

year

no.

1693 1731 1752 1784 1807 1826

337 566 506 625 283 171

Source: Armstrong and Worden, ‘The Slaves, 1652–1834’, p. 124; Worden et al., Cape Town: The Making, p. 50. The fullest data for Company slaves is R. Shell, From Diaspora to Diorama: The Old Slave Lodge in Cape Town, CD Rom (Cape Town, Ancestry 24, 2005).



The Data

The data for privately owned urban slaves is taken from the household inventories preserved in the archives of the Master of the Orphan Chamber (MOOC). These do not provide by any means a complete listing of the slave population of the colony, nor of Cape Town. There were many slave owners whose possessions were not recorded at all, or whose estates were listed in other records, such as wills.20 Only 10,676 slaves are recorded altogether for the period up until the ending of the slave trade in 1807. We have no complete records to show the total number of privately owned slaves that existed in the colony, but Robert Shell has estimated that approximately 63,000 were imported, and to this must be added the considerable and increasing number who were born at the Cape.21 Census opgaaf records give annual totals for the whole colony,

20  The Orphan Chamber inventoried estates of owners who died intestate, who nominated the Chamber as executors or who did not specify an executor, or who left unmarried heirs under 25 years old or living outside the colony. The MOOC inventories do, however, ‘form the single most cohesive record of privately owned slaves at the Cape’, TEPC Transcription Team, ‘Introduction to the Inventories of the Orphan Chamber at the Cape of Good Hope’, available at http://databases.tanap.net/mooc/, retrieved 1 August 2014, pp. 34, 49. 21  Shell, Children of Bondage, p. 40. There are no estimates of locally born slave totals, although Shell has calculated that their number exceeded imported slaves (the ‘moment of creolisation’) in the 1760s and 1770s, and again after 1810; see pp. 47–8.

Indian Ocean Slaves In Cape Town, 1695–1807

1375

which show a rise from 337 in 1692 to 29,861 in 1806.22 The database thus includes only a small proportion of the total slave population. None the less the Orphan Chamber inventories include a much wider range of households than any other series hitherto used by researchers, both in chronological span and in wealth distribution. They include the estates of some of the wealthiest inhabitants of the town, such as Jan Daniel Herhold, who owned 50 slaves in 1807, alongside many who owned only one slave. The main advantage of this data is that the recording of the location of the properties means that researchers can now identify truly urban households from rural and semi-rural ones. For the purposes of this study, inventories that list both rural and urban properties have been excluded. This has removed many slaves who may well have been living in the town, but, since there is no way of telling exactly where they were located, their inclusion would skew the overall picture. The database therefore includes only a relatively small sample of urban slaves, but we do know for certain that they were living in Cape Town, and they seem to be broadly representative of a range of urban households. When extracted from the rest, there are 2,266 clearly identifiable Cape Town slaves in the data for the years between 1695 and 1807. Of these, 132 were repeated in inventories that were duplicated, leaving a total of 2,134. Table 3 shows the distribution of these slaves by gender and by date of inventory, in three broad categories corresponding to differing periods of the town’s history: up to 1749, when Cape Town was still in a relatively early stage of growth, from 1750 to the eve of the end of VOC rule in 1795, and the years of political transition and change between 1795 and the ending of the slave trade in 1807. More specific periodisation is problematic, since the date of each inventory reflects the chance factor of the timing of death of the owner rather than any specific chronological information about the slaves in the estate. There is a danger of over-interpreting these raw figures. Since they represent only a selection of Cape Town slaves—those who happened to be listed in deceased estate inventories—their numbers cannot be said to be representative of the total size of the urban slave population in any of the periods demarcated. It is unsurprising none the less that there are more than double the number of slaves listed for the period 1750–94 than for the preceding 60 years, given both that the number of inventories increased during that time and that the size of Cape Town’s population, both free and enslaved, grew considerably in the decades after the 1750s. 22  J. Armstrong and N. Worden, ‘The Slaves, 1652–1834’, in R. Elphick and H. Giliomee (eds), The Shaping of South African Society, 1652–1840, second edition (Cape Town, Maskew Miller Longman, 1989), pp. 129–32.

1376 Table 3

Worden Number of Cape Town slaves in MOOC database by period and gender

period

male

female

male:female ratio

unknown gender

Total

1695–1749 1750–1794 1795–1807 TOTAL

346 863 228 1,437

106 313 108 527

3.26:1 2.75:1 2.11:1 2.72:1

57 63 50 170

509 1,239 386 2,134

What is more revealing is the sex ratio for those slaves whose gender is identifiable, since the randomness of the incidence of inventories would not have affected the overall gender patterns. At no stage did the number of male and female slaves in the town reach equivalence although, as might be expected, the imbalance lessened as time went on and locally born slaves came to form a higher percentage of the total. In this regard, the urban data is almost identical to that for slave sex ratios throughout the Cape colony, which was 3.98:1 (male:female) in 1738 and 1.84:1 in 1806.23 Cape Town did not therefore follow the pattern of Asian port towns, where the slave sex ratio was more balanced—for instance in Batavia ranging from 1.06:1 in 1673 to 1.18:1 in 1797.24 Figures for the Asian ports include slaves owned by indigenous inhabitants, which may explain some of the difference, since female domestic slaves were more frequent in Asian households.25 Overall, it appears that women were a minority in Cape Town’s slave population, as they were in the rest of the colony. Where the data is most revealing is in the geographical information provided by slave toponyms. Here another caveat must be made. Despite the claims of most Cape historians, myself included, the toponyms do not necessarily indicate the place of origin of each slave but rather the region from which 23  Ibid., p. 133. 24  Raben, ‘Batavia and Colombo’, Table 4.14, p. 112. 25  The position of female slaves (as opposed to elite indigenous women) in VOC Asia has only recently been the focus of historical research. See, for example, H., Niemeijer, ‘Slavery, Ethnicity and the Economic Independence of Women in Seventeenth-Century Batavia’, in B. Andaya (ed.), Other Pasts: Women, Gender and History in Early Modern Southeast Asia (Honolulu, Center for Southeast Asian Studies, University of Hawai’i, 2000), pp. 174–94; E. Jones, Wives, Slaves And Concubines: A History of the Female Underclass in Dutch Asia (DeKalb, Northern Illinois University Press, 2010).

1377

Indian Ocean Slaves In Cape Town, 1695–1807

they were sold to the Dutch. These were not necessarily exactly the same, although the toponym was usually in the same broad region as the place of origin. Furthermore, in some of the Cape judicial records there are disparities between the toponyms given to the same slave, which reflects the carelessness with which these were recorded by official clerks, although it is likely that there was greater accuracy in the inventories, where issues of inheritance and the value of individual slaves were at stake.26 The full data for slave toponyms is shown in the Appendix. They are grouped into five regional categories: 1. 2. 3. 4. 5

south Asia, including the Indian subcontinent and Ceylon; south-east Asia, primarily Java and the islands of modern Indonesia plus a scattering from Melaka and Macao; Madagascar and a small number from the Mascarenes (Mauritius and Bourbon); mainland Africa, primarily Mozambique but including a few from West African coast and islands, but excluding those born in the Cape; Cape-born.

Table 4

Number and percentage of Cape Town slaves in MOOC database by regional toponyms, 1695–1807

S Asia SE Asia Madagascar & Mascarenes Africa Cape TOTAL

498 501 87 139 627 1,852

27% 27% 5% 8% 34% 100%

26  For examples of the disparities of Cape slave toponyms, see N. Worden and G. Groenewald, Trials of Slavery: Selected Documents Concerning Slaves from the Criminal Records of the Council Of Justice at the Cape of Good Hope, 1705–1794, second series, Vol. 36 (Cape Town, Van Riebeeck Society, 2005), pp. 5n1, 80n11, 169n1, 189n1, 455n2, 487n1, 609n2, 616n1.

1378 Table 5

Worden Number and proportion of imported Cape Town slaves in MOOC database by regional toponyms, 1695–1807

S Asia SE Asia Madagascar & Mascarenes Africa TOTAL

498 501 87 139 1,225

40.6% 40.8% 7.1% 11.3% 100%

Some 282 slave names (13 per cent of the total) lacked toponyms. The proportions of toponyms of the remainder in each category are shown in Table 4.27 These indicate that just over one-third of the total were born in the colony, while the remainder were divided between south and south-east Asia, which each formed just over one-quarter of the total, and a smaller proportion from Madagascar—Mascarenes and Africa. When the Cape-born slaves are removed, focusing on those slaves who were imported (see Table 5), the proportions rise to over one-third of the imported slaves from each of south and south-east Asia and the remainder from Madagascar, Mascarenes and Africa. These figures make an interesting comparison to Robert Shell’s estimations of the origins of all slaves imported to the Cape between 1652 and 1808, although his data did not enable him to distinguish Cape Town from the rest of the colony. Shell calculated a proportion of 22.7 per cent for Indonesia, 25.9 per cent for India, 25.1 per cent for Madagascar and the Mascarenes and 26.4 per cent for Africa.28 This may suggest that a higher proportion of south and south-east Asian slaves lived in the town than in the rural areas of the colony, thus lending some credence to Robert Ross’s earlier impression before statistical data was available, that ‘earlier in the century, and especially in Cape Town, Indonesian [slaves] were perhaps the most numerous’.29 However, as Ross’s emphasis also suggests, overall totals conceal important differences when these figures are broken down by period. Figure 1 shows the percentage for each period for all Cape Town slaves, while Figure 2 includes only those imported into the colony. Percentage proportions of the sample are 27  These figures exclude 306 slaves whose toponyms were not recorded (288 names) or were unidentifiable (18 names). 28  Shell, Children of Bondage, Figure 2–1, p. 41. 29  Ross, Cape of Torments, p. 13.

Indian Ocean Slaves In Cape Town, 1695–1807

Figure 1

Percentage of Cape Town slaves in MOOC database by toponym and period, 1695–1807.

Figure 2

Percentage of imported Cape Town slaves in MOOC database by toponym and period, 1695–1807.

1379

used because the raw figures (given in the Appendix) are dependent on the incidence of inventories and the length of the selected time period rather than the number of slaves in the town. The proportions are thus more reflective of the wider urban pattern than absolute numbers. This breakdown confirms a clear pattern of change over time. Unsurprisingly, the percentage of locally born slaves increases throughout the series, from 23 per cent up to 1749, to 33 per cent in 1750–94 and 46 per cent in 1795–1807. This confirms the increasing dependence on locally born slaves, and indeed Cape

1380

Worden

Town had almost reached a ‘creolisation moment’ in the decade before the ending of the slave trade. Once the Cape-born slaves are removed from the sample, the changes in regional origins of imported slaves over time become very evident. Up to 1749, over half of the recorded slaves in Cape Town in the MOOC database were from south Asia. This proportion declined to 30 per cent by the period between 1795 and 1807. The percentage of south-east Asian slaves was less than that of south Asian in the earlier period (34 per cent) but then increased to 46 per cent in 1750–94 before declining to a similar proportion of 29 per cent in 1795–1807. By contrast, the percentage of African imported slaves in the urban inventories increased from 7 per cent and 6 per cent in the VOC period to 38 per cent in 1795–1807. The percentage of Malagasy and Mascarene slaves remained lower throughout, at around 10 per cent or less. These changing patterns over time broadly accord with the picture of slave origins that have been identified for the Cape Colony as a whole, although now we are able to identify the urban component of this population more precisely.30 In particular, we can compare the figures for the period of an open Indian Ocean slave trade with those calculated by Bank for the decades after it ended. Between 1816 and the abolition of slavery in 1834, 71 per cent of all Cape Town slaves were locally born, while only 9.6 per cent came from south-east and south Asia combined. Nineteen per cent were from Africa (primarily East Africa and Mozambique).31 The growth of a locally born slave population after the ending of the slave trade is unsurprising, while the predominance of African over Asian slaves in the 19th century is in marked contrast to the VOC period, but an accentuation of the trend observable between 1795 and 1807.

Cape Town and Indian Ocean Trading Networks

As Gwyn Campbell has stressed, ‘the IOW [Indian Ocean world] slave trade was multidirectional, changed over time and included slaves of many different

30  For example, Shell, Children of Bondage, pp. 40–46; Bradlow and Cairns, Early Cape Muslims, pp. 118–24; Worden, ‘Indian Ocean Slavery’, pp. 29–38. 31  Bank, The Decline of Urban Slavery, p. 232. Bank’s figures drawn from the slave registers include the surrounding arable Cape district, and so are not directly comparable with the solely urban data used here.

Indian Ocean Slaves In Cape Town, 1695–1807

1381

origins’.32 Slaves brought to Cape Town were a product of these shifting and multidirectional forces in the Indian Ocean and its associated trading regions. It has long been a lament of historians of Cape slavery that we lack information about precisely how individual slaves were obtained. This is a problem for analysis of Dutch slave trading as a whole in the Indian Ocean region, and especially where Dutch trading overlapped with indigenous Asian slaving networks.33 There are few equivalents to the shipping records of transatlantic slavers, since, in the Indian Ocean, ‘slaves rarely constituted a special cargo’.34 Very occasionally, clues are provided from other records, as for example the information given by Ari (no toponym given), a slave prisoner in the Cape Town castle in 1706, that ‘during his childhood years [in the area] between Surat and Persia, when he was playing on the beach, he was carried off by the Dutch and was eventually sold as a slave’.35 There is more detailed information about specific slave trading voyages organised by the VOC from the Cape to Madagascar, some of which have recently been published, but these are a small proportion of the total.36 Similar problems face historians of slavery in other VOC port towns. However, a consideration of the patterns of trade and policy in the wider Indian Ocean arena of VOC activity reveals how dependent Cape Town was on these networks and how typical it was of Asian VOC settlements. The predominance of slaves from India and Ceylon in the first part of the 18th century is a feature of Cape Town that is little appreciated today. The focus in both academic and popular writing on south-east Asian slaves, and especially on their contributions to the influence of Islam, language and cuisine, has obscured the equal, and in earlier decades greater, dependence on regions of south Asia for slave imports into the town. South Asian sources of slaves were particularly important for VOC trading posts in both India and south-east Asia in the 17th century, and this continued well into the subsequent decades and was reflected in the slave composition of Cape Town.37 A recent publication by the Indian scholar Ansu Datta has drawn attention to the number of Bengali 32  G. Campbell, ‘The Question of Slavery in Indian Ocean World History’, in A. Sheriff and E. Ho (eds), The Indian Ocean: Oceanic Connections and the Creation of New Societies (London, Hurst Publishing, 2014), p. 125. 33  Van Welie, ‘Patterns of Slave Trading’, pp. 185–6. 34  Campbell, ‘Question of Slavery’, p. 127. 35  Worden and Groenewald, Trials of Slavery, p. 8. 36  P. Westra and J. Armstrong, Slave Trade with Madagascar: The Journals of the Cape Slaver Leijdsman, 1715 (Cape Town, Africana, 2006); D. Sleigh and P. Westra, The Taking of the Slaver Meermin, 1766 (Cape Town, Africana, 2013). 37  M. Vink, ‘“The World’s Oldest Trade”: Dutch Slavery and Slave Trade in the Indian Ocean in the Seventeenth Century’, Journal of World History, 14, 2 (2003), pp. 140–3.

1382

Worden

slaves at the Cape, although he admits that ‘detailed information is lacking as to how slaves were acquired from Bengal’.38 We do know that in the earlier 17th century Arakanese raiders captured slaves from Bengali coastal villages and traded some of them to the Dutch on the Burmese coast. The Company also used Indian slave labour in its Burmese trading posts between the 1630s and the 1670s.39 A precedent was thus established for Indian slave usage that was extended to other VOC settlements. Remco Raben has shown that in Batavia ‘from the beginning, the Company had claimed most of the slaves imported from the Indian subcontinent’, and Hendrik Niemeijer has argued that the VOC deliberately distanced itself from local slave trading in Java by turning to south Asian slave supplies, following Portuguese precedents that had made Melaka a ‘collecting place for Indian slaves’.40 Such slaves were brought both by Indian and Dutch trading vessels carrying a variety of goods such as rice, saltpetre and silk to Batavia from the Bay of Bengal, and from VOC trading posts on the Coromandel coast. Many Cape Town Indian slaves were likely to have been transferred from Batavia and to have been owned there before being taken to the Cape on the numerous vessels of the return fleets. This may explain in part why their Indian cultural and linguistic background was submerged in the more prominent south-east Asian characteristics of the Asian Cape slave population. However, sometimes the transfer from south Asia was more direct. One of the rare detailed cases we know about illustrates such a route: a 10–year-old girl named China was sold by her mother ‘because of her utter poverty and lack of means of livelihood’ to a VOC employee in the trading post at Nagapattinam, and, after three transfers of ownership and a change of name, was finally shipped to the Cape, where she ended up as ‘Rosa’ working at Groot Constantia wine estate outside Cape Town.41 Vink has stressed that slave supplies from the Coromandel coast in the 17th century were often directly connected to a ‘famine—slave cycle’, by which slaves were obtained through sale at a time of 38  A. Datta, From Bengal to the Cape: Bengali Slaves in South Africa (Bloomington, XLibris, 2013), p. 40. 39  R. Eaton, ‘Locating Arakan in Time, Space and Historical Scholarship’, in J. Gommans and J. Leider (eds), The Maritime Frontier of Burma: Exploring Political, Cultural and Commercial Interaction in the Indian Ocean World, 1200–1800 (Leiden, KITLV Press, 2002), pp. 227–8; W. Dijk, Seventeenth-Century Burma and the Dutch East India Company, 1634– 1680 (Singapore, NIAS Press, 2006), pp. 140–2. 40  Raben, ‘Batavia and Colombo’, p. 122; Niemeijer, Batavia, p. 53. 41  Cape Archives Depot, Miscellaneous 49, Serrurier papers, file (n), transfer deeds of slaves, 1763–6. This is a single rare example of a record detailing the gradual movement of a slave to Cape Town.

Indian Ocean Slaves In Cape Town, 1695–1807

1383

economic and social crisis.42 China-Rosa’s case indicates that such trends continued into the 18th century. Slaves from the Malabar coast were also prominent among the urban population of Colombo and Galle, and some were then exported to the Cape on ships travelling directly from Ceylon to Cape Town, although slaves from other regions of India as well as Batavia were found in Colombo, and some reached the Cape via this route.43 The Cape Town slaves with ‘Ceylon’ as toponyms could thus equally well have originated from the Indian mainland or even from south-east Asia. These patterns of intra-Asian slave trading in the VOC ports account for the predominance of south Asian slaves in Cape Town in the earlier 18th century and their continued presence to the end of VOC rule. They also explain the declining proportion of such slaves in the subsequent periods. Although Bengali and Malabari slaves continue to regularly appear in the Cape Town inventories in equal numbers, they were subsequently outnumbered by slaves from south-east Asia. This matched the pattern in Batavia, where south-east Asians increasingly replaced Indian slaves. These came especially from Bali and Sulawesi, where Dutch control over the Goa sultanate after 1667 led to ‘the grafting of the Makassarese slaving network to the Batavian trade system’.44 Although supplies to Batavia from these areas were sometimes circumscribed in the 18th century, they came to dominate the town. Meanwhile slave imports from India to Dutch Java lessened when VOC trading in Bengal and the Coromandel coast diminished in the face of rivalry from other European traders, especially in the period after the 1770s. At the same time, the number of ships sailing directly from Ceylon to the Cape declined markedly, since most were now directed to Batavia as part of the VOC’s involvement in intra-Asian trade. This, as Shell has persuasively argued, was a key reason for the declining importation of Indian slaves to the Cape. He maintains that it was only the activities of Danish traders from Tranquebar that maintained a supply of south Asian slaves to the colony.45 The dependence of Cape Town on Asian slave-trading patterns is still further revealed by the data for south-east Asian slaves. These slaves were brought to the Cape directly from Batavia on the return fleets and so were a filter from the larger slave population of the VOC’s Asian administrative and trading centre. Not only do their number and proportion increase markedly in the second half of the 18th century, but their toponyms closely reflect VOC 42  Vink, ‘“The World’s Oldest Trade”’, p. 142. 43  Raben, ‘Batavia and Colombo’, p. 131; Ekama, ‘Slavery in Dutch Colombo’, p. 13. 44  Raben, ‘Batavia and Colombo’, p. 122. 45  Shell, Children of Bondage, pp. 43–4.

1384

Worden

activities in the region. As Kerry Ward has emphasised, the names of southeast Asian slaves did not necessarily indicate the precise place of origin, ‘especially if we consider how mobile the archipelagic seafaring, trading and refugee populations were’.46 None the less, the Cape Town database names for south-east Asian slaves are a clear reflection of the directions that Batavian trading took in the course of the 18th century. ‘Boegies’ or ‘Makassar’ toponyms predominate, forming half of the total of south-east Asians for the whole sample. This was the result of the VOC’s active slave trading and raiding in the Sulawesi region from the late 17th century, which drew on the existing activities of indigenous slave traders.47 Although there were periodic restrictions on the importation of such slaves from the eastern regions of the Indonesian archipelago, Makassar, as Raben has argued, ‘remained the nodal point in an extended slaving network’ and continued to be the main region of Batavian slave imports from the late 17th century, forming over 40 per cent of the total.48 Their visibility in Cape Town and its surrounds was equally marked. The Bugis language was spoken—and written—in Cape Town, and the combination of Bugis slave literacy and adherence to Islam was a cause of considerable concern to the Cape authorities.49 Secondary regions of importance for Batavian slave supplies also figure in the Cape Town data. These include Bali, another ‘axis’ of slave trading, not only from the Balinese kingdoms themselves but also from islands further to the east.50 Sumbawa, an island to the east of Bali and part of its catchment area, was specified for a number of Cape Town slaves. The islands of Timor and Ternate, both further removed from Batavia, were also important sources, incorporated into the wider Makassarese network by the 1700s. The long tentacles of slave trading into the islands of the eastern Indonesian archipelago were thus extended via Batavia to the other end of the Indian Ocean. The shift from south to south-east Asian sources of slaves in Cape Town was gradual and never complete, but it was the product of a changing pattern of slave trading in VOC Asia, over which the Cape authorities had little control. Such imports continued to the end of the VOC period, despite some attempts by the Batavian authorities to limit the transportation of male slaves from Asia 46  K. Ward, ‘Southeast Asian Migrants’, in Worden (ed.), Cape Town Between East and West, p. 86. 47  Van Welie, ‘Patterns of Slave Trading’, pp. 196–7. 48  Raben, ‘Batavia and Colombo’, p. 125. 49  Worden and Groenewald, Trials of Slavery, pp. 355–84, 537–56; R. Ross and S. Koolhof, ‘Upas, September and the Bugis at the Cape’, Archipel, 70 (2005), pp. 281–308. 50  Raben, ‘Batavia and Colombo’, p. 125.

Indian Ocean Slaves In Cape Town, 1695–1807

1385

to Cape Town in the late 1760s as part of a move to prevent private slave trading by Company officials returning to Europe.51 The other regions of slave origins in Cape Town stemmed from the very different character of trading in the south-western Indian Ocean. Throughout the 18th century, Malagasy slaves appear in relatively small numbers in the inventories. Madagascar was a major provider of slaves to a range of European powers between the 17th and 19th centuries, notably to the sugar plantations of French Bourbon and Mauritius.52 The VOC in Cape Town also relied on the island for slave supplies, sending between 33 and 38 ships to Madagascar.53 Although it thus had greater control over the way in which it could obtain slaves from this region than was the case for Asian slaves, it was singularly unsuccessful in establishing a regular source of supply to satisfy its requirements, often lamenting the competition from other European powers, particularly the French and English, as well as high prices, low number and ‘poor quality’ of the slaves that were offered to them by Malagasy local polities. Shell has argued that some 25 per cent of all Cape slaves were imported from Madagascar, a figure considerably higher than the 2,870 estimated by Armstrong or 3,753 by Ravell.54 It is also higher than the percentages shown in the MOOC database. However, Malagasy slaves also formed a large proportion of the slaves owned by the Company, as enumerated in Table 2, who were not recorded in the MOOC inventories. Records of the deaths of Company slaves indicate that very many were from Madagascar, as Table 6 indicates. Moreover, the figures for Madagascar are a considerable under-estimate, since a large proportion of the toponyms that could not be precisely identified appears from their spelling to be Malagasy names. All in all, at least half of the total number of Company slaves was imported from Madagascar. There were chronological variables in the flow of the Malagasy slave trade that these overall figures only partially reveal. Up to the 1720s, Madagascar was the main source of Company slaves, but when the VOC held a trading post at Rio de la Goa between 1721 and 1730 the south-east African coast became more a 51  Van Welie, ‘Patterns of Slave Trading’, p. 206; Shell, Children of Bondage, p. 43 n. 8. 52  J-M. Filliot, La Traite des Esclaves vers les Mascereignes au XVIIIe Siècle (Paris, ORSTOM, 1974); D. Scarr, Slaving and Slavery in the Indian Ocean (Basingstoke, Macmillan, 1998); M. D. North-Coombes, ‘Labour Problems in the Sugar Industry of Ile de France or Mauritius, 1790–1842’, MA dissertation, University of Cape Town, 1978, especially Ch. 2. 53  The figures differ slightly because of divergent lists by the key historians of the trade: Armstrong and Worden, ‘The Slaves’, p. 112; J. Ravell, ‘The VOC Slave Trade between Cape Town and Madagascar’, unpublished paper, Bilthoven, 1979. 54  Armstrong and Worden, ‘The Slaves’; Ravell, ‘The VOC Slave Trade’; Westra and Armstrong, Slave Trade, p. 9.

1386 Table 6

Worden Identified Company slave toponyms from death records, 1719–1789

S Asia SE Asia Madagascar Rio de la Goa Mozambique

1719–1749

1750–1789

36 24 96 219 2

29 38 601 70 415

Source: Data drawn from Iziko Slave Lodge database of Company slave deaths, derived from Cape Archives Depot, C 2432–2652, Attestatien, 1719–1789.

significant source of supply. After the abandonment of Rio de la Goa, the slaves held there by the VOC were brought back to Cape Town, thus explaining the high number of such slaves in the period 1719–49. Supplies from Madagascar resumed between 1740 and 1785, although trading with Mozambique grew markedly in this period, while the Malagasy slave trade tailed off with the decline of the VOC from the late 1780s.55 Not all Company slaves worked in Cape Town, although the majority were housed in the Slave Lodge in the centre of the settlement and would have formed an important part of the urban population. However, the visibility of both Company- and privately owned slaves from Madagascar declined in the 19th century, as the number of mainland African slaves in the town increased. None the less, as Pier Larson has argued, they formed a higher proportion of the slave population of Cape Town than that of the colony’s rural districts, and there is evidence of the continued use of Malagasy in the town and its environs by slaves and their descendants until at least the 1840s.56 For it was the mainland African continent that became the primary source of slave imports towards the end of the VOC period and into the final years of the legal slave trade under British and Batavian authorities. In the earliest years of the VOC Cape settlement in the 1660s and 1670s, a few slaves were imported from the Dutch West India Company’s regions of influence in Angola and 55   N. Worden, ‘Slavery and Amnesia: Towards a Recovery of Malagasy Heritage in Representations of Cape Slavery’, in I. Rakoto (ed.), L’esclavage à Madagascar: Aspects Historiques et Résurgences Contemporaines (Antananarivo, Institut de Civilisations— Musée d’Art et d’Archéologie, 1997), p. 54. 56  Larson, Ocean of Letters, p. 244.

Indian Ocean Slaves In Cape Town, 1695–1807

1387

Guinea, but this was halted after West India Company protests against VOC infringement of their trading monopoly, and only one of them survived into the Cape Town inventories that start in 1695.57 Subsequently the VOC trading post at Rio de la Goa had provided a source of slaves for the Company, and especially after the transfer of the slaves from there when it was shut down in 1730, but very few of these ended up in private hands. The major replacement of the town’s slaves by African imports came rather with the development of private slave trading from the colony in the 1790s and with the availability of slaves transported by Portuguese traders en route from south-east Africa to Brazil and later by Brazilian traders directly. Both of these developments have recently come under focus, providing a wholly new perspective on the integration of the Cape into the south-west Indian Ocean—south Atlantic slave trade. Michael Reidy has shown how private traders operated from Cape Town in the period of the first British occupation, after VOC slaving operations had ended in the late 1780s. Their sources were the coast of Mozambique and Mozambique Island, thus leading to Cape mercantile participation in the burgeoning slave trade of the region. Reidy’s research continues, but his earlier work demonstrated that despite the diffidence of the British authorities at the Cape, who were aware of the growing anti-slave trade sentiments in London, some 7,200 slaves were imported into the colony by Cape merchants or purchased in the town from passing Portuguese traders between 1797 and 1808.58 Patrick Harries’s recent research, some of which is published in this issue, has also exposed the existence of a ‘middle passage’ from Madagascar, Mozambique and the east African coast, which he traces back to French, Spanish and Portuguese traders who brought slaves to Cape Town from the 1770s, a trade in forced labour that continued well into the decades after 1808 and indeed after the ending of slavery itself in the 1830s.59 His work reminds us that foreign traders also played a part in bringing slaves into Cape Town in earlier decades, although there has been little research into these activities. In addition, there were the ‘prize negroes’ or slaves captured after 1808 by British patrols and indentured at the Cape. There is currently much research under 57  Van Welie, ‘Patterns of Slave Trading’, p. 203. 58  M. Reidy, ‘The Admission of Slaves and “Prize Negroes” into the Cape colony, 1797–1818’, MA dissertation, University of Cape Town, 1997, p. 111. 59  P. Harries, ‘Middle Passages of the Southwest Indian Ocean: A Century of Forced Immigration from Africa to the Cape of Good Hope’, Journal of African History, 55, 2 (2014), pp. 173–90; P. Harries, ‘Slavery, Indenture and Migrant Labour: Maritime Immigration from Mozambique to the Cape, c.1780–1880, African Studies, 73, 3 (2014), pp. 323–40.

1388

Worden

way on these, and on the ways in which they provided an important supplement to the unfree labour pool of the colony after the ending of the slave trade. A further possibility, although one as yet completely unresearched, is that some African slaves were brought to the Cape as part of the regular shipping trade that existed between Mauritius and Cape Town in the last decades of the 18th century. It is likely that slaves were in this way brought from the active east African—Malagasy—Mascarenes slaving networks to Cape Town, where they were sold by traders who returned with Cape agrarian produce needed to feed the rapidly increasing number of slaves on the Mauritian sugar plantations.60 There are relatively few slaves in the inventories with toponyms from Mauritius, although one famous example is Louis van Mauritius, leader of the 1808 Cape slave revolt, who was brought from the Mascarenes as a young boy in the late 1780s.61 A number of slaves of Malagasy and east African origin could also have been imported in this way via the Mascarenes. The result of these shifts for the composition of Cape Town’s slave population was striking. By the period of British rule, this was coming more to resemble Port Louis in Mauritius, with its predominance of slaves of Malagasy and Mozambican origin, rather than the VOC ports of south and south-east Asia.62 Cape Town’s slave networks were now detached from their Asian roots and instead firmly embedded in the African and south-west Indian Ocean nexus.

Indian Ocean Slavery and the Politics of Race

These shifts, firstly from south to south-east Asia, then from the Asian to the African Indian Ocean, have been noted by historians of Cape slavery, but their implications for the social history of the town are still little appreciated. A historiographical focus on the survival and growth of south-east Asian elements, particularly Islam, in Cape Town has instead predominated. This has more 60  I owe this observation to one of the anonymous readers of this article, for which I am most grateful. 61  N. Worden, ‘“Armed with Ostrich Feathers”: Cultural Revolution and the Cape Slave Uprising of 1808’, in R. Bessel, N. Guyatt and J. Rendall (eds), War, Empire and Slavery, 1770–1830 (London, Palgrave, 2010), pp. 121–38. 62  There is now a thriving literature on Mauritian slavery, but, as at the Cape, very little of this specifies the character of Port Louis as distinct from the plantations of the island where most of the slaves laboured. For an earlier comparative study of both ports, see N. Worden, ‘Cape Town and Port Louis in the Eighteenth Century’, in G. Campbell (ed.), The Indian Ocean Rim: Southern Africa and Regional Co-operation (London and New York, Routledge Curzon, 2003), pp. 42–53.

Indian Ocean Slaves In Cape Town, 1695–1807

1389

recently taken a transnational turn, notably in the research of Kerry Ward, who has shown how not only south-east Asian slaves but also freed slaves, exiles and convicts played a key role in cementing Cape Town’s links to Asian networks within the VOC empire.63 New work on free black kinship and contacts is now beginning to show the ways in which south Asian connections were also maintained in the 18th-century town, while the continued meshing of Cape Town and Robben Island in the convict system of the 19th-century Indian Ocean and the migrations that accompanied it are also receiving attention.64 This article indicates that similar research needs to be undertaken on the social and cultural effects of the Africanisation of Cape Town’s slave population, especially in the light of Patrick Harries’s and Michael Reidy’s exposure of its scale and character. Pier Larson’s study of the linguistic and cultural connections that drew Cape Town into the Malagasy diaspora of the early 19th century provides an important blueprint for such work. Such approaches will need to confront not only academic but also popular beliefs about the character of slavery in Cape Town. For both the south Asian and the African components of Cape slave history have been obliterated in public memory and in heritage representations. This amnesia has a long history, which can be traced back to the period of slavery itself. By the beginning of the 19th century, contemporary visitors and commentators described the ethnic composition of the slave population at the Cape in ways that excluded, or at least submerged, the variety of regions of origin that this article has identified. This was not the case in the 18th century. Otto Mentzel, one of the most perceptive writers about the colony, who lived at the Cape in the 1730s and wrote his three-volume account of the colony in 1785, gave a reasonably full description of the range of Indian Ocean origins of Cape slaves and the changing patterns of importation: [t]he first slaves were brought from Madagascar, but the French afterwards put a stop to this traffic, and importations were made from India and the East Indies, from Bengal, Amboina, Ternate, Tutegrin, Negapatam, Trincomali, Java, Kandy, Surat, Malabar, and also from the Portuguese settlements. All these races live indiscriminately together … but one type of slave is regarded as inferior and has to be kept apart.

63  Ward, Networks of Empire; Ward, ‘Southeast Asian Migrants’. 64  S. Newton-King, ‘Family, Friendship and Survival among Freed Slaves’, in Worden (ed.), Cape Town between East and West, pp. 153–75; Anderson, Subaltern Lives.

1390

Worden

These have been brought from the land of Terletan, situated on the Rio de la Goa [Mozambique] …65 By the 19th century, however, the shifting nature of slave importation meant that contemporary observers simplified their characterisation of Cape slaves into two categories: ‘Malay’ and ‘Mozambique’. Thus an anonymous writer of 1806 commented that, ‘The Malay and Mosambique slaves principally compose those of the Cape’. He continued by entrenching racial stereotypes in this division, with the former … distinguished by their copper colour; and the latter by a jet black and thick lips. The Malay, cruel and revengeful in his disposition, is allowed to possess a better capacity for instruction, and when taught a trade becomes a source of profit to his master. The Mosambique slave has a dull, inanimated appearance strongly stamped upon him, indicating, as it were, an inferiority of intellect.66 This crude, racially stereotyped division of slaves into ‘Malays’ and ‘Mozambicans’ became the standard fare of early 19th-century commentators. Thus the administrator John Barrow’s detailed description of the Cape divided ‘Malay’ slaves from ‘the negroes of Mozambique and Madagascar’, and the British traveller William Burchell commented that ‘the Mozambique and Madagascar slaves are at once distinguished from the Malays by their black colour, woolly hair and negro countenance’. William Bird, the Customs Controller in Cape Town in the 1820s, divided slaves at the Cape ‘into three classes: the Negro, the Malay and the Africander’ (locally born slaves).67 In this simplified and bifurcated racial hierarchy, the complexity of Indian Ocean origins was lost to sight. In particular, the south Asian component of Cape Town’s slave population, which had been so significant in the earlier 18th century, and which continued to be evident in the inventories into the 65  O. Mentzel, A Complete and Authentic Geographical and Topographical Description of the Famous and (All Things Considered) Remarkable African Cape of Good Hope, Vol. II (Cape Town, Van Riebeeck Society, 1925), p. 125. 66  Anon., Gleanings in Africa (London, Cundee,1806), p. 58. 67  J. Barrow, An Account of Travels into the Interior of Southern Africa in the Years 1797 and 1798, Vol. I (London, Cadell and Davis,1801), p.108; W. Burchall, Travels in the Interior of Southern Africa (London, Longman, Hirst Rees, Orme and Brown, 1822), p. 33; F. Krauss, ‘A Description of Cape Town and its Way of Life, 1838–40’, Quarterly Bulletin of the South African Library, 21, 1 (1966), p. 5. For further discussion of these sources in relation to their representation of Malagasy slaves, see Worden, ‘Slavery and Amnesia’, pp. 57–9.

Indian Ocean Slaves In Cape Town, 1695–1807

1391

19th century, was completely obliterated, and continues to be so to this day. The nomenclature ‘Malay’, a term that came to define Muslim slaves in Cape Town in particular, was instead swiftly associated with the geographical region of south-east Asia, although it is highly probable that at least some Cape slaves from India and Sri Lanka were also Muslims, and that a number of south-east Asian slaves were not. The close ties of south and south-east Asia in the trading networks of the VOC, and the likelihood that many Indian slaves came to the Cape via Batavia, further elided the distinctiveness of a south Asian Cape slave memory. Only with more recent scholarship has the importance of regions such as Bengal, Malabar and Coromandel to Cape Town’s demographic composition become evident.68 The ending of slavery in 1834 brought a further perceptual redefining of its racial characteristics, which has also continued to the present. The racial order of the post-emancipation Cape came to define ex-slaves, along with the Khoe and San inhabitants of the Cape, as ‘coloured’, in opposition not only to the descendants of ‘white’ colonists but also to the indigenous African population of the colony. Thus masters and servants legislation proposed by the Cape legislative council in 1839 defined emancipated slaves as ‘persons of colour’ in contrast to ‘native’ Africans.69 As Vivian Bickford-Smith has argued, the term ‘coloured’ was used in mid—later 19th-century Cape Town as an antonym to ‘white’, and could refer to all black Capetonians, including Africans. However, it was also increasingly used in distinction from ‘native’ Africans, to include the categories ‘Malays’ and ‘Mixed and others’, as used in the 1871 census, but excluded other Africans.70 As eastern Cape migrant workers came to Cape Town to work on the harbour from the 1870s, the division between ‘coloured’ and ‘native’ was further entrenched. Cape Town’s slave descendants were in this context perceived as ‘coloured’, in particular as ‘Malay’, rather than ‘African’. As early as the 1850s, and certainly by the 1870s, a distinct Malay ethnicity, associated with the south-east Asian roots of Cape Islam, had emerged. Celebrations of slave emancipation day on 1 December were marked with picnics and parades, at which distinctive ‘Malay’ dress, dancing and music predominated. Cape Town slavery was now associated firmly with an Asian rather than an African past. This contrasted with the situation in the rural hinterland, where, as Patrick Harries has argued, the 68  Datta, From Bengal to the Cape. 69  T. Keegan, Colonial South Africa and the Origins of the Racial Order (Cape Town, David Philip, 1996), p. 125. 70  V. Bickford-Smith, Ethnic Pride and Racial Prejudice in Victorian Cape Town (Cambridge, Cambridge University Press, 1995), p. 31.

1392

Worden

descendants of African slaves, as well as more recent migrant farm workers, were known as ‘Mozbiekers’ in the mid to late 19th century.71 The politics of coloured identity in 20th-century Cape Town did little to change this perception. One trend was to accentuate the differences between ‘coloured’ and African Capetonians, and in the process to link slavery further to an Islamic and south-east Asian heritage that ignored its African components. This was evident in the conscious construction of a ‘Malay’ heritage and culture in the 1940s and 1950s and was accentuated by the separation of ‘black’ African and ‘Malay’ culture and history in the racial schema of the apartheid era, although the emphasis of this segregated ‘Malay’ heritage lay on the romanticized stories of Muslim exiles and princely rulers rather than on the realities of chattel slavery.72 In opposition to this, a radical and anti-segregationist tradition emphasised the unity of working-class struggles that rejected racially determined distinctions, but in the process the legacy of Cape slavery was also played down, since it could not be linked to the broader history of indigenous land dispossession and the migrant labour system of the bantustans.73 The ending of apartheid led to a major recovery of awareness of the importance of slavery to the heritage of Cape Town, marked by popular publications, a variety of cultural outputs, and museum and heritage memorialisations.74 Yet while the diversity of the Indian Ocean roots of the city’s slave past was often celebrated and acknowledged in all of this, the emphasis remained on southeast Asia, while the African elements of Cape slavery were usually neglected. There are several reasons for this. The ‘Malay’ tradition, so strongly emphasised since the mid 19th century, is deeply rooted in public consciousness and has continued to be a focus of celebration and attention. One of the first major public heritage events to be held at the time of apartheid’s collapse in 1994 was a commemoration of ‘300 years of Islam’ held at Cape Town’s Castle, which was accompanied by books and museum exhibits stressing the Muslim background of Cape slaves and political exiles from south-east Asia. Although a 71  P. Harries, ‘Making Mozbiekers: History, Memory and the African Diaspora at the Cape’, in B. Zimba, E. Alpers and A. Isaacman (eds), Slave Routes and Oral Tradition in Southeastern Africa (Maputo, Filson, 2005), pp. 91–123. 72  S. Jeppie, ‘Historical Process and the Constitution of Subjects: I. D. du Plessis and the Reinvention of the “Malay”’, BA (Hons) dissertation, University of Cape Town, 1986–7; L. Witz, Apartheid’s Festival: Contesting South Africa’s National Pasts (Bloomington and Cape Town, Indiana University Press and David Philip, 2003), pp. 178–9. 73  K. Ward and N. Worden, ‘Commemorating, Suppressing and Invoking Cape Slavery’, in S. Nuttall and C. Coetzee (eds), Negotiating the Past: The Making of Memory in South Africa (Cape Town, Oxford University Press, 1998), pp. 208–9. 74  Worden, ‘The Changing Politics of Slave Heritage’.

Indian Ocean Slaves In Cape Town, 1695–1807

1393

wider focus included India and Madagascar, this emphasis on Islam excluded slavery from the African mainland.75 As Pumla Gqola has recently pointed out, Cape Malay identity has now recovered its slave roots, but the ‘African’ dimension of Cape creolisation, in which slavery formed such an important part, is still neglected.76 Instead slaves from Africa are firmly associated in school textbooks, museum displays and popular publications with the transatlantic slave trade, in which the Cape and South Africa played no part. There is, by contrast, a lack of awareness of the use of African slaves in the Indian Ocean trade, which included supplies to the Cape. This was apparent when the South African government, with no sense of irony, joined the demand for reparations and compensation for the export of slaves from Africa at the Durban World Conference Against Racism in 1991. The neglect of the African and (to a lesser extent) south Asian elements of Cape Town’s slave heritage is a highly politicised issue in the context of the contested contemporary politics of the Western Cape region of South Africa. As the only province without a majority African-speaking population, and without an African National Congress local government, many see the region as distinct in its historical and social character from the rest of the country. Africans, it is widely believed, came to Cape Town only in the 20th century, or, at the earliest, with the migrant dockworkers of the late 19th century. To stress the African elements of Cape Town’s slave past would thus not only challenge the dominant public perceptions of urban slave heritage as Muslim and Asian, but also emphasise that Africans have an equally long-standing place in the city’s history. To appreciate the full implications of Cape Town’s past thus requires greater awareness of the complexities of the Indian Ocean Asian and African slave-trading networks, an issue which academic work needs to continue to address.

75  K. Ward, ‘The “300 Years: Making of Cape Muslim Culture” Exhibition, Cape Town, April 1994: Liberating the Castle?’, Social Dynamics, 21, 1 (1995), pp. 96–131. The exhibition was organised by the Sheikh Yusuf Tercentenary Commemoration Committee, led by the historian and linguist Achmat Davids. The main research publication associated with this was Y. da Costa and A. Davids (eds), Pages from Cape Muslim History (Pietermaritzburg, Shooter and Shuter, 1994). 76  P. Gqola, What is Slavery to Me?: Postcolonial/Slave Memory in Post-Apartheid South Africa (Johannesburg, Wits University Press, 2010), pp. 140–1.

1394

Worden

Appendix: Cape Town Slaves Listed in MOOC Database, 1695–1807

South Asia M

Bengal Malabar Cust’ /Kust Westcust Madras Nagapatnam Surat Coromandel or East coast Trancquebar Tuticorin Pulicat Goa Cochin Manjeri Ceylon Sub-total

41 29 5

12 1

3

2 2

1

3 4 1

1

8

2

12 107

2 19

South-East Asia M

Batavia Bantam Java Madura Bali Sumbawa Bima/ ‘Dima’ Bougies Macassar Mandaar Bouton Ambon

1695–1749 F U

26 12 2

1 1

6 1 18 3 257

2

4

1

1

9

2 3

99 99 16 10 1

M

6 24 21 1

M

1

2

3 46

2 7 53

M

4

215 162 24 10 2 3 3 5

164 149 22 10 1 2 2 4

42 13 2 0 0 1 1 1

9 0 0 0 1 0 0 0

1

7 8 1 0 19 1 38 498

6 7 1 0 16 1 32 417

1 1 0 0 3 0 6 71

0 0 0 0 0 0 0 10

Total

M

Total F

U

60 2 17 0 32 39 5 172 70 16 6 4

46 5 1 1 6 2 0 0 24 3 35 3 4 1 142 24 55 12 8 3 6 0 3 1

1 6

0

M

1795–1807 F U

2

3 1 1

3

15

21 3 27 3 4 1 117 12 36 8 6 3 2 3 1

5 1

1 3

6 2 4

15 4

Total F

Total

1 2

1750–1794 F U

19 1 5

1795–1807 F U

1

3 3

12

10 15 2 4

1750–1794 F U

1

1695–1749 F U

2 5

M

10 1

1 1

U

9 0 9 0 5 1 0 6 3 5 0 0

1395

Indian Ocean Slaves In Cape Town, 1695–1807

South-East Asia

1695–1749 M F U

Banda Timor Ternate Banjar Banjarmasin Palembang Padang Sumatra Nias Macao Malakka Sub-total

61

Madagascar & Mascarenes

M

Mauritius Bourbon Madagascar Sub-total

1 2 4

18 13

2

13 13

8

1 16 1 1 3 13 294 44

1695–1749 F U

2 2

0

1695–1749 F U

M

5 1 2

1 1 5

1795–1807 M F U

3

1 2

1

1

Africa

Rio de la Goa Mozambique [Terra de] Natal Angola Guinea Cape Verde St Jago/Santiago Sao Tome Sub-total

1750–1794 M F U

M

2 1 42 45

M

7

1

25

41

11

4

1750–1794 F U

M

1795–1807 F U

2

1

1

2 4

4 5

1750–1794 F U

M

17 17

4 27

5 12

1 32

17

0

1

1795–1807 F U

1 51

21

52

21

1 1 10

7

0

0

0

Total F

Total

M

2 25 24 0 0 1 1 1 20 1 3 501

1 1 21 1 19 5 0 0 0 0 1 0 1 0 1 0 18 1 1 0 3 0 396 63

Total

M

Total F U

6 1 80 87

3 1 59 63

0 0 19 19

Total

M

Total F

22 112 2 0 1 0 1 1 139

10 79 2 0 1 0 1 1 94

U

0 3 0 0 0 0 0 0 1 0 0 42

3 0 2 5

12 33 0 0 0 0 0 0 45

U

0 0 0 0 0 0 0 0 0

1396

Worden

Appendix (cont.) Africa M

Unknown Total non-Cape Cape born Total Total known

1695–1749 F U

119 41 35 191 36 16 36 29 6 346 106 57 227 65 22

M

35 628 200 863 828

1695–1749 M F U

Overall % % S Asia % SE Asia % Madagascar and Mascarenes % Africa % Unknown % Cape born % Non-Cape born % of Known Toponyms % S Asia % SE Asia % Madagascar & Mascarenes % Africa % Cape born % Non-Cape born % M/F in period known total % M/F Cape born in period

1750–1794 F U

12 124 177 313 301

M

1795–1807 F U

Total

16 9 5 10 282 36 151 38 5 1,225 11 68 65 35 627 63 228 108 50 2,134 47 219 103 40 1,852

1750–1794 M F U

1795–1807 M F U

Total F

M

163 970 304 1,437 1,274

Total M

U

58 198 271 527 469

61 57 52 170 109

Total F U

31% 18% 5% 30% 15% 11% 23% 6% 0% 23% 29% 13% 6% 18% 8% 23% 34% 14% 40% 18% 10% 8% 23% 28% 12% 25% 4% 2% 0% 5% 5% 6% 2% 0% 2% 4% 4% 4% 3% 3% 34% 10% 55%

7% 39% 27% 34%

0% 61% 11% 28%

4% 4% 23% 73%

5% 4% 57% 40%

0% 25% 17% 57%

23% 4% 30% 66%

19% 5% 60% 35%

0% 7% 7% 9% 20% 13% 11% 11% 70% 29% 21% 51% 10% 57% 68% 38%

0% 36% 31% 34%

47% 29% 14% 31% 15% 15% 24% 6% 0% 27% 33% 15% 9% 27% 12% 59% 36% 15% 53% 19% 11% 10% 27% 31% 13% 39% 6% 3% 0% 5% 6% 9% 2% 0% 3% 5% 5% 4% 5% 4% 11% 0% 4% 6% 0% 24% 20% 0% 8% 7% 10% 0% 16% 45% 27% 24% 59% 23% 31% 63% 88% 34% 24% 58% 48% 84% 55% 73% 76% 41% 77% 69% 37% 13% 66% 76% 42% 52% 78% 22%

73% 27%

68% 32%

55% 45%

53% 47%

51% 49%

Slavery, Forced Labour and Resistance in Indian Ocean Africa and Asia1 Gwyn Campbell and Edward A. Alpers

Historiographical Background

Conventional Western historiography based on the Atlantic slave system has characterized the slave as a chattel who could be sold or transferred at will by the owner, and a capital asset who, through coercion, bore dividends both in productive labour and through reproduction. Some authors have considered that economies in which slave labour predominated were characterized by a ‘slave mode of production’. The slave/free divide, the major social division in slave-owning societies, was accentuated by the slave’s legal inferiority and his/ her foreign and allegedly uncivilized origin. Only in exceptional cases was a slave granted freedom. Again, conventional historiography portrays the forced labour policies of colonial regimes as highly exploitative, fomenting resentment, the formation of nationalism and a desire for independence. In such historical contexts, rebellion was to be expected. This is because there comes a point of exploitation at which the slave or unfree worker has nothing to gain from complying with the labour system he/she is subject to, and nothing to lose by rebelling against it. Thus there developed a vast range of reactions, from passive resistance to marronage and outright revolt. However, it is clear from this and other recent studies on slavery and unfree labour in Asia (including aboriginal Australia) and Indian Ocean Africa (IOA)2 that the range of reactions to forced labour regimes by those subject to them were much more varied than those presented in conventional literature on slavery.3 This, in turn, reflects the greater complexity of forced labour regimes in IOA Source: Campbell, Gwyn and Alpers, Edward A., “Introduction: Slavery, Forced Labour, and Resistance in Indian Ocean Africa and Asia,” Slavery & Abolition, 25(2) (2004): ix–xxvii. © Taylor & Francis Ltd, reprinted by permission of the publisher Taylor & Francis Ltd, http://www .tandfonline.com. 1  We would like to thank Michael Salman for his comments on the drafts of the introduction. 2  Indian Ocean Africa, hereafter IOA, is defined as those regions of continental Africa, from Egypt in the north to South Africa, bordering the Indian Ocean or its Red Sea extension, and the islands of the Western Indian Ocean. 3  See also: Edward A. Alpers, Gwyn Campbell and Michael Salman (eds.), Resisting Bondage in Indian Ocean Africa and Asia (London: Routledge, 2005). © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_045

1398

Campbell and Alpers

and Asia, some of which resembled those of the Atlantic world, but most of which contrasted sharply in their structure and operation. An important part of the challenge for historians of slavery and unfree labour is that received distinctions between ‘slave’ and ‘free’ are not particularly helpful tools of historical analysis in most regions of IOA and Asia. The concept of ‘slavery’ was highly specific to time and place, often embracing or overlapping with other forms of forced labour. Moreover, Western concepts of individual liberty were inconceivable in most IOA and Asian societies where the majority of people subject to slave or other forms of ‘unfree’ labour sought to better integrate themselves within existing systems, frequently by resisting its abuses or constraints. This was the goal even of some ‘rebellions’. Otherwise, protest was generally muted, frequently manifested in small-scale, short-term marronage. Major rebellions were the exceptions, often a reflection of largescale state exploitation, and were defined as ‘revolts’ as much, if not more, by the political authorities, European or indigenous, as by the rebels themselves.

Servitude in Indian Ocean Africa and Asia

Slavery and other regimes of servitude in IOA and Asia varied greatly both geographically and over time, rarely approximating to the image of New World plantation slavery except in a few sectors, represented for example by the Mascarene sugar islands of Réunion and Mauritius in the eighteenth and early nineteenth centuries, and the nineteenth-century date plantations in the Gulf and coffee plantations around Lake Tana in Ethiopia.4 Many female and male slaves were employed in activities removed from direct production, although many slave activities that were not directly productive, such as in the military, trade and administration, often had an important, sometimes vital economic impact. However, coercion, generally considered a defining feature of slave work, was not exclusive to slavery. Coerced labour was probably extracted from the majority of non-elite populations of Africa and Asia well into the twentieth century.5 As Roger Knight in his study of Javanese sugar-cane Cultuurstelsel 4  See the contributions in this volume by Edward A. Alpers and Richard Allen; Hussein Ahmed, ‘Benevolent Masters and Voiceless Subjects: Slavery and the Slave Trade in Southern Wallo, Ethiopia in the Nineteenth and Early Twentieth Centuries’, paper presented at the IC/ULBR. 5  See e.g. Robert Castle, Jim Hagan and Andrew Wells, ‘“Unfree Labor” on the Cattle Stations of Northern Australia, the Tea Gardens of Assam, and the Rubber Plantations of Indo-China, 1920–50’ in Alpers, Campbell and Salman (eds.), Resisting Bondage.

Slavery, Forced Labour and Resistance in IOA and Asia

1399

points out with reference to the Cuban sugar-cane plantations that used slaves, global comparisons can and should be made between productive regimes using slavery and similar regimes using non-slave but coerced labour.6 Again, the complexity and overlap between slave and non-slave systems of servile labour defy the traditional binary definitions used to analyse the Atlantic system and frequently extended by Western scholars to other regions of the world. Thus slavery in IOA and Asia cannot be understood fully in terms of ‘open’ and ‘closed’ systems, characterized respectively by slave assimilation into the dominant society and exclusion from it. The key question in societies using forms of slave and bonded or servile labour was not an individual’s legal status, as is emphasized for the Americas. Social mobility was built into even so-called ‘closed’ systems like China and Madagascar, and an individual’s status could change, for the better or for the worse.7 Rather, a social hierarchy of dependence in which each possessed a status with concomitant rights and obligations embraced both slaves and non-slaves. As the exact meanings of each status varied geographically and over time, any definition of slavery in IOA and Asia must of necessity be highly culture and time specific. Institutions of servitude, including slavery and debt bondage, are best judged against this backdrop. The key issue throughout was the dynamics of the relationship between people of inferior and superior status, and the extent to which a superior could and did offer an inferior protection against external threats in return for labour and other services.8 Of notable significance here were the human (wars) and natural disasters (notably cyclones, volcanoes and the ‘Southern Oscillation’ or ‘El-Niño’ (‘ENSO’) effect that constituted ever-present threats over much of the Africa–Asia region.9 In good times, the dominant indigenous social group attempted to increase the number of 6  G. Roger Knight, ‘Sugar and Servility: Themes of Forced Labour, Resistance and Accommo­ dation in Mid-Nineteenth Century Java’ in Alpers, Campbell and Salman (eds.), Resisting Bondage. 7  James L. Watson, ‘Transactions in People’ in Watson (ed.), Asian and African Systems of Slavery (Berkeley and Los Angeles: University of California Press, 1980), pp. 227, 229–30, 235–6, 240–44; chapter by Michael Lambek. For African comparisons, see Igor Kopytoff and Suzanne Miers, ‘African “Slavery” as an Institution of Marginality’ in idem (eds.), Slavery in Africa: Historical and Anthropological Perspectives (Madison: University of Wisconsin Press, 1977), pp. 10–11, 25. 8  See, e.g., Anthony Reid, ‘Introduction: Slavery and Bondage in Southeast Asian History’ in Reid (ed.), Slavery, Bondage and Dependency in Southeast Asia (St. Lucia: University of Queensland Press, 1983), pp. 6–8; Kopytoff and Miers, ‘African “Slavery” ’, pp. 17–19. 9  Gwyn Campbell, ‘Introduction’ to Campbell (ed.), Abolition and Its Aftermath in Indian Ocean Africa and Asia (London: Routledge, forthcoming).

1400

Campbell and Alpers

its dependents, but in bad times they sold the most expendable and marketable to further the survival of the group. Some voluntarily entered slavery. Others ‘sold’ or ‘pawned’ their children.10 Possibly the majority entered slavery through indebtedness, a process largely ignored in the conventional historiography of slavery. Violence Traditional slavery literature assumes that violence was universally employed in order to extract labour from slaves. In IOA and Asia, violence or the threat of violence could certainly induce slaves to work, but was relatively little employed outside European- and Arab-managed plantations where economies of scale made higher levels of coercion profitable. This is not to deny that in some places, and at certain times, harsh working conditions existed and could provoke revolt, suicide and attempts to curtail reproduction; low birth rates, characteristic of the Mauritian slave plantations, may have marked even milder slave regimes, as in Indonesia and the Gulf.11 However, slaves represented a capital asset the value of which was worth maintaining or even enhancing. Indeed, maximum slave productivity could only be achieved through acknowledging the essential humanity of slaves.12 Nevertheless, as Frederick Cooper has noted for Zanzibar, ‘Violence was a part of paternalism’,13 which was the hallmark of most slave systems in IOA and Asia.

10  A theme covered in the international conference on Children in Slavery, held in Avignon from 19–21 May 2004. 11  Peter Boomgaard, ‘Human Capital, Slavery and Low Rates of Economic and Population Growth in Indonesia, 1600–1910’ in Gwyn Campbell (ed.), The Structure of Slavery in Indian Ocean Africa and Asia (London and Portland, Or: Frank Cass, 2004), pp. 83–96; Edward A. Alpers, ‘Flight to Freedom: Escape from Slavery among Bonded Africans in the Indian Ocean World’ in Campbell (ed.), Structure of Slavery, pp. 51–68; Abdul Sheriff, ‘The Slave Trade and its Fallout in the Persian Gulf’ in Campbell (ed.), Abolition and Its Aftermath. 12  Martin Klein, ‘Introduction: Modern European Expansion and Traditional Servitude in Africa and Asia’ in Klein (ed.), Breaking the Chains: Slavery, Bondage and Emancipation in Modern Africa and Asia (Madison: University of Wisconsin Press, 1993), pp. 11–12; Claude Meillassoux, The Anthropology of Slavery: The Womb of Iron and Gold (Chicago: University of Chicago Press and London: Athlone Press, 1991), pp. 9–10. 13  Frederick Cooper, Plantation Slavery on the East Coast of Africa (New Haven and London: Yale University Press, 1977), p. 154.

Slavery, Forced Labour and Resistance in IOA and Asia

1401

The lot of female slaves, who were less likely to be sold than male slaves,14 could also be generally easier than for males. There existed for favourite concubines and their children born into important families special routes of social mobility. Rulers and the wealthy, most of whom were men, surrounded themselves with female slaves who, as secondary wives, concubines, entertainers and domestic servants, enjoyed a lifestyle and, Anthony Reid argues for Southeast Asia, a respect superior to that of female peasants.15 There are instances of concubines in the Middle East sending for family members to join them—albeit as non-slaves.16 However, this is not to underestimate the ‘labour’ of such female slaves. As Janet Hoskins shows, ‘slaves of the house’ in Sumba were sexually available to males of the owning family. More importantly, their ability to experience social mobility through marriage depended upon agency and strategic calculations in a society in which both slavery and marriage practices were arenas of power and status.17 Furthermore, Indrani Chatterjee observes of the systems of female sexual labour in colonial India that ‘The system did not allow for gendered tenderness: male and female holders and hirers were equally merciless, or arbitrarily benevolent’.18 Enslavement and surveillance costs were higher for male than for female or child slaves, and males were more likely to revolt or attempt escape. Once a decision had been made to enslave rather than kill male captives, their captors and subsequent owners were inclined to keep them in good health and not mistreat them, or use excessive punishment, as this could lead to the slave committing suicide, seeking revenge or fleeing.19 Thus most European colonial authorities legislated for relatively mild treatment of slaves, although maroons were mercilessly punished. Indeed, the Dutch and English East India Companies reserved for government authorities the meting out of severe punishments to slaves. However, private European masters generally needed little prompting. They were aware that many mistreated slaves would flee, and that some might seek revenge.20 Malagasy and Sulu slaves, in particular, had 14  Reid, ‘Introduction’, pp. 25–6; Ahmed, ‘Benevolent Masters’. 15  Reid, ‘Introduction’, pp. 25–6; Joseph Miller, ‘A Theme in Variations: A Historical Schema of Slaving in the Atlantic and Indian Ocean Regions’ in Campbell (ed.), Structure of Slavery, pp. 169–94. 16  Suzanne Miers, ‘Slavery and the Slave Trade in Saudi Arabia and the Arab States on the Persian Gulf 1921–1963’ in Campbell (ed.), Abolition and Its Aftermath. 17  Contribution in this volume by Janet Hoskins. 18  Indrani Chatterjee, Gender, Slavery and Law in Colonial India (New Delhi: Oxford University Press, 1999), p. 234. 19  Boomgaard, ‘Human Capital’. 20  Alpers, ‘Flight to Freedom’; Reid, ‘Introduction’, p. 16.

1402

Campbell and Alpers

a reputation for rebelliousness and violence.21 In the larger Indonesian city ports lacking the centralized authorities necessary to ensure the recapture of runaways, owners treated domestic slaves particularly well in order to retain them.22 For owners, a successful escape meant the loss of the costs of acquisition and maintenance of the slave. Recapture and subsequently heightened surveillance would significantly add to the overall ‘slave’ cost.

Class Consciousness amongst Slaves and Bonded Labour

Authors including Claude Meillassoux have argued that slaves comprised a social class oppressed and exploited by the slave-owning class.23 In the process of class formation, the development of class-consciousness was critical. In the case of slave class formation, the development of slave class-consciousness required the existence of large concentrations of slaves, experiencing similar work and living conditions, with clearly identifiable ‘opponents’ in the form of the slaveholders, and a leadership that articulated their interests. In IOA and Asia, the greatest potential for the development of a slave class was on plantations where there existed relatively large concentrations of slave or bonded labour. However, there is little indication of a slave consciousness within the ranks of plantation labourers, even on the Mascarenes. Clare Anderson’s study of Indian convict ‘slave’ labour on Mauritius in the first half of the nineteenth century reveals group consciousness, but it was caste rather than class based, so could not act as a unifying force for unfree labour.24 The same was true of labour on the colonial plantations of tea in Assam,25 as initially of rubber in French Indo-China, although the influence there of communist cadres eventually led to many plantation workers being embraced by a

21  Richard Allen, ‘The Mascarene Slave-Trade and Labor Migration in the Indian Ocean during the Eighteenth and Nineteenth Centuries’ in Campbell (ed.), Structure of Slavery, pp. 33–50; Alpers, ‘Flight to Freedom’; Nigel Worden, ‘Indian Ocean Slavery and its Demise in the Cape Colony’ in Campbell (ed.), Abolition and Its Aftermath; see also Reid, ‘Introduction’, pp. 3, 19. 22  Reid, ‘Introduction’, p. 15. 23  Meillassoux, Anthropology of Slavery. 24  Clare Anderson, ‘The Bel Ombre Rebellion: Indian convicts in Mauritius, 1815–53’ in Campbell (ed.), Abolition and Its Aftermath. 25  Keya Dasgupta, ‘Plantation Labor in the Brahmapura Valley: Regional Enclaves in a Colonial Context, 1881–1921’ in Campbell (ed.), Abolition and Its Aftermath.

Slavery, Forced Labour and Resistance in IOA and Asia

1403

revolutionary consciousness.26 An added complication in the case of bonded labour was the interest of government, indigenous and colonial, in suppressing any expression of a collective consciousness amongst bonded labourers.27 Outside the plantations, the vast variety of activities engaged in by slaves and bonded workers, and their disparate and often isolated living and working conditions, hindered any rise of ‘slave’ or ‘bonded labour’ consciousness.28 In extreme circumstances, such as on the Australian ranches where proprietors encouraged drug-related dependency, the social basis for collective consciousness and action simply failed to materialize.29 There existed, however, exceptions where slave consciousness developed. One was the secret societies in Korea that embraced slaves in a tight organizational structure that engaged in violent anti-establishment activities, including the formation of political reform factions. However, a combination of government suppression and the progressive transformation from slave to small farmer status following the 1801 liberation of state slaves effectively prevented the development of a slave class consciousness.30 Another exception was the royal slaves in Aden Protectorate who in 1943 ‘were sufficiently numerous, in close enough contact, and bound by common interests to act en masse’ in rebellion. However, the primary aim of that revolt was not for but against freedom.31 Indeed, the goal of most servile labour was to secure a niche within the dominant society, improve that position over time and, if granted non-slave status, to assume a new ethnicity. Individual slaves therefore sought to forge linkages not with other slaves but with slaveholders who alone could ameliorate their conditions and station. Central to this process was acculturation. For most owners, it was vital that the slave speak the local language. The general 26  Castle, Hagan and Wells, ‘“Unfree Labor” ’. 27  Anderson, ‘The Bel Ombre Rebellion’; Dasgupta, ‘Plantation Labor’; Castle, Hagan and Wells, ‘Unfree Labor’; Gwyn Campbell, ‘Slavery and Fanompoana: The Structure of Forced Labour in Imerina (Madagascar), 1790–1861, Journal of African History, 29, 2 (1988), pp. 463–86. 28  See e.g. contribution by Suzanne Miers. 29  Castle, Hagan and Wells, ‘Unfree Labor’; Ahmed, ‘Benevolent Masters’. This same point was made by Jan-Georg Deutsch, ‘Absence of Evidence is No Proof. Slave Resistance under German Colonial Rule in East Africa’, paper presented at the IC/ULBR. 30  Kim, Bok-Rae, ‘Korean Nobi Resistance under the Chosun Dynasty (1392–1910)’ in this volume. 31  Suzanne Miers, ‘Slave Rebellion and Resistance in the Aden Protectorate in the MidTwentieth Century’, in this volume.

1404

Campbell and Alpers

preference was thus for young slaves who could learn quickly. At the very minimum, slaves needed to understand orders but, in marked contrast to slaves in the Americas, many—if not the majority—of slaves in IOA and Asia were employed in sensitive positions—within the household, court, administration or commerce—where they needed to fully comprehend and assimilate the cultural and ideological values of the slave-holding society. In such circumstances, the native languages of slaves quickly ceded to that of the dominant society. In Southeast Asia acculturation was facilitated by the fact that many slaves originated from societies with linguistic and cultural backgrounds similar to that of the dominant society.32 In China and Imperial Madagascar, language classes were established in slave reception camps to facilitate the process.33 Slaves also often accepted the religious ideology of slaveholders, which in turn justified slavery. Vestiges of cultural origin, such as the Zar healing ceremony practised by ex-slaves in the Gulf and the atete cult introduced by southern Ethiopian female slaves into local practices of Zar in Wallo, were insufficient bases for a separate consciousness to be maintained, although they did create important space for enslaved people within slave societies.34 In nineteenth-century Madagascar, many slaves sought religious roles as ‘traditional’ mediums.35 In China and Korea, Confucian concepts of hierarchical social, familial and sexual relations, impregnated into the minds of daughters sold into slavery, were vital to the maintenance of the system.36 As crucial was the conversion of imported child brides and concubines to local belief and 32  Reid, ‘Introduction’, pp. 13, 25–6. 33  Karine Delaye, ‘Slavery and Colonial Representations in Indochina from the Second Half of the Nineteenth to the Early Twentieth Century’ in Campbell (ed.), Structure of Slavery, pp. 129–42; Gwyn Campbell, ‘Madagascar and the Slave Trade, 1810–1895’, Journal of African History, 22 (1981), p. 224. 34   Sheriff, ‘The Slave Trade and its Fallout’; Behnaz Mirzai Asl, ‘Identity, Cultural Transformation, and the African Diaspora in Iran’, paper presented to the symposium, ‘Africans Across the Indian Ocean’, Spelman College, Atlanta, April 2004; Ahmed, ‘Benevolent Masters’. 35  Gwyn Campbell, ‘Abolition and its Aftermath in Madagascar, 1877–1949’ in Campbell (ed.), Abolition and Its Aftermath; Campbell, ‘Crisis of Faith and Colonial Conquest. The Impact of Famine and Disease in late nineteenth-century Madagascar’, Cahiers d’Études Africaines, 32, (3)127 (1992), pp. 409–53. 36  Kim, Bok-Rae, ‘Nobi: A Korean System of Slavery’ in Campbell (ed.), Structure of Slavery, pp. 155–68; James Francis Warren, Chinese Prostitution in Singapore: Recruitment and Brothel Organisation’ in Maria Jaschok and Suzanne Miers (eds.), Women and Chinese Patriarch: Submission, Servitude and Escape (London and New Jersey: Zed Books, 1994), pp. 79–80.

Slavery, Forced Labour and Resistance in IOA and Asia

1405

value systems so that these might in turn be transmitted to their children.37 Acceptance of the dominant society’s ideology was a pre-requisite for slaves employed in sensitive posts in the royal household, army and administration. Again, internal divisions within slave groups militated against the development of slave class-consciousness. Ethnic differences were often apparent; for instance, the fierce ethnic loyalty of Bugis slaves in Sulu separated them as much from other slaves as from their owners.38 This was also evident amongst Malagasy slaves in North America who sought to distinguish themselves from slaves of ‘African’ descent.39 However, in contrast to the Americas, creolization in those African and Asian societies under review tended to develop slowly and rarely engendered a ‘Creole’ consciousness, with the notable exception of the Mascarenes.40 In general, slaves originating from the slave holding society were considered superior in status to those procured from outside. In a movement resembling that of immigrant waves into nineteenth century America, new arrivals raised the status of resident slaves. Thus the mass influx of African slaves in nineteenth-century Madagascar placed the newcomers at the bottom of the slave hierarchy.41 Similarly placed were recaptured fugitive slaves, whose inferior status was often visibly clarified through branding or tattoos.42 A sign of superior slave status was avoidance of menial or ritually degrading activities: Those engaged in harsh manual tasks such as mining, irrigation work, agriculture and diving were considered inferior to those employed within a household or court, or by a trader. Moreover, those at the top of the slave hierarchy, and any other slaves who could afford to do so, became slaveholders.43 In addition, while ‘bonded labour’ often experienced brutally exploitative work regimes that in some cases, as in Imperial Madagascar and amongst 37  Kim, ‘Nobi’; Angela Schottenhammer, ‘Slaves in Late Imperial China, from the Late 18th to the Early 20th Centuries’ in Campbell (ed.), Structure of Slavery, pp. 143–54; Sheriff, ‘The Slave Trade and its Fallout’. 38  James Francis Warren, The Structure of Slavery in the Sulu Zone in the Late 18th and 19th Centuries’ in Campbell (ed.), Structure of Slavery, pp. 111–28. 39  Wendy Fall, ‘Madagascar in Maryland and Virginia in the Eighteenth and Nineteenth Centuries’. Paper presented at the IC/ULBR. 40  Nigel Worden, ‘Revolt in Cape Colony Slave Society’ in Alpers, Campbell and Salman (eds.), Resisting Bondage; Fall, ‘Madagascar in Maryland and Virginia’; Megan Vaughan, Creating the Creole Island: Slavery in Eighteenth Century Mauritius (Durham, NC: Duke University Press, forthcoming). 41  Campbell, ‘Abolition and its Aftermath in Madagascar’; Ahmed, ‘Benevolent Masters’. 42  Reid, ‘Introduction’, p. 12. 43  See e.g. Schottenhammer, ‘Slaves and Forms of Slavery’; Klein, ‘Introduction’, p. 7.

1406

Campbell and Alpers

Aboriginal ranch hands, led to self-mutilation as a means of protest and escape,44 slaves in Africa and Asia enjoyed a range of traditional and prescribed rights unknown on the American plantations. In some regions (outside the Mascarenes, where European law was applied), treatment of slaves was tempered by local economic and political forces.45 Even in Korea and China, where the most extreme systems of hereditary slavery were practised, slaves were immune from state corvées, could be punished but not killed by their owners, and their marriages were in general respected. Such rights, it could be argued, meant that they were not true outsiders as they had entered into the dominant society’s system of reciprocity.46 In such contexts, it could prove futile to search for a slave class or ‘slave modes of production’, constructs which, as the case of Imperial Madagascar illustrates, sometimes bear little relation to historical reality.47

The Gender Issue

In contrast to the Atlantic slave trade, in which predominantly male Africans were shipped to plantations to serve as field hands, the majority of slaves trafficked in IOA and Asia were females, notably girls and young women, valued particularly for their sexual attractiveness and reproductive capacity.48 The issue was complicated by the overlap between slavery and forms of debt-bondage that involved the sale or involuntary transfer of people, including

44  For Madagascar, see Gwyn Campbell, Brigandry and Revolt in Pre-Colonial Africa: Imperial Madagascar, 1750–1900 (Rochester: University of Rochester Press, forthcoming), ch. 7; for Australia see Castle, Hagan and Wells, ‘Unfree Labor’. 45  Boomgaard, ‘Human Capital’; Worden, ‘Indian Ocean Slavery’. 46  Schottenhammer, ‘Slaves and Forms of Slavery’; Kim, ‘Nobi’; see also Michael Salman, ‘Colonialism, Nationalism and the Meaning of Slavery: the Genealogy of “an Insult to the American Government and to the Filipino People” ’ in Campbell (ed.), Abolition and Its Aftermath. 47  Paul E.  Lovejoy, Transformations in Slavery (Cambridge: Cambridge University Press, 1983), pp. 234, 238–9; Klein, ‘Introduction’, pp. 10–11; Maurice Bloch, ‘Modes of Production and Slavery in Madagascar’, in Watson (ed.), Asian and African Systems of Slavery, pp. 110– 12; Campbell, ‘Abolition and its Aftermath in Madagascar’; Campbell, ‘Madagascar and the Slave Trade’; Campbell, ‘Unfree Labor, Slavery and Protest in Imperial Madagascar’ in Alpers, Campbell and Salman (eds.), Resisting Bondage. 48  Miller, ‘Theme in Variations’; Goody, ‘Slavery in Time and Space’, pp. 20–21.

Slavery, Forced Labour and Resistance in IOA and Asia

1407

‘serfs’ (Asia and Africa), ‘pawns’ (Africa) and ‘mui tsai’ girl servants (China).49 For instance, in Africa, a kinship group might during a period of hardship, in return for goods or money, transfer to another lineage kinship members who, if not redeemed, were retained by the creditor lineage.50 In China, the tendency in bad times was to sell daughters or secondary wives.51 In many African and Asian societies, parents also offered their children for adoption in exchange for money.52 This was of critical importance in the bond-woman’s attitude to revolt. From early times females in bondage worked as concubines, entertainers, prostitutes and domestic servants. Although some also laboured as water carriers, and in agriculture, textile production and mining,53 many entered into close social and often sexual relations with their owners that frequently resulted in bearing children fathered by the master. In Wallo, Ethiopia, the offspring of such unions were free and their children could inherit equally with the children of the legal wife of their father.54 Many women were thus obliged to act both for their own immediate well-being and for the future security, welfare and prospects of their children.55 In southern Somalia during the period of Italian colonialism, women were forced into ‘Italian marriage’ (as it was known to the Zigula of the Gosha region) with men who themselves were pressed into agricultural labour. This bizarre form of union which enabled men to avoid the expense of a properly arranged marriage between families56—something that made slave wives

49  For serfs, see e.g. Kim, ‘Nobi’; Klein, ‘Introduction’, pp. 5–6; for pawns, see Paul E. Lovejoy and Toyin Falola (eds.), Pawnship, Slavery, and Colonialism in Africa (Trenton, NJ: Africa World Press, 2003); and for ‘mui tsai’ see Watson, ‘Transactions in People: The Chinese Market in Slaves, Servants, and Heirs’ in Watson (ed.), Asian and African Systems of Slavery, pp. 240–45. 50  Kopytoff and Miers, ‘African Slavery’, pp. 10–11. 51  Watson, ‘Transactions in People’, pp. 227–30. 52  See e.g. Boomgaard, ‘Human Capital’. 53  Goody, ‘Slavery in Time and Space’, pp. 21, 32. 54  Ahmed, ‘Benevolent Masters’. 55  For a broader reading of women’s reproductive household labor, see Margaret Strobel, ‘Slavery and Reproductive Labor in Mombasa’ in Claire C. Robertson and Martin A. Klein (eds.), Women and Slavery in Africa (Madison: University of Wisconsin Press, 1983), pp. 111–29. 56  See Francesca Declich, ‘Unfree Labour, Forced Labour and Resistance among the Zigula of the Lower Juba’ in Alpers, Campbell and Salman (eds.), Resisting Bondage.

1408

Campbell and Alpers

equally attractive to some in Wallo57—undermined traditional gender relations and marriage customs. Manumission Nevertheless, a number of IOA and Asian societies developed the legal structure for manumission. Individual manumission/redemption posed no threat to the slave system: The number of slaves affected was limited, sometimes by law; the possibility of manumission acted as an important psychological safety valve that helped temper the slaves’ impulse to revolt; and manumissions encouraged slave imports to replace those thus ‘freed’.58 Even so, manumission was a relatively rare occurrence. Moreover, ‘liberty’ in an abstract sense was not necessarily as appealing as Western minds have imagined. For most IOA and Asian populations, security, food and shelter rather than an abstract concept of liberty, were the primary aims. ‘Liberty’ in the sense of individual freedom from inherited status and responsibilities, would have effectively destroyed the web of obligations that offered protection from man-made and natural dangers.59 This helps explain in IOA and Asian societies the remarkable absence of class-consciousness and of revolt amongst slaves who generally sought to integrate themselves into the slave-holding society that provided them with basic sustenance and sometimes the chance of an enhanced lifestyle.60 It also explains why some slaves who were presented with the opportunity to gain ‘freedom’ through manumission or redemption rejected it in favour of retaining their slave status.61 Marronage Marronage occurred on a scale sufficient to worry authorities, both in plantation societies such as the Mascarenes, and in non-plantation economies across 57  Ahmed, ‘Benevolent Masters’. 58  Suzanne Miers, ‘Slavery: A Question of Definition’ in Campbell (ed.), Structure of Slavery, pp. 1–16; see also David Brion Davis, The Problem of Slavery in Western Culture (Harmondsworth: Penguin, 1970), pp. 70–72, 110–11, 289–97. 59  Boomgaard, ‘Human Capital’; Miers, ‘Slavery and the Slave Trade’; Salman, ‘Colonialism, Nationalism and the Meaning of Slavery’. 60  Kim, Bok-Rae, ‘Korean Nobi Resistance’. 61  Suzanne Miers, ‘Slave Rebellion’.

Slavery, Forced Labour and Resistance in IOA and Asia

1409

region.62 In Korea, where the crown tacitly encouraged flight, up to 8 per cent of all privately owned slaves ran away from their owners.63 In Mauritius, at the end of the eighteenth century, 4–5 per cent of slaves deserted their masters each year, rising to 11–13 per cent in the early 1820s.64 Comparisons with bonded labour reveal similar statistics: thus on Mauritius in the 1840s, 6–17 per cent of Indian indentured workers engaged in short-term absenteeism.65 Nevertheless, escape tended to be a spontaneous, individual and short-term reaction to mistreatment. It generally occurred in seasons when wild fruits and roots, and animals, were comparatively common. Thus, in the Australian bush, temporary marronage, or ‘walkabouts’, tended to occur immediately after the rainy season had induced fresh plant growth.66 The same pragmatic concerns were evident elsewhere. For example, Cape slaves tended to abscond in times of harvest, when labour was intensive, and Javanese peasants in 1944, faced with compulsory seizures of rice by the occupying Japanese, rebelled for the duration of the rice harvest season,67 although for bonded labour with a stake in harvests, the reverse was the case. For instance, the 1856 Mayotte ‘rebels’ returned home in time to harvest their fields.68 Flight was rare for female slaves who preferred to remain behind with their children, although during the nineteenth century some women in Iran and Arabia seized the opportunity offered by abolition to flee their masters and seek refuge in the British Consulate.69 Flight was most common amongst newly acquired young adult male slaves. In one study of slave runaways in the 62  Anderson, ‘The Bel Ombre Rebellion’; Richard Allen, ‘A Serious and Alarming Daily Evil: Marronage and Its Legacy in Mauritius and the Colonial Plantation World’, in this volume. Edward A. Alpers, ‘The Idea of Marronage: Reflections on Literature and Politics in Reunion’, in this volume. Alpers, ‘Flight to Freedom’; see also Allen F. Isaacman, Mozambique: The Africanization of a European Institution, The Zambesi Prazos, 1750–1902 (Madison: The University of Wisconsin Press, 1972), pp. 107–8; Campbell, Brigandry and Revolt. 63  Kim, ‘Nobi’. 64  Richard Allen, ‘A Serious and Alarming Daily Evil’. 65  Ibid. 66  Castle, Hagan and Wells, ‘Unfree Labor’. 67  Worden, ‘Revolt in Cape Colony Slave Society’; Shigeru Sato, ‘Forced Laborers and their Resistance in Java under Japanese Military Rule, 1942–1945’ in Alpers, Campbell and Salman (eds.), Resisting Bondage. 68  Isabelle Denis, ‘Forced Labour and the Mayotte Revolt of 1856’ in Alpers, Campbell and Salman (eds.), Resisting Bondage. 69  See Asl, ‘Slave Emancipation in Iran: Gender and Freedom’, paper presented at the IC/ ULBR.

1410

Campbell and Alpers

Cape from 1806 to 1809, 90 per cent were male.70 Again, the comparison holds true for bonded labour. Thus, on Mauritius in the mid-nineteenth century, the vast majority of Indian indentured workers to engage in short-term absenteeism were young adult males.71 However, escapees were vulnerable to recapture as few possessed local kin or spoke the local language. Also, they could often be easily identified by their physiognomy or by their ‘ethnic’ brand marks. Fugitive slaves risked reenslavement by ‘free’ people unconnected to their owner. In larger urban areas, runaway slaves might even seek the protection of alternative ‘masters’.72 Many who returned to their home communities, such as Filipino slaves fleeing Sulu, were rejected and subsequently lived a marginal existence on the periphery of the dominant society.73 Those runaway slaves wishing for permanent escape commonly joined maroon bands. Predominantly male, maroons inhabited geographically remote and often economically marginal areas where subsistence cultivation or foraging proved difficult.74 To survive, most were forced to raid surrounding communities, often kidnapping people for sale. The authorities crushed most maroon communities, although a minority flourished. Some transformed themselves into major slave traders.75 The colonial advance in the late nineteenth century weakened slave-holding powers. Where this happened, as on the Benadir coast 70  Worden, ‘Revolt in Cape Colony Slave Society’. 71  Allen, ‘A Serious and Alarming Daily Evil’. 72  See Vaughan, ‘Maroons and Masquerade: Mauritius in the Eighteenth Century’, paper presented at the IC/ULBR; Allen, Slaves, Freedmen, and Indentured Laborers, pp. 46, 50–51. See also Declich, ‘Unfree Labour’. 73  Warren, ‘Structure of Slavery’. 74  See, e.g., Charles Camara, ‘The Siddis of Uttara Kannada: History, Identity and Change among African Descendants in Contemporary Karnataka’, in Amy Catlin-Jairazbhoy and Edward A. Alpers (eds.), Sidis and Scholars: Essays on African Indians (Delhi: Rainbow Publishers/Trenton, NJ: Red Sea Press, 2004), pp. 100–101. 75  Miers, ‘Slavery and the Slave Trade’; Martin Klein, ‘The Emancipation of Slaves in the Indian Ocean’ in Campbell (ed.), Abolition and Its Aftermath; Campbell, ‘Abolition and its Aftermath in Madagascar’; Omar Eno, ‘Abolition and its Impact on the Benadir Coast’ in Campbell (ed.), Abolition and Its Aftermath; Alpers, ‘Flight to Freedom’; see also Gwyn Campbell, ‘The Menalamba Revolt and Brigandry in Imperial Madagascar, 1820–1897’, International Journal of African Historical Studies 24, 2 (1991), pp. 259–91; Goody, ‘Slavery in Time and Space’, p. 29; Luísa Martins, ‘Esclavage et main-d’œuvre forcée d’après des sources orals recueillies dans la province de Nampulam, Mozambioque’, paper presented at the IC/ULBR; for comparisons in the Americas, see Richard Price (ed.), Maroon Societies: Rebel Slave Communities in the Americas, 3rd edn (Baltimore and London: The Johns Hopkins University Press, 1996).

Slavery, Forced Labour and Resistance in IOA and Asia

1411

of Somalia and in German East Africa, it could, at least temporarily, offer maroons greater possibilities of establishing independent existences.76 Indeed, as Jan-Georg Deutsch suggests for German East Africa, ‘The run-away slaves of the later nineteenth century became the labour migrants of the early twentieth century. Labour migration brought slavery to an end.’77 Revolt Slave revolt in IOA and Asia was remarkably rare for many of the same reasons that slave flight was limited: most slaves were women. Often involved in intimate relationships with their owners, and frequently offered greater opportunity to assimilate into the dominant society than male slaves, they were reluctant to take risks that might damage their children’s’ interests. Again, assimilation and ethnic or caste divisions, as well as ‘social heterogeneity’, hindered the development of a slave consciousness and leadership.78 As noted, rebel slaves often survived by participating as agents in the slave trade.79 Male slaves in Islamic India and Arabia were frequently employed in the military, but the use of slave soldiers was sometimes forbidden, as in Imperial Madagascar, for fear of a slave revolt.80 Even in ‘perhaps the most successful slave rebellion of all times’, that of the ‘Zanj’ in Iraq from 869 to 883 CE, it would appear that non-slave influences were crucial in promoting the revolt.81 Also, in some cases of revolt, it was not possible to distinguish between slaves and other forms of bonded labour. Thus in the short-lived revolt on 76  Alpers, ‘Flight to Freedom’; Declich, ‘Unfree Labour’; see also Eno, ‘“Gosha/Heer-Goleet” [people of the forest]: Runaway Slaves in the Juba Valley of Southern Somalia’, paper presented at the IC/ULBR; Suzanne Miers and Martin Klein, ‘Introduction’ to Miers and Klein (eds.), Slavery and Colonial Rule in Africa (London: Frank Cass, 1999), pp. 6–7. 77  Deutsch, ‘Absence of Evidence’. For the full study, see Deutsch, Slavery under German Colonial Rule in East Africa, c.1860–1914’, Habilitationsschrift, Humboldt University at Berlin, October 2000. 78  Deutsch, ‘Absence of Evidence’. See also Ahmed, ‘Benevolent Masters’. 79  Campbell, Brigandry. 80  Chauhan, Africans in India; Shanti Sadiq Ali, The African Dispersal in the Deccan (Hyderabad: Orient Longman, 1996); Campbell, ‘Madagascar and the Slave Trade’, p. 209; Campbell, ‘Slavery and Fanompoana’. 81  Alan Fisher, ‘Zanj’ in Paul Finkelman and Joseph C. Miller (eds.), Macmillan Encyclopaedia of World Slavery (New York: Macmillan Reference USA Simon & Schuster Macmillan, 1998) vol. II, p. 967; Alexandre Popovic, The Revolt of African Slaves in Iraq in the 3rd/9th Century (Princeton: Markus Weiner, 1999), p. 22.

1412

Campbell and Alpers

Mayotte from March to April 1856, rebels included both east African ‘slaves’ recruited under the infamous French contract labour schemes of the epoch and local Sakalava driven to revolt by a French colonial regime that encouraged French planters to seize the land of, and impose forced labour on, the local population.82 Moreover, ‘protests’ were often intended to reform the existing system, rather than overthrow it. It might be inferred from the evidence that systematic denial of non-violent protests against violations of established and legal rights led to the development of other strategies of ‘resistance’.83 Richard Allen cites the example of slaves on Mauritius in the 1820s who preferred a prison sentence (for maroonage) to working on an estate where they received insufficient rations.84 Reform was also the underlying motivation of those Zigula who organized themselves in a strike against Italian labour exactions in the later 1930s.85 The aim of reform even inspired violent rebellion. Such was the case, for instance, of the short-lived revolts on Mayotte (1856) and in the Cape (1808 and 1825).86 These examples reflect the complexity of traditional labour relations in IOA and Asia, in which workers and owners were tied into an unequal system of rights and obligations. Moreover, this system was perpetuated by many European colonial regimes, using forced labour, who invariably worked through indigenous collaborators.87 Nevertheless, there were cases of successful revolt in the sense of the formation of relatively stable communities composed of runaways. Intrinsically, their former ‘masters’ regarded such communities as a challenge to the legally constituted authority. As such, they were declared to be rebels, subject to the extreme penalties of the law. As true ‘outsiders’, they were invariably forced into a covert existence of theft, brigandry and often slave trading.88 Nevertheless, they gained some ‘popular’ legitimacy in the eyes of slave and bonded labour that remained in servitude. Thus the slave runaways of Table 82  Denis, ‘Forced Labour and the Mayotte Revolt’. 83  Allen, ‘A Serious and Alarming Daily Evil’; Kim, ‘Korean Nobi Resistance’. 84  Allen, ‘A Serious and Alarming Daily Evil’. 85  Declich, ‘Unfree Labour’. 86  Worden, ‘Revolt in Cape Colony Slave Society’; Denis, ‘Forced Labour and the Mayotte Revolt’. 87  Eric Jennings, ‘Forced Labor in Madagascar under Vichy, 1940–42: Autarky, Requisitions, and Resistance on the ‘Red Island’ in Alpers, Campbell and Salman (eds.), Resisting Bondage; Knight, ‘Sugar and Servility’. 88  Martins, ‘Esclavage et main-d’œuvre forcée’.

Slavery, Forced Labour and Resistance in IOA and Asia

1413

Mountain traded with the slaves of Cape Town,89 the ‘Refugee Republics’ of Imperial Madagascar cooperated with anti-Merina forces, free and bonded,90 and the WaGosha maroon communities of the Juba River valley in southern Somalia fought against Italian encroachment.91 Of critical importance to the issue of rebellion is the role of the state, which possessed the greatest potential for coercive exploitation of the people subject to its control. The literature has tended to emphasize the role of European colonial regimes that supported the interests of European settlers and/or plantation owners and companies against those of the local population. However, indigenous states, notably those engaged in ‘secondary imperialism’, were frequently as, if not more, exploitative. Indeed, it was common from the late eighteenth century for the traditional system of dependency to be appropriated by the state, whether indigenous or colonial, in various forms of forced labour, some remunerated and others not.92 An interesting variation on this theme is the overthrow of Dutch rule in Java by the Japanese who, while maintaining the myth of ‘liberating’ the Javanese, provoked revolt through maintaining systems of economic exploitation.93 This, in turn, needs to be put in the context of the emergence and rapid development of the international economy during the nineteenth century and subsequent evolution towards full globalization. Invariably, this was accompanied in its initial stages by a huge growth in demand for cheap labour—a feature that has become an embedded characteristic of Third World economies. Given the systematic lack of capital resources, states—indigenous and European—who attempted to exploit the enormously expanded commercial opportunities presented by the growing international economy inevitably turned to labour-intensive means of production. In view of the initial lack of labour, this entailed the implementation and exploitation of forced labour. Indeed, in most African and Asian contexts, probably the greatest revolts of servile people, like that of the Khoi and Xhosa in South Africa in 1850–53, and the 1895–97 Menalamba uprising and 1947 revolt in Madagascar were

89  Worden, ‘Revolt in Cape Colony Slave Society’. 90  Campbell, ‘Unfree Labor, Slavery and Protest’. 91  Declich, ‘Unfree Labour’. 92  See e.g. Sato, ‘Forced Laborers’; Knight, ‘Sugar and Servility’; Campbell, ‘Unfree Labor, Slavery and Protest’; and Ahmed, ‘Benevolent Masters’. 93  Sato, ‘Forced Laborers’; Suzanne Miers, Slavery in the Twentieth Century: The Evolution of a Global Pattern (Walnut Creek, CA: AltaMira Press, 2003), pp. 66–86, 174–9, 239–53, 300–303.

1414

Campbell and Alpers

instigated by ‘unfree’ peoples subject to ‘state’ exploitation, whether European or indigenous.94

Resistance to ‘Freedom’

Another aspect of ‘slavery’ and ‘resistance’ is resistance to manumission and abolition. Traditional historiography has emphasized the resistance of indigenous authorities to moves to emancipate slaves. For instance, Imperial Madagascar and Ethiopia adopted anti-slave trade legislation, but slave trading and slavery persisted in Madagascar until the French abolished slavery in 1896, and in Ethiopia until the Second World War.95 Again, abolition proclaimed in China just prior to the end of Manchu rule in 1911 had little impact on rural areas until after the communist seizure of power in 1949.96 In traditional historiography, Islam represented the forces most opposed to abolition. Nevertheless, although most Muslims defended slavery, some violently, because the institution was legitimized by the sharia, debate over slavery never ceased because the main orthodox Islamic legal schools pronounced freedom to be the normal condition of humanity. Some unorthodox Islamic groups, notably the Ismaili Druzes sect, opposed slavery from their foundation in the eleventh century.97 Thus in 1808, Muslim muftis in Calcutta ruled that sale of family members and enslavement for debt, possibly the main cause of enslavement, was wrong,98 while ‘modernist’ Islam supported abolition.

94  Worden, ‘Revolt in Cape Colony Slave Society’; Jennings, ‘Forced Labor in Madagascar’; Campbell, ‘Unfree Labor, Slavery and Protest’; see also Knight, ‘Sugar and Servility’; Gwyn Campbell, ‘Coffee Production in Madagascar’ in William Gervase Clarence-Smith and Steven Topik (eds.), Coffee under Colonialism and Post-Colonialism: The Global Coffee Economy in Africa, Asia, and Latin America, 1500–1989 (Cambridge: Cambridge University Press), pp. 67–99. 95  Campbell, Abolition and its Aftermath in Madagascar’; Klein, Emancipation of Slaves’; Timothy Derek Fernyhough, ‘Serfs, Slaves and Shefta: Modes of Production in Southern Ethiopia from the Late Nineteenth Century to 1941’, PhD, University of Illinois at UrbanaChampaign, 1986. 96  Watson, ‘Transactions in People’, p. 245. 97  W. G. Clarence Smith, ‘Islamic Attitudes towards the Abolition of the Slave Trade and Slavery in the Indian Ocean, c1800–c1940’ in Campbell (ed.), Abolition and Its Aftermath. 98  Clarence Smith, ‘Islamic Attitudes’.

Slavery, Forced Labour and Resistance in IOA and Asia

1415

In 1847–48, Ottoman rulers banned the maritime slave traffic in the Gulf, and Islamic support for abolition was also marked in Egypt and East Africa.99 Support for abolition also came from non-Islamic regimes. In 1744, the Korean monarchy abolished enslavement for debt; in 1783 they ended enslavement for a convict’s immediate family; and in 1801 they emancipated all 66,000 state-owned slaves thereby effectively undermining slavery well before it was officially abolished in 1894.100 Again, the Thai monarchy effectively ended slavery by 1900.101 It should also be noted that in some regions, slaves themselves resisted ‘free’ status where the latter resulted in lower standards of working and living. For instance, some slaves in Imperial Madagascar refused manumission because as ‘freedmen’ they became vulnerable to harsh state-imposed forced labour.102 Similarly, in 1943, some royal slaves revolted in the Aden Protectorate against a freedom which threatened to undermine their traditional privileges—one of the few cases where emancipation compensation to slaves was considered.103 Although experiencing vastly different lifestyles, some slaves in both Aden and Imperial Madagascar could and did earn more than most of the nominally ‘free’ population. It has been traditionally assumed that the abolitionist movement in IOA and Asia was led by Western powers. However, in the Cape, the 1826 removal of measures protecting wine diverted investment to the far less labour-intensive wool-producing sector, effectively undermining the need for slaves.104 Elsewhere, the labour-intensive nature of the economy and lack of free wage labour rendered academic any arguments that slave labour might be inefficient105 while European authorities in the region often supported slavery until formal abolition. Thus, Robert Farquhar, the first British governor

99  Ibid.; Behnaz Mirzai, ‘The Shari’a and the Anti-Slave Trade “Farman” of 1848 in Iran’ in Campbell (ed.), Abolition and Its Aftermath; Abdul Sheriff, Slaves, Spices & Ivory in Zanzibar: Integration of an East African Commercial Empire into the World Economy, 1770–1873 (London: James Currey, 1987), pp. 223–35. 100  Kim, ‘Nobi’. 101  Klein, ‘Emancipation of Slaves’. 102  Campbell, ‘Abolition and its Aftermath in Madagascar’. 103  Miers, ‘Slave Rebellion’. 104  See e.g. Worden, ‘Indian Ocean Slavery’. 105  William Gervase Clarence-Smith (ed.), The Economics of the Indian Ocean Slave Trade (London: Frank Cass, 1989), pp. 4–5.

1416

Campbell and Alpers

of Mauritius, delayed anti-slave import measures in order to meet the cheap labour needs of local sugar planters.106 The British and French officially abolished slavery in most of their imperial domains in 1834 and 1848 respectively. However, abolition did not initially extend to British India—under East India Company rule—where in 1841 Sir Bartle Frere estimated slave numbers at eight to nine million107 and not until 1860 was slave holding completely banned in British India.108 In Portuguese India, officials and settlers devised a number of strategies to avoid the abolitionist implications of the 1842 Britanno-Portuguese treaty, and perpetuate both illicit slave imports and slave holding.109 Meanwhile, French authorities on Réunion and neighbouring French-held colonies such as Mayotte condoned the continued import of slaves under the guise of contract labour.110 Abolition constituted a core justification for the imposition of colonial rule during the ‘scramble’ for colonies from 1870 to 1914. However, colonial regimes, constrained by tight budgets and insufficient private investment, sought to maximize local resources, notably through imposing money and labour taxes on the colonized. However, as much manpower was tied up in forms of servitude to the local elite whose cooperation was vital to the colonial administration, the latter generally moved slowly on implementing abolitionist measures lest it provoke revolt.111 Colonial governments also feared that abolition would deprive European colonists of labour. The colonial regime in Somalia initially granted European settlers slave labour,112 and in German East Africa permitted

106  Anderson, ‘The Bel Ombre Rebellion’; see also Samuel Pasfield Oliver, ‘Sir Robert Townsend Farquhar and the Malagasy Slave Trade’, Antananarivo Annual and Madagascar Magazine 15 (1891), pp. 319–21; Anthony J. Barker, Slavery and Anti-Slavery in Mauritius, 1810–33 (Houndsmills and London: Macmillan/New York: St. Martin’s Press, 1996). 107  Timothy Walker, ‘Abolishing the Slave Trade in Portuguese India: Documentary Evidence of Popular and Official Resistance to Crown Policy, 1842–60’, in this volume. 108  Klein, ‘Emancipation of Slaves’; Indrani Chatterjee, ‘Abolition by Denial? Slavery in South Asia after 1843’ in Campbell (ed.), Abolition and Its Aftermath. 109  Walker, ‘Abolishing the Slave Trade’, this volume. 110  Denis, ‘Forced Labour and the Mayotte Revolt’; Gwyn Campbell, ‘The East African Slave Trade, 1861–1895: The “Southern” Complex’, International Journal of Southern African Studies, 21, 4 (1988), pp. 1–27. 111  Clarence Smith, ‘Islamic Attitudes’; Salman, ‘Colonialism, Nationalism and the Meaning of Slavery’. 112  Eno, ‘Abolition and its Impact on the Benadir Coast’; Declich, ‘Unfree Labour’.

Slavery, Forced Labour and Resistance in IOA and Asia

1417

planters to ‘ransom’ slaves who were obliged to work for their ‘liberators’ until the ransom had been paid off.113 Most colonizing powers, including the British, declared newly conquered territories to be ‘protectorates’, thus avoiding the necessity of enforcing full abolition as in the ‘colonies’. Many argued that traditional servitude, because it involved a degree of ‘protection’ for those subject to it, did not constitute true slavery. American colonial officials and Filipino nationalists who opposed slavery ideologically used precisely this argument to avoid confronting slavery in minority regions of the Philippines.114 Moreover, they rarely moved decisively against overland slave trades. Slavery was not abolished in former German East Africa until 1922 and continued in parts of British-held Burma and Sudan until the late 1920s.115 In the Netherlands Indies, slavery was banned in 1860, but the Dutch then possessed only one quarter of the Indonesian territory that was to pass under their control by 1910, in much of which they tolerated slavery. Slavery endured in remote regions of French Indochina and the Dutch Indies into the 1940s.116 In the Middle East, where British influence was paramount following the First World War, the desire to control oil resources led to official tolerance of both the slave trade and slave-holding until the 1930s. In some regions slavery endured until the 1950s, abolition not being formally imposed in Saudi Arabia until 1962, and 1963 in the Gulf Emirates—although it lingered on into the late 1960s in many regions. Only in 1970 did Sultan Qaboos abolish slavery in Oman upon overthrowing his father.117

113  Suzanne Miers and Martin Klein, ‘Introduction’ to Miers and Klein (eds.), Slavery and Colonial Rule in Africa (London: Frank Cass, 1999), p. 6; see also Kopytoff and Miers, ‘African Slavery’ pp. 73–4. 114  Michael Salman, ‘Resisting Slavery in the Philippines: Ambivalent Domestication and the Reversibility of Comparisons’, in this volume; Idem, The Embarrassment of Slavery: Controversies over Bondage and Nationalism in the American Colonial Philippines (Berkeley: University of California Press, 2001). 115  Miers, ‘Slavery and the Slave Trade’; see also Kopytoff and Miers, ‘African Slavery’, p. 72; Miers and Klein, ‘Introduction’, pp. 1–2, 4–5. 116  Klein, ‘Emancipation of Slaves’; Boomgaard, ‘Human Capital’; Delaye, ‘Slavery and Colonial Representations in Indochina’ see also Reid, ‘Introduction’, p. 34; Klein, ‘Introduction’, p. 24. 117  Miers, ‘Slave Rebellion’; Idem, ‘Slavery and the Slave Trade’; Asl, ‘The Shari’a’; Klein, ‘Emancipation of Slaves’; Clarence Smith, ‘Islamic Attitudes’.

1418

Campbell and Alpers

Conclusion The contributions in this volume address a wide range of resistance against slavery and forced labour in Indian Ocean Africa (IOA) and Asia, as well as the responses of different societies to those revolts. Taken together, they bear out the implicit argument we have been making in this introduction that slavery and the responses of the enslaved in IOA and Asian societies need to be understood on their own terms and not through abstractions that have long had a dominant influence on the study of slavery in the Americas. They further bear out our contention that conventional distinctions between ‘slave’ and ‘free’ based upon Western concepts collapse under the historical conditions of bondage, unfree labour and resistance in IOA and Asia. Above all, these contributions demonstrate the ways in which people in bondage, whether as slaves or as forced labourers, sought to act on their own behalf and in their own interests despite their subjugation. Surely in this aspect they shared a common history with those who were enslaved in other parts of the world.

Carrying Away the Unfortunate from India and Southeast Asia, 1500–1800 Richard B. Allen Early in March 1793, Lieutenant Thomas Riddell, the commander of an East India Company detachment of sepoys stationed at Coringa on India’s Coromandel Coast, received a letter from Pierre Sonnerat, commandant of the French factory at nearby Yanam. In his letter, Sonnerat informed Riddell that although he had done everything in his power to prevent slave trading at Yanam, “I was not able to prevent many of the people at the time from searching, and taking advantage of the misery which reigned to the Northward, to carry away the unfortunate from their Country and from their families.” While he asserted that his orders to punish slave traders severely would prevent future exports from Yanam, Sonnerat added that unless the British adopted the same measures elsewhere in India that they had implemented at Coringa to suppress this trade, “there will always be found some private people without delicacy, greedy and allured by the prospect of gain, who will go into the Country to carry on that disgraceful traffick.”1 Company officials dismissed Sonnerat’s claims about suppressing slave trading at Yanam because of the reports they received from Matthew Yeats, their agent at Ingeram, also near Yanam. Early in February 1792, Yeats informed his superiors in Madras that there were hardly any Europeans at the French factory “from the Chief downwards” who were not involved in slave trading, and that almost every French vessel sailing from Coringa or the Dutch factory at Jaggernatpooram [sic] had slaves on board.2 The following month, Yeats advised Madras that the French at Yanam “now send off their slaves in small parties by night, to avoid detection, which, when they have passed the English settlements in this area, unite and proceed by land to Pondicherry” and that while he had intercepted three such parties containing a total of 19 slaves, Source: Allen, Richard B., “Carrying Away the Unfortunate from India and Southeast Asia, 1500– 1800,” in Richard B. Allen, European Slave Trading in the Indian Ocean, 1500-1850, Athens: Ohio University Press, 2014, 108–140. © 2014, Ohio University Press. This material is used by permission of Ohio University Press, www.ohioswallow.com. 1  I OR: P/241/38, pp. 1229–30, Sonnerat to Thomas Riddell, 7 March 1793. 2  I OR: P/241/31, p. 562, Matthew Yeats to Maj.-Gen. William Medows, 3 February 1792.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_046

1420

Allen

other groups shepherding a total of about 30 slaves had eluded capture.3 Later that same year, he reported the discovery of 6 women and a child confined in a warehouse less than twenty yards from Sonnerat’s house, a discovery that was followed by the hasty departure from Coringa of two French ships carrying as many as 900 slaves.4 Shortly thereafter, the governor at Madras advised the governor-general in Calcutta that this traffic was tolerated, if not actually carried on, by “persons of authority” at Yanam and that the inability of the French governor at Pondichéry to redress this problem meant that the British would probably have to deal with it themselves.5 While Riddell’s instructions to lead a detachment of troops to Coringa diplomatically downplayed official French and Dutch involvement in slave trading near Coringa, they noted that “many” Indian slaves had been exported from Yanam and Jaggernatpooram and directed him accordingly “to employ such means as circumstances may require” to suppress this traffic.6 Madras also dispatched troops to Bimilipatnam for the same purpose following the discovery of 565 “young persons” being held for export there.7 This was not the first time that the exportation of slaves taken from British territories by other European nationals had come to the attention of company officials in India. Eight years earlier, Matthew Day, Collector at Dacca (Dhaka) in Bengal, had reported that “many hundreds” of enslaved children from his district had been “landed in the Foreign Settlements [near Calcutta] from whence … they are embarked in Vessels to different Parts.”8 The magnitude of this traffic, Day continued, was illustrated by the Hon. Mr. Lindsay’s report that he had encountered more than a hundred boats loaded with such children while traveling between Dacca and Calcutta. Nor was such trafficking confined to Bengal and the Coromandel Coast. In May 1792, W. G. Farmer reported from Calicut on the “very extensive” trade carried on by the French at Mahé on the Malabar Coast to supply Mauritius and Réunion with slaves.9 Several months later, Farmer informed his superiors that although the French government in 3  P P 1828 XXIV [125], 491, Matthew Yeats to Maj.-Gen. William Medows, 12 March 1792. 4  I OR: P/241/36, pp. 16, 19, Matthew Yeats to Sir Charles Oakeley, 22 December 1792. 5  I OR: P/241/36, pp. 249–50, Charles Oakeley to Earl Cornwallis, 18 January 1793. 6  I OR: P/241/37, pp. 503–5, Anthony Sadleir to Lt. Thomas Riddell, 14 January 1793. 7  P P 1828 XXIV [125], 477–78, Governor-in-Council at Fort St. George to Court of Directors, 2 May 1793. 8  I OR: P/50/60, M. Day to William Cowper, 2 March 1785, following L.R. No. 311, Wm Cowper to the Hon. John Macpherson, 14 March 1785, in Fort William proceedings of 9 September 1795. 9  I OR: P/E/5, Report of W. G. Farmer, Calicut, 17 May 1792.

Carrying Away The Unfortunate From India And Southeast Asia

1421

India had banned slave trading, planters in the Mascarenes “would hardly also forbear [termination of] this Trade which, by a late examination taken at the Adawlut Cutcherry [court house] and referred to me by Mr Handley they appear to have carried on in a manner equally prejudicial to this Country and repugnant to all humanity.”10

Historiographical Issues

Slavery and slave trading were integral components of socioeconomic life in many of the lands washed by the Indian Ocean long before Europeans arrived on the scene in 1498. In spite of a growing appreciation by historians that hundreds of thousands of enslaved men, women, and children were exported from sub-Saharan Africa to Arabia, the Persian Gulf, and South Asia in the centuries before 1500,11 histories of the Indian Ocean either ignore this traffic in chattel labor or pay only passing attention to it.12 Compounding this historiographical silence is the Africa-centric focus of much of the recent scholarship on slavery and slave trading in the region.13 The attendant emphasis on the northwestern Indian Ocean14 poses a potentially significant obstacle to a more comprehensive understanding of the slave and other migrant labor systems within and ultimately beyond this oceanic world’s confines. The exportation of chattel laborers from India and Southeast Asia into the wider Indian Ocean, by comparison, remains largely unexplored.15 Major studies of South Asian trade and commerce,16 slavery and bondage in India17 and 10  IOR: P/E/6, p. 448, [Letter from] William Gamul Farmer, Calicut, 16 August 1793. 11  Paul E. Lovejoy, Transformations in Slavery: A History of Slavery in Africa, 3rd ed. (Cambridge: Cambridge University Press, 2012), 25–27, 46, 61. 12  See chap. 1, n. 16. 13  Markus Vink, “‘The World’s Oldest Trade’: Dutch Slavery and Slave Trade in the Indian Ocean in the Seventeenth Century,” Journal of World History 14, no. 2 (2003): 133. 14  William Gervase Clarence-Smith, ed., The Economics of the Indian Ocean Slave Trade During the Nineteenth Century (London: Frank Cass, 1989); Gwyn Campbell, ed., The Structure of Slavery in Indian Ocean Africa and Asia (London: Frank Cass, 2004); Gwyn Campbell, ed., Abolition and Its Aftermath in Indian Ocean Africa and Asia (London: Routledge, 2005). 15  For a recent overview of the Indian slave trade in the Indian Ocean, see Marina Carter, “Slavery and Unfree Labour in the Indian Ocean,” History Compass 4, no. 5 (2006): 800–813. 16  See chap. 1, n. 24. 17  See chap. 1, n. 23.

1422

Allen

Southeast Asia,18 and European commercial activity in this oceanic world19 make little or no reference to the exportation of slaves. Although the presence of significant numbers of slaves of Indian and Southeast Asian origin in the Mascarenes20 and South Africa21 is well documented and such individuals are 18  Bruno Lasker, Human Bondage in Southeast Asia (Chapel Hill: University of North Carolina Press, 1950); Anthony Reid, ed., Slavery, Bondage and Dependency in Southeast Asia (New York: St. Martin’s Press, 1983). 19  C. R. Boxer, The Dutch Seaborne Empire, 1600–1800 (New York: Alfred A. Knopf, 1965); S. P. Sen, The French in India, 1763–1816, 2nd ed. (New Delhi: Munshiram Manoharlal, 1971); Philippe Haudrère, La Compagnie française des Indes aux XVIIIe siècle (1719–1795) (Paris: Librairie de l’Inde, 1989); Sanjay Subrahmanyam, Improvising Empire: Portuguese Trade and Settlement in the Bay of Bengal, 1500–1700 (Delhi: Oxford University Press, 1990); George D. Winius and Marcus P. M. Vink, The Merchant-Warrior Pacified: The VOC (The Dutch East India Company) and Its Changing Political Economy in India (Delhi: Oxford University Press, 1994); Om Prakash, European Commercial Enterprise in Pre-Colonial India (Cambridge: Cambridge University Press, 1998); Sudipta Sen, Empire of Free Trade: The East India Company and the Making of the Colonial Marketplace (Philadelphia: University of Pennsylvania Press, 1998); Arvind Sinha, The Politics of Trade: Anglo-French Commerce on the Coromandel Coast, 1763–1793 (New Delhi: Manohar, 2002); Nick Robins, The Corporation That Changed the World: How the East India Company Shaped the Modern Multinational (London: Pluto Press, 2006); Philip J. Stern, The Company-State: Corporate Sovereignty and the Early Modern Foundations of the British Empire in India (Cambridge: Cambridge University Press, 2011). 20  Charu Chandra Ray, “A Note on Slaves and Slavery in Old Chandernagore,” Bengal, Past and Present 6 (1910): 257–65; K. K. N. Kurup, “Slavery in 18th Century Malabar,” Revue historique de Pondichéry 11 (1973): 56–60; Hubert Gerbeau, “Des minorités mal-connues: Esclaves indiens et malais des Mascareignes au XIXe siècle,” in Migrations, minorités et échanges en océan Indien, XIXe–XXe siècle, Études et Documents No. 11 (Aix-en-Provence: Institut d’Histoire des Pays d’Outre-Mer, 1978), 160–242; Hubert Gerbeau, “Les esclaves asiatiques des Mascareignes au XIXe siècle,” Annuaire des pays de l’océan Indien 7 (1980): 169–97; Teotonio R. de Souza, “French Slave Trading in Portuguese Goa (1771–1793),” in Essays in Goan History, ed. Teotonio R. de Souza (New Delhi: Concept Publishing Company, 1989), 119–31; Richard B. Allen, “A Traffic of Several Nations: The Mauritian Slave Trade, 1721– 1835,” in History, Memory and Identity, ed. Vijayalakshmi Teelock and Edward A. Alpers (Moka, Mauritius: Nelson Mandela Centre for African Culture/University of Mauritius, 2001), 157–77; Richard B. Allen, “The Mascarene Slave-Trade and Labour Migration in the Indian Ocean during the Eighteenth and Nineteenth Centuries,” in Campbell, The Structure of Slavery, 33–50. 21  Robert C.-H. Shell, Children of Bondage: A Social History of the Slave Society at the Cape of Good Hope, 1652–1838 (Middletown, CT: Wesleyan University Press, 1994), 41; Nigel Worden, Slavery in Dutch South Africa (Cambridge: Cambridge University Press, 1995), 46–48.

Carrying Away The Unfortunate From India And Southeast Asia

1423

known to have reached Portugal during the sixteenth century22 and British and French colonies in the Americas23 during the eighteenth century, we still know little about the dynamics of these trades.24 The presence of such slaves throughout the Indian Ocean basin and in the Atlantic world, coupled with Markus Vink’s pioneering research on Dutch slave trading from India25 and Kerry Ward’s work on forced migration in the VOC’s commercial empire,26 highlights the need to examine this European commerce in Asian slaves in greater detail and situate these trades in relevant regional and global contexts. Doing so requires us to ask many of the same questions and cope with many of the same limitations that govern research on the East African and Malagasy trades.

Trading Indians

Europeans first exported slaves from India no later than 1510, when 24 slaves were shipped to Portugal from Cochin (Kochi) on the Malabar Coast. The volume of Portuguese slave trading in India is difficult to determine with any 22  A. C. de C. M. Saunders, A Social History of Black Slaves and Freedmen in Portugal, 1441–1555 (Cambridge: Cambridge University Press, 1982), 72. 23  On Indian slaves in British North America, see: Francis C. Assisi, “Indian Slaves in Colonial America,” India Currents, 16 May 2007 [http://www.indiacurrents.com/ articles/2007/05/16/indian-slaves-in-colonialAmerica]; Thomas Costa, The Geography of Slavery in Virginia [http://www2.vcdh.virginia.edu/gos.news/browse/browse_main.php], specifically the advertisements in the Virginia Gazette, 27 April 1737, 4 August 1768, and 13 July 1776. The French ship La Cibèle delivered 386 (or 406) slaves “from the coasts of India” to Saint Domingue in 1778 (Jean Mettas, Répertoire des expeditions négrières françaises au XVIIIe siècle, vol. 1: Nantes, ed. Serge Daget [Paris: Société Française d’Histoire d’Outre-Mer et Librairie Orientaliste Paul Geuthner S. A., 1978], 613–14). 24  J.-M. Filliot, La traite des esclaves vers les Mascareignes au XVIIIe siècle (Paris: ORSTROM, 1974), esp. 175–81; Marina Carter, “Indian Slaves in Mauritius (1729–1834),” Indian Historical Review 15, nos. 1–2 (1988): 233–47, esp. 2.33–38; Nigel Worden, “Indian Ocean Slavery and Its Demise in the Cape Colony,” in Campbell, Abolition and Its Aftermath, 29–49, esp. 30– 37. See also Sudeshna Chakravarti, “The Dutch East India Company and Slave Trade in Seventeenth Century India: An Outline by Pieter van Dam, an Advocate of the Company,” Journal of the Asiatic Society 39, no. 2 (1997): 73–99. 25  Markus Vink, “A Work of Compassion? Dutch Slavery and Slave Trade in the Indian Ocean in the Seventeenth Century,” The History Cooperative (2003) [www.historycooperative .org/proceedings/seascapes/vink.html]; Vink, “ ‘The World’s Oldest Trade.’ ” 26  Kerry Ward, Networks of Empire: Forced Migration in the Dutch East India Company (Cambridge: Cambridge University Press, 2009).

1424

Allen

certainty, but the second half of the sixteenth century reportedly witnessed the exportation of “large” numbers of slaves from the subcontinent, with exports totaling perhaps 5,000 to 6,000 in some years during this period.27 Some of the vessels that engaged in this traffic clearly carried large cargoes, as the arrival at Aceh of a small Portuguese vessel from Nagapatnam (Nagapattinam) with some 400 slaves on board in early October 1646 attests.28 Portuguese ships also transported several hundred slaves from southern India, Burma, China, Malaya, Java, and other parts of Asia to the Philippines each year during the union of the Portuguese and Spanish crowns (1580–1640), with Malacca frequently serving as the collection point from which these laborers were funneled to Manila. Spanish colonists in the Philippines preferred Asian over African slaves, while Portuguese traders also sought to satisfy the local demand for skilled Asian workers.29 Some of the slaves shipped to Manila from South and Southeast Asia subsequently reached Mexico. Estimates of the number of Asian slaves and other immigrants, including sailors and servants, who did so vary, but as many as 6,000 such slaves may have landed in New Spain each decade during the early seventeenth century.30 Recent scholarship suggests that at least 40,000 to 60,000 and more probably some 100,000 Asians reached Mexico between the late sixteenth and late seventeenth centuries, with slaves and servants comprising the second largest component (after sailors) of this immigrant population.31 Dutch interest in Indian slaves followed closely on the heels of the VOC’s establishing itself in the East Indies early in the seventeenth century and was driven by the demand for labor in Batavia and the company’s other possessions in the Indonesian archipelago.32 In 1618, the VOC factory at Masulipatam 27  K. S. Mathew, Portuguese Trade with India in the Sixteenth Century (New Delhi: Manohar, 1983), 137–38. 28  William Foster, The English Factories in India, 1646–1650: A Calendar of Documents in the Indian Office, Westminster (Oxford: Clarendon Press, 1914), 54, n. 2. 29  Tatiana Seijas, “The Portuguese Slave Trade to Spanish Manila: 1580–1640,” Itinerario 22, no. 1 (2008): 19–38. See also: George Bryan Souza, The Survival of Empire: Portuguese Trade and Society in China and the South China Sea, 1630–1754 (Cambridge: Cambridge University Press, 1986), 74–75, 78; William Henry Scott, Slavery in the Spanish Philippines (Manila: De La Salle University Press, 1991), 27–28, 31. 30  Jonathan Israel, Race, Class, and Politics in Colonial Mexico, 1610–1670 (Oxford: Oxford University Press, 1975), 75–76. 31  Edward R. Slack, Jr., “The Chinos in New Spain: A Corrective Lens for a Distorted Image,” Journal of World History 20, no. 1 (2009): 37, 41–42. 32  Om Prakash, The Dutch Factories in India, 1617–1623: A Collection of Dutch East India Company Documents Pertaining to India (New Delhi: Munshiram Manoharlal, 1984), 11.

Carrying Away The Unfortunate From India And Southeast Asia

1425

on the Coromandel Coast received orders to send as many slaves as possible to Bantam; the following year, Masulipatam reported that company personnel at Pulicat had been directed to purchase both male and female slaves, whom they hoped to forward to Java forthwith.33 The extent of this demand is revealed in a letter in which Jan Pieterszoon Coen, Batavia’s founder and principal architect of the VOC’s eastern empire, informed Masulipatam on 18 October 1621 that Batavia and the company’s spice plantations on Amboina and Banda needed 2,000 to 3,000 slaves.34 The following year, East India Company officials at Pulicat asserted that VOC personnel on the Coromandel Coast had been ordered to acquire as many as 4,000 to 5,000 men, women, and children.35 This demand spurred the exportation of 150 to 400 slaves a year from the ArakaneseBengali coast to Indonesia between 1626 and 1662, as well as the shipment of “large” numbers of slaves from the Malabar Coast before 1663, after which exports from Malabar declined to 130 to 220 a year. Dutch slave trading along the Coromandel Coast tended to be more sporadic. Markus Vink identifies six periods of relatively intense trafficking along this coast: 1622–23 (1,900 exports from ports such as Pulicat and Devanampatnam); 1645–46 (2,118 exports from the southern Coromandel Coast); 1659–61 (8,000–10,000 exports from Nagapatnam, Pulicat, and other ports, mostly to Ceylon); 1673–77 (a minimum of 1,839 exports from Madurai and elsewhere along the southern Coromandel Coast); 1688 (an unknown number of exports); and 1694–96 (3,859 imports into Ceylon).36 The scale of this traffic is also indicated by recent assertions that some 15,000 Indian slaves reached Batavia between 1618 and 1680.37 Bengali and “Malabar”38 slaves also reached Dutch Mauritius39 and South Africa during the seventeenth century, although the number who did so is uncertain. The period from 1695 to 1707 witnessed the importation of 1,844 slaves 33  Ibid., 75, 119. 34  Ibid., 179. 35  William Foster, The English Factories in India, 1622–1623: A Calendar of Documents in the India Office and British Museum (Oxford: Clarendon Press, 1908), 105. 36  Vink, “ ‘The World’s Oldest Trade,’ ” 143. On slavery in Ceylon and the Dutch need to import slaves into the island, see Nandadeva Wijesekera, “Slavery in Sri Lanka,” Journal of the Sri Lanka Branch of the Royal Asiatic Society 18 (1974): 1–22. 37  Daniel Perret, “From Slave to King: The Role of South Asians in Maritime Southeast Asia (From the Late 13th to the Late 17th Century),” Archipel 82 (2011): 165, 192. 38  In addition to referring to those originating from Malabar province or the Malabar Coast, this descriptor was commonly used to describe any slave of southern Indian origin. 39  Gabrielle de Nettancourt, “Le peuplement néerlandais à l’Ile Maurice (1598–1710),” in Mouvements de populations dans l’océan Indien (Paris: Librairie Honoré Champion, 1979), 224.

1426

Figure 1

Allen

Arakanese selling slaves to the Dutch at Pipely/Baliapal, India, in January 1663. From Wouter Schouten, Reys-togten naar en door Oost-Indien: in welke, de voornaamste landen, koningryken, steden, eylanden, Bergen, en rivieren, met haare eigenschmeten, beneffens de wetten, godsdienst, zeden en dragten der inwoonders, en watverder zoo van dieren, vrugten, en planten aanmerkelyks in die gewesten is; naauwkeurig word beschreven [Travels in the East Indies …], 2nd ed. (Amsterdam: Andries van Damme, 1708). (Author’s collection)

Carrying Away The Unfortunate From India And Southeast Asia

1427

into the Cape Colony, with approximately 1,400 such imports between 1718 and 1729; imports averaged about 100 to 200 each year thereafter.40 The number of Indian slaves among these imports is difficult to ascertain, but colonial inventories and auction sales that record the presence of 633 Indian slaves in the colony between 1697 and 1750 and 54 in the years between 1790 and 1793 confirm the presence of Indian slaves throughout the eighteenth century.41 Overall, Indians probably comprised slightly more than one-fourth of the 63,000 slaves estimated to have been imported into the Cape between 1652 and 1808.42 These various data indicate that the VOC exported at least 26,000 to 38,000 Indian slaves to Batavia, Ceylon, Malacca, and elsewhere in the region during the seventeenth century and underpin other estimates that Dutch exports from India during this century may have totaled 100,000, if not more.43 Indigenous Hindu and Muslim traders also capitalized on the demand for chattel labor in the East Indies. In 1622, for example, two ships that had sailed from the Orissa coast landed 300 slaves at Aceh,44 while Asian traders shipped thousands of slaves from Nagapatnam to Southeast Asia during the export boomlet of 1688.45 The total number of Indian slaves exported to Southeast Asia by Indian and other Asian traders cannot be determined with any certainty, but the possible scale of this traffic is suggested by a conservative estimate that Southeast Asian states (exclusive of Portuguese and Dutch settlements) imported an average of 2,000 South Asian slaves, captives, and fugitives each year between the fifteenth and seventeenth centuries.46 The seventeenth century also witnessed Indian slaves being traded by the British and French. Information about British involvement in this traffic is fragmentary.47 The earliest report of such activity dates to 1622, when East India Company officials at Batavia asked their counterparts at Pulicat to 40  James C. Armstrong and Nigel A. Worden, “The Slaves, 1652–1834,” in The Shaping of South African Society, 1652–1840, ed. Richard Elphick and Hermann Giliomee, 2nd ed. (Middletown, CT: Wesleyan University Press, 1988), 119–20. 41  Worden, “Indian Ocean Slavery,” 31. More than three-fourths of these slaves came from Bengal and the Malabar Coast. 42  Shell, Children of Bondage, 40–41. 43  Rik Van Welie, “Slave Trading and Slavery in the Dutch Colonial Empire: A Global Comparison,” New West Indian Guide/Nieuwe West-Indische Gids 82, nos. 1–2 (2008): 72. 44  Prakash, The Dutch Factories, 197. 45  Vink, “ ‘The World’s Oldest Trade,’ ” 140–41. 46  Perret, “From Slave to King,” 164. 47  B. Natarajan, “Slave Trade in Madras,” in The Madras Tercentenary Commemoration Volume (Madras: Oxford University Press, 1939), 247–54.

1428

Allen

send them 10 to 15 male slaves between sixteen and twenty-two years of age. Although Pulicat could not honor this request immediately because the Dutch had reportedly already purchased all available slaves in the area, the factory shipped 22 slaves to Batavia later that year.48 Two years later, John Bickley, captain of the Hart, received orders to purchase 30 slaves at Karikal for a factory to be established at Sericall [sic].49 In 1654, Venkata, a Brahmin who served as a company merchant at Madras, was fined 16 pagodas for accepting bribes to facilitate the sale of 4 kidnapped children to Dutch brokers at Pulicat.50 On 11 July 1662, the company’s directors directed Madras to purchase 10 “Gentues” to be sent to St. Helena; the following year, officials at Fort St. George received additional instructions to purchase a couple of Gentue barbers, also for St. Helena.51 Orders to procure Gentue or “Arracan” (Arakanese) slaves for St. Helena followed in 1668 as well.52 A 1683 report noted the exportation of a “great” number of slaves from Madras each year.53 September 1687 witnessed the exportation of 665 slaves from Madras,54 some of whom may have been sent to St. Helena.55 This traffic was large enough to attract the attention and incur the displeasure of local Mughal officials, as a result of which Governor Elihu Yale and his council banned the purchase and exportation of slaves from Madras and neighboring ports in May 1688 on pain of a 50–pagoda fine for each slave illegally purchased and exported.56 The late seventeenth century 48  Foster, The English Factories in India, 1622–1623, 85, 105. 49  William Foster, The English Factories in India, 1624–1629: A Calendar of Documents in the India Office, Etc. (Oxford: Clarendon Press, 1909), 10. 50  Henry Davison Love, Vestiges of Old Madras, 1640–1800: Traced from the East India Company’s Records Preserved at Fort St. George and the India Office, and from Other Sources (London: J. Murray, 1913), vol. 1, 127–35; William Foster, The English Factories in India, 1651– 1654: A Calendar of Documents in the India Office, Westminster (Oxford: Clarendon Press, 1915), 235, 238, 244, 285. Twenty-seven or more children may actually have been involved. 51  William Foster, The English Factories in India, 1661–1664 (Oxford: Clarendon Press, 1923), 271, 276. 52  Ethel Bruce Sainsbury, A Calendar of the Court Minutes, Etc. of the East India Company, 1668–1670 (Oxford: Clarendon Press, 1929), 122. 53  Basanta Kumar Basu, “Notes on Slave Trade and Slavery in India during the Early Days of John Company,” Muslim Review [Calcutta] 4, no. 4 (1930): 22. 54  Love, Vestiges of Old Madras, vol. 1, 545. 55  In September 1686, officials at Madras ordered 30 slaves to be sent to St. Helena (IOR: G/19/4, fol. 62v, Fort St. George Pub. Cons., 30 September 1686). The following January, every ship leaving Madras for Britain was ordered to carry 10 slaves to the island (IOR: G/19/4, fol. 94v, Fort St. George Pub. Cons., 17 January 1686/7). 56  IOR: G/19/5, fol. 43r, Fort St. George Pub. Cons., 14 May 1688; BL: Eur. Mss. Mack General 55, pp. 46–47, Diary and Consultation Book of the Honble Elihu Yale Esqre President

Carrying Away The Unfortunate From India And Southeast Asia

1429

also found the French Compagnie des Indes transporting Indians to the Ile de Bourbon. Of the island’s 58 “black” laborers in 1674, 15 came from South Asia,57 and the island housed at least 13 Indian females in 1678 and 12 Indian males in 1686.58 While the legal status of these individuals remains uncertain, the sale of François, a twelve-year-old Indian boy, on the island in 1687 confirms the de facto presence of South Asian slaves in the Mascarenes by the late seventeenth century.59 With the exception of the years between 1785 and 1793, information about European slave trading in India during the eighteenth century is equally fragmentary. We currently know little about Dutch or Portuguese activity during this period. British traders may have been largely responsible for carrying small groups of Bengali slaves to the Cape Colony during the first half of the eighteenth century.60 Much of the evidence of the East India Company’s trafficking in Indian slaves during this era comes from Bencoolen, the company’s factory on Sumatra’s west coast. A 1712 slave census recorded Malabars and Bengalis among the factory’s slaves.61 Thirty-five years later, another slave census reported that the settlement housed 12 or 13 adult Malabars (10 or ii men, 2 women).62 Small numbers of Indian slaves continued to arrive at Fort Marlborough during succeeding decades, and in 1766 25 (22 men, 3 women) of the 853 slaves at the settlement and its outstations were described as Malabars.63 The company also shipped enslaved Indians to other establishments, as in 1758, when 10 “Malabar” officers formerly in the ruler of Travancore’s service were sent to

Governour & Councill their proceedings & transactions for the affairs of the Rt Honble English East India Company in the Presidency of the Coast of Choromandell & the Bay of Bengala. Begun the 18th of February 1687/8. 57  Filliot, La traite des esclaves, 31. 58  Hubert Gerbeau, “The Indians of the Mascarenes—A Success in Diaspora: Mauritius and Réunion (17th–20th Centuries),” Bengal, Past and Present 116, nos. 1–2 (1997): 35–79. 59  Hai Quang Ho, “L’esclavage à l’Ile Bourbon de 1664 à 1714,” in L’esclavage à Madagascar: Aspects historiques et resurgences contemporaines, ed. Ignace Rakoto (Antananarivo, Madagascar: Institut de Civilisations—Musée d’Art et Archéologie, 1997), 29–51. 60  Worden, Slavery in Dutch South Africa, 48. 61  IOR: G/35/8, fol. 62, York Fort Bencoolen Mo Septembr 1712 An Acct of the Rt Honbe Company Slaves where they are & how Employ[ed]. 62  IOR: G/35/9, fols. 177r–180r [No. 1175], List of the Honble Company’s Slaves at their Settlement at Fort Marlbro the 20th Jany 1747/8; fols. 221r–224r [No. 1155], A List of Slaves belonging [to] the Honble Companys Settlement of Fort Marlborough [17 January 1748]. 63  IOR: G/35/14, fols. 87r–98r [No. 177], List of the Hon’ble Companys Slaves at Fort Marlborough & its Subordinates [31 December 1766].

1430

Allen

St. Helena as slaves.64 The last known and ultimately problematic instance of company involvement in the exportation of Indian slaves dates to 1789, when Peter Horrebow, the Danish captain of the country ship Friendship, reportedly received permission to sail as a slaver under English colors and with an East India Company pass only to be subsequently tried and convicted of illegally trading slaves.65 Three years later, Pierre Sonnerat asserted that three English vessels were about to sail for Mauritius with slave cargoes,66 allegations that company officials promptly dismissed as false.67 In addition to exporting slaves, Britons and other Europeans living in India purchased Indian slaves for use locally, usually as domestic servants. There is unquantifiable evidence of domestic slavery in European households at Madras until at least the end of the eighteenth century, with most of these slaves being of South Indian origin.68 Some Europeans appear to have owned sizable numbers of such slaves. When the Calcutta police raided the house of Mr. Borel, a Swiss officer, in September 1789, for example, they found eight children on the premises, six of whom Borel claimed to have purchased to be his servants.69 As formal requests by company employees in the early 1770s attest, individuals returning to England also took slave boys and girls, some of whom were identified explicitly as Indians, with them when they left the subcontinent.70 Although slaves from the “coasts of India” reached Saint Domingue on at least one occasion,71 French slave trading in the subcontinent was driven largely by the demand for chattel labor in the Mascarenes. This demand was fueled in part by perceptions that Indians were cleaner and more docile, faithful, 64  IOR: E/4/996, p. 597, [Despatch] to President and Council at Bombay, 12 May 1758. Five of these men hanged themselves shortly after arriving on the island. 65  PP 1828 XXIV [125], 19, Extract from the Calcutta Gazette, Thursday, 30 July 1789. See chap. 6 for more on Horrebow’s prosecution. 66  IOR: P/241/37, Sonnerat to Mr. Yeats, 14 December 1792. 67  IOR: P/241/37, Anthony Sadlier, W. A. Dobbyn, and Andrew Scott to Sir Charles Oakeley, 18 February 1793. 68  Ravi Ahuja, “Labour Relations in an Early Colonial Context: Madras, c. 1750–1800,” Modern Asian Studies 36, no. 4 (2002): 804. Ahuja also notes that indigenous Indian slaves, the majority of whom worked in private households, outnumbered the town’s Coffree slaves. 69  IOR: P/3/47, pp. 859–60, T. Motts and Edwd Maxwell to [Earl Cornwallis], 14 September 1789. Borel claimed that the other two children belonged to a Mr. Milliat at Chandernagore. 70  IOR: P/1/47, fol. 283r, J. Richard to John Cartier, 10 December 1770; P/1/49, p. 630, George Hadley to John Cartier, 1 November 1771; P/1/49, p. 631, John Deffell to John Cartier, 1 November 1771; P/2/3, p. 79, Mills [to John Stewart], 25 January 1773. 71  See n. 23 above.

Carrying Away The Unfortunate From India And Southeast Asia

1431

and intelligent than African and Malagasy slaves, who were usually deemed fit only for manual labor.72 During the first decades of the eighteenth century, 10 to 20 Indian slaves probably reached the Mascarenes each year, after which the volume of such imports increased. The Île de Bourbon housed 45 Indian slaves in 1704; four years later, the 85 Indians (71 men, 14 women) on the island comprised 31.7 percent of its slave population. These numbers climbed to 93 in 1709 and then to 487 in 1735. The island held 1,100 Indian slaves by 1765, a number that increased to 1,955 in 1808 before declining to about 1,800 in 1826.73 Mauritius, however, received a substantial majority of the Indian slaves landed in the Mascarenes. Two hundred sixty-nine enslaved Indians reached the island between 1728 and 1735, as did some of the 200 or more Indians imported into the islands during 1737–38.74 The Île de France housed approximately 1,000 Indian bondmen and -women by the late 1760s, a number that soared to 6,162, or 10.2 percent of the island’s total slave population, in 1806.75 The outbreak of war between Britain and France in 1792 and British occupation of French and Dutch possessions in India in 1793 and 1795, respectively, effectively ended the large-scale European exportation of slaves from the subcontinent. These developments, coupled with high slave mortality rates in the Mascarenes during the early nineteenth century, reduced the number of Indian slaves on Mauritius to 2,351 in 1827.76 Despite this traffic’s de facto abolition in 1793, various reports suggest that small numbers of Indian slaves probably continued to reach the Mascarenes 72  IOR: Add. Mss. 33765, fol. 10r and v, Account of Mauritius by C. F. Noble, 1756; J. H. Bernardin de St. Pierre, Voyage à l’île de France (Paris: Armand-Aubrée, 1834), vol. 2, 120–21 [Lettre XII, 25 avril 1769]; M. J. Milbert, Voyage pittoresque à l’Ile de France, au Cap de Bonne-Espérance et à l’Ile de Ténériffe (Paris: A. Nepveu 1812), vol. 1, 217–27, and vol. 2, 162–74; Baron d’Unienville, Statistique de l’Île Maurice et ses dépendances suivie d’une notice historique sur cette colonie et d’un essai sur l’Île de Madagascar, 2nd ed. ([Île] Maurice: Typographie The Merchants and Planters Gazette, 1885–86), vol. 1, 256. 73  Filliot, La traite des esclaves, 177–78; André Scherer, Histoire de La Réunion (Paris: Presses Universitaires de France, 1974), 26–27; J. V. Payet, Histoire de l’esclavage à l’Ile Bourbon (Paris: L’Harmattan, 1990), 17; Sudel Fuma, L’esclavagisme à La Réunion, 1794–1848 (Paris: L’Harmattan, 1992), 37. 74  Philip Baker, “On the Origins of the First Mauritians and of the Creole Language of Their Descendants: A Refutation of Chaudenson’s ‘Bourbonnais’ Theory,” in Isle de France Creole: Affinities and Origins, ed. Philip Baker and Chris Corne (Ann Arbor, MI: Karoma Publishers, 1982), 193; Filliot, La traite des esclaves, 179. 75  Evenor Hitié, Histoire de Maurice (Ancienne Ile de France) (Port Louis, Mauritius: Imprimerie Englebrecht, 1897), 9; Milbert, Voyage pittoresque, vol. 2, 233 bis. 76  MNA: ID 14, Return of the slave population of Mauritius [1827]. On Mauritian slave mortality, see Allen, “The Mascarene Slave-Trade,” 37–38.

1432

Allen

into the early nineteenth century. In April 1794, the superintendent of police at British-occupied Mahé reported that the Portuguese-flagged Phoenix St. José, then anchored in the harbor ostensibly taking on a cargo only of rice, was actually a French-owned ship that he believed would be returning to the Île de France, from whence it had come.77 In February 1812, the company’s resident at Travancore reported the discovery of 24 persons being held in irons at Quilon (Kollam) by a Frenchman named Valley “for the purpose of being transported as slaves to the French islands.”78 The following month, reports about freeborn persons in Travancore being kidnapped and transported to Malabar by sea and kidnapped children being carried to Mahé79 prompted company officials to consider how best to end “the nefarious traffic in Slaves which has prevailed on the Malabar Coast.”80 Hubert Gerbeau reports that at least one partial cargo of Indian slaves reached Réunion after France abolished its slave trade in 1818.81 That ports such as Quilon and Mahé probably exported small numbers of Indian slaves to the Mascarenes during the 1810s and 1820s is also suggested by general observations by company officials in 1825 that “children are not unfrequently [sic] transported by sea from provinces subject to this government [i.e., Madras], to the eastward, or elsewhere, for the purposes of being sold as slaves.”82 A review of census and other data suggests that 5,000 to 5,700 Indian slaves were exported to the Mascarenes between 1670 and 1769, mostly after 1721.83 The 1769 decree that opened the islands to free trade increased the importation of Indian as well as Malagasy and Mozambican slaves, so much so that in May 1774 the “savage commerce” being conducted by Dutch and especially French traders in Bengal prompted Governor Warren Hastings and his council to attempt to regulate slave trading in the province.84 The scale of this traffic 77  IOR: P/E/7, p. 343, George Parry, Supt Police, Mahé to James Stevens, Supervisor, 10 April 1794. 78  PP 1828 XXIV [125], 729, T. H. Baber to the Chief Secretary to Government, 9 January 1813. On slavery in Travancore, see K. K. Kusuman, Slavery in Travancore (Trivandrum: Kerala Historical Society, 1973). 79  IOR: P/322/65, pp. 1296–1408, Thos Hervey, Judge and Magistrate Zilla N[orth] Malabar, to the Secretary to Government, 29 February 1812. 80  IOR: P/322/68, p. 2850, Wm Thackeray, Chief Secretary to Govt, to Judge and Magistrate Zillah North Malabar, 29 May 1812. 81  Hubert Gerbeau, “L’esclavage et son ombre: L’île Bourbon au XIXe et XXe siècles” (Thèse pour le doctorat d’État, Université de Provence [Aix-Marseille I], 2005), 1310. 82  PP 1828 XXIV [125], 933, J. M. Macleod to the Advocate General, 9 August 1825. 83  Allen, “The Mascarene Slave-Trade,” 41. 84  PP 1828 XXIV [125], 3, Extract Bengal Revenue Consultations, 17 May 1774.

Carrying Away The Unfortunate From India And Southeast Asia

1433

during the late eighteenth century is suggested by estimates that 14,750 to 18,200 Indian men, women, and children were exported to Mauritius and Réunion between 1770 and 1810, mostly before 1793.85 French slave trading was centered at Chandernagore in Bengal, Pondichéry and Yanam on the Coromandel Coast, and Mahé on the Malabar Coast. Europeans also acquired slaves at Chittagong and Kedgeree (Kagery) in Bengal, at Bimilipatnam, Cuddalore, Ganjam, Porto Novo (Parangipettai), Wattara, and Vizagapatam (Visakhapatnam) along the Coromandel Coast, and at Calicut (Kozhikode) and Cochin on the Malabar Coast during the 1780s and early 1790s. The degree to which other European nationals participated in the Mascarene trade is difficult to ascertain, but they clearly did so. In August 1789, GovernorGeneral Cornwallis advised the Court of Directors that “an infamous traffic has, it seems, long been carried on in this country by the low Portuguese, and even by several foreign European seafaring people and traders, in purchasing and collecting native children in a clandestine manner, and exporting them for sale to the French islands, and other parts of India.”86 Company officials in southern India were aware of Dutch involvement in the Mascarene trade, especially during the early 1790s, when VOC factories at Pulicat, Bimilipatnam, and Jaggernatpooram facilitated the acquisition and shipment of slave cargoes to Mauritius and Réunion.87 In January 1792, for instance, Anthony Sadlier and his colleagues at Masulipatam wrote to their Dutch counterparts at Jaggernatpooram that they had received information that contractors reside openly at your factory, who furnish the commanders of vessels with slaves, by which means upwards of five hundred persons have lately been exported. We are also informed, that this traffic is openly countenanced by you, and written passes granted, in consideration of which, seven rupees per head is paid for each slave exported; that two subjects of France, named Burelle and La Touche, as also another, Portuguese, reside as slave contractors at your factory, by whom two vessels named L’Aimable Marie and La Baillie de Souffrein, have lately been dispatched, as also a sloop belonging to a Mr. Boulle, about three weeks

85  Allen, “The Mascarene Slave-Trade,” 41. 86  PP 1828 XXIV [125], 13, Extract of a Letter from Lord Cornwallis, Governor General of India, to the Court of Directors; dated 2d August 1789. 87  IOR: P/252/76, Jacob Eilbracht et al. to Sir Charles Oakeley, 2 February 1792; P/252/77, Jacob Eilbracht et al. to Sir Charles Oakeley and Members of the Council of Fort St. George, 19 February 1792.

1434

Allen

ago; and that there is at this time a French ship, called Les Deux Amis Reunis, laying at your factory, for the express purpose of taking in slaves.88 A Mauritian déclaration d’arrivée dated 11 May 1792 confirms that Les Amis Réunis arrived at Port Louis with a cargo of slaves from the Coromandel Coast.89 While no record exists of L’Aimable Marie arriving from India about the same time, the ship’s status as a slaver is confirmed by the fact that it left Port Louis on 10 November 1792 to trade for slaves along the Malagasy coast.90 Except for the acquisition of Mozambican slaves at Goa for shipment to the Mascarenes during the 1770s and 1780s,91 information about the mechanics of European slave trading in India remains largely hidden from our view.92 In spite of their importance as slave trading centers, studies of the French factories in India make little or no reference either to slavery or slave trading.93 The archival record has also yet to give up accounts similar to those left by Dutch and French captains who traded along the East African and Malagasy coasts.94 88  PP 1828 XXIV [125], 480, Anthony Sadlier et al. to Cas Leon Eilbracht, Chief, &c. Council at Jaggernautporam, 16 January 1792. 89  MNA: F 10/251, 11 mai 1792. 90  MNA: F 10/382, 15 janvier 1793. The ship returned to Port Louis on 13 January 1793 with a cargo of unknown size. 91  de Souza, “French Slave-Trading.” 92  Gerrit J. Knaap, “Europeans, Mestizos and Slaves: The Population of Colombo at the End of the Seventeenth Century,” Itinerario 5, no. 2 (1981): 96. 93  The one notable exception is Ray, “A Note on Slaves and Slavery.” Cf. Dominique Taffin, “Citoyens et malabars à Pondichéry pendant la Révolution française,” in Révolution française et océan Indien: Prémices, paroxysms, heritages et deviances, ed. Claude Wanquet and Benoît Jullien (Paris: L’Harmattan, 1996), 235–48; Ernestine Carreira, “La factorie française de Surat et les ports de l’Inde portugaise (1668–1778),” in Les relations entre la France et l’Inde de 1673 à nos jours, ed. Jacques Weber (Paris: Les Indes Savantes, 2002), 23–38; Frédéric Mantienne, “Le commerce intra-asiatique française aux XVIIe et XVIIIe siècles,” in Weber, Les relations entre la France et l’Inde, 49–69; K. S. Mathew, “Missionaries from the Atlantic Regions and the Social Changes in French Pondicherry from the Seventeenth to Nineteenth Centuries,” in Weber, Les relations entre la France et l’Inde, 349–69. 94   Robert Ross, “The Dutch on the Swahili Coast, 1776–1778: Two Slaving Journals,” International Journal of African Historical Studies, Part I, 19, no. 2 (1986): 305–60, and Part II, 19, no. 3 (1986), 479–506; Piet Westra and James C. Armstrong, Slave Trade with Madagascar: The Journals of the Cape Slaver Leijdsman, 1715/Slawe-Handel met Madagaskar: Die Joernale van die Kaapse slaweskip Leijdsman, 1715 (Cape Town: Africana Publishers, 2006); Andrew Alexander, “Negotiation, Trade and the Rituals of Encounter: An Examination of the Slave-Trading Voyage of De Zon, 1775–1776,” Itinerario 31, no. 3 (2007): 39–58; Thomas Vernet, “La première traite française à Zanzibar: Le journal de bord

Carrying Away The Unfortunate From India And Southeast Asia

1435

While Europeans usually acquired slaves from Indian suppliers,95 the extent to which these dealers operated in tandem with or independently of indigenous merchants who delivered textiles and other export commodities to European factories is unknown.96 As the instructions that Alexis Joseph Bartro, captain of Le Chevalier d’Entrecasteaux, received from Mr. de la Villecollet, the vessel’s armateur, at Port Louis in April 1789 indicate, European merchants in India clearly facilitated, if not participated directly in, supplying ships with slave cargoes. Bartro was directed that upon reaching Calcutta or Chandernagore, he was to work with de la Villecollet’s correspondent in Bengal, a Mr. de Vérinne, to procure 80 male and female slaves in addition to other merchandise for sale in the Mascarenes. The captain also had specific instructions about the slaves he was to purchase: all had to be between twelve and eighteen years old, good looking, in good physical condition, and have had smallpox, while males were to be at least five feet two inches tall and females as tall as possible.97 Whether Bartro succeeded in obtaining his human cargo is unknown. In some instances, Europeans were directly involved in procuring South Asian slaves for export. Portuguese freebooters sold captives taken along the Arakanese coast in 1611 into slavery in the East Indies and subsequently participated in raids on the Bengali coast organized by the Arakanese kingdom of Mrauk-U, which supplied slaves to Batavia and other VOC establishments in Southeast Asia between circa 1630 and the mid-1660s.98 Scattered reports from du vaisseau L’Espérance, 1774–1775,” in Civilisations des mondes insulaires (Madagascar, îles du canal de Mozambique, Mascareignes, Polynésie, Guyanes), ed. Chantal Radimilahy and Narivelo Rajaonarimanana (Paris: Karthala, 2011), 477–521. 95  Carter, “Indian Slaves,” 237; Vink, “‘The World’s Oldest Trade,’” 153; Seijas, “The Portuguese Slave Trade,” 24. 96  Cf. Ajit K. Neogy, “Early Commercial Activities of the French in Pondicherry: The Pondicherry Authorities, the Jesuits and Mudaliars,” in Mariners, Merchants and Oceans: Studies in Maritime History, ed. K. S. Mathew (New Delhi: Manohar Publishers, 1995), 333–40; Joseph J. Brennig, “Chief Merchants and the European Enclaves of SeventeenthCentury Coromandel,” Modern Asian Studies 11, no. 3 (1977): 321–40; Bhaswati Bhattacharya, “The Hinterland and the Coast: The Pattern of Interaction in Coromandel in the Late Eighteenth Century,” in Politics and Trade in the Indian Ocean World: Essays in Honour of Ashin Das Gupta, ed. Rudrangshu Mukherjee and Lakshmi Subramanian (Delhi: Oxford University Press, 1998), 19–51; K. S. Matthew, “Indian Merchants and the French during the Seventeenth and Eighteenth Centuries—A Study in Partnership,” in K. S. Mathew, French in India and Indian Nationalism (1700 A.D.–1963 A.D.) (Delhi: B. R. Publishing Corporation 1999), vol. 2, 611–28. 97  MNA: JH 10, Instructions et conditions de Mr Bartro Capitaine du Nre le Chr Dentrecasteaux. 98  Sinnappah Arasaratnam, “Slave Trade in the Indian Ocean in the Seventeenth Century,” in Mathew, Mariners, Merchants, 198; Michael W. Charney, “Crisis and Reformation in a

1436

Allen

the late eighteenth century provide additional evidence of instances when British and French nationals were directly involved in securing captives. In April 1790, Andrew Perry, master and owner of a small coasting vessel, asserted that the suffering he witnessed at Juggernaut [sic], near Cuttack, because of the famine then ravaging the area had prompted him to order his dubash (interpreter) to go to Juggernaut with money “to contribute to the necessities of those who most needed it.” The dubash returned with children sold willingly by their parents who, Perry asserted disingenuously, he wished only to convey “to other parts where they might procure a livelihood” and thereby better their situation.99 Two years later, a Frenchman named Guilliard, commander of a small schooner anchored at Ganjam, apparently personally enticed a boy on board the ship, from which he was subsequently rescued by British authorities together with four women and a man who had been spirited aboard the vessel earlier in the dead of night.100 As the actions of a Mr. St. Croix demonstrate, Europeans were not averse to enslaving individuals elsewhere in the region if the opportunity presented itself. During a stop at the Andaman Islands, probably in January 1790, St. Croix and his crew forcibly took a man, two boys, and a woman with her child from the canoe in which they were traveling. According to Francis Light, officer-in-charge at the East India Company’s settlement at Penang, St. Croix stated that following this incident, “his intention was to get a large vessel, and return to the Andamans, where he made no doubt of being able to procure a cargo of Caffrees, that would sell exceedingly well at the French islands.”101 Indian men, women, and children were enslaved for various reasons. Arakanese raids along the Bengali coast during the seventeenth century generated thousands of slaves, many of whom were sold to the VOC and shipped to Southeast Asia.102 The scale of this activity is suggested by a report that Angarcale and Dianga, two ports on the Arakanese coast that had originally Maritime Kingdom of Southeast Asia: Forces of Instability and Political Disintegration in Western Burma (Arakan), 1603–1701,” Journal of the Economic and Social History of the Orient 41, no. 2 (1998): 197, 203. On operations by Portuguese acting independently of the Estado da India, see George Winius, “Portugal’s ‘Shadow Empire’ in the Bay of Bengal,” Camões Center Quarterly 3, no. 1–2 (1991): 36–45. 99  PP 1828 XXIV [125], 473, Petition of Andrew Perry, 13 April 1790. 100  PP 1828 XXIV [125], 492, Thomas Snodgrass to Sir Charles Oakeley, 26 August 1792. 101  PP 1828 XXIV [125], 29, Francis Light to Edward Hay, 1 April 1791. 102  Arasaratnam, “Slave Trade in the Indian Ocean,” 197–98; Charney, “Crisis and Reformation,” 201. On deltaic Bengal as a slave-raiding zone, see also Rila Mukherjee, “Mobility in the Bay of Bengal World: Medieval Raiders, Traders, States and the Slaves,” Indian Historical Review 26, no. 1 (2009): 109–29.

Carrying Away The Unfortunate From India And Southeast Asia

1437

been just small fishing villages, handled 18,000 new slaves during a five-year period in the mid-seventeenth century.103 Political instability and warfare in central and southern India generated many of the slaves exported from the Coromandel Coast during the same century.104 In other instances, those funneled into this traffic were the victims of kidnapping and deception. Thomas Mills and John Milward reported in July 1622 that most of the slaves bought by the Dutch near Pulicat had been “stollen uppon the highwayes” or taken forcibly from their parents and friends. So great was the fear of being kidnapped, they continued, that country people no longer frequented local markets, one consequence of which was a shortage of provisions in Masulipatam.105 Fortyfive years later, in a case redolent with irony, the sentence of eight of the eleven members of a Bombay-based gang convicted of kidnapping people for sale was commuted from death to transportation to St. Helena as slaves.106 The forcible abduction and sale of men, women, and children not only continued during the eighteenth century but was also reported to be a wellestablished practice in some parts of the subcontinent. In 1792, W. G. Farmer reported, “It has long been the custom of the Moplas [in Malabar] to steal the Children of the Nairs and other Gentoo Casts and carry them to the coast for sale.”107 Two years later, officials in Bombay, citing the commission that had investigated conditions in Malabar province during 1792–93, informed London that the traffic in children kidnapped by the Moplas was particularly active not only at the French settlement at Mahé but also at the Dutch factory at Cochin and even, “though we believe in a less degree,” at the company’s own establishment at Tellicherry (Thalassery).108 The kidnapping and sale of children in Malabar would resurface as a subject of official concern twentyseven years later.109 103  Charney, “Crisis and Reformation,” 204. 104  Basu, “Notes on Slave Trade,” 26; Vink, “ ‘The World’s Oldest Trade,’ ” 140–43. 105  Foster, The English Factories in India, 1622–1623, 105. 106  Sir Charles Fawcett, The English Factories in India, vol. 1 (n.s.), The Western Presidency, 1670–1677 (Oxford: Clarendon Press, 1936), 173. 107  IOR: P/E/5, Report of W. G. Farmer, Calicut, 17 May 1792. “Moplah” usually referred to the indigenous Muslim population in Malabar but could also refer to indigenous Christians in Cochin and Travancore (Henry Yule and A. C. Burnell, Hobson-Jobson: The Anglo-Indian Dictionary [Ware, Herts., UK: Wordsworth Reference, 1996], 585). 108  IOR: E/4/473, George Dick, Daniel Crokatt (?), and William Lewis to Court of Directors, 25 September 1794 (para. 4). 109  IOR: F/4/1128/30154, Attention shown by the French authorities at Mahé to the representations of the Magistrate at Malabar on the subject of Kidnapping Children from the Company’s Territories [1819].

1438

Allen

Others, especially children, fell victim to deceit and duplicity.110 Boys and girls recovered by the police in the early 1790s often reported that they had enticed by “sweet words” or the promise of food and clothing to go to places where they were held for sale.111 Fourteen-year-old Joomun, a day laborer rescued from the Hero at Calcutta in October 1789, reported that the ship’s serang (boatswain) had him carry a load aboard the vessel, where he was then detained against his will, while Sahebodin, a petty criminal who had been confined in jail and was also recovered from the Hero, stated that police officers had delivered him to the vessel’s captain.112 On occasion, the forcible abduction of children incited strong reactions from local populations. Such was the case in Yanam in December 1792, when the news that a nine-year-old Brahmin boy numbered among those sent aboard two French slavers anchored at Coringa prompted “the Inhabitants and Country People in the place,” already disturbed that many higher-caste children had been carried away from Yanam, to demand “clamorously” that their children be restored to them.113 Following the ships’ hasty departure from Coringa shortly thereafter, at least eight hundred local inhabitants attacked and demolished the house belonging to the owner of one of the two vessels in question.114 The kidnapping and sale of children and the deceptions used to lure slaves away from their legal owners for sale elsewhere remained a subject of official concern well into the early nineteenth century.115 Other acts of enslavement entailed less overt violence or duplicity but were no less insidious. The famines that followed in the wake of the natural disasters that periodically plagued the subcontinent generated many of the slaves

110  On the regional trafficking in child slaves, see Richard B. Allen, “Children and European Slave Trading in the Indian Ocean during the Eighteenth and Early Nineteenth Centuries,” in Children in Slavery through the Ages, ed. Gwyn Campbell, Suzanne Miers, and Joseph C. Miller (Athens: Ohio University Press, 2009), 35–54. 111  PP 1828 XXIV [125], 56, No 59—Midnapore, the 12th November 1793. 112  IOR: P/3/48, pp. 500–501, Police Office, 28 October 1789. 113  IOR: P/241/36, pp. 14–15, Matthew Yeats to Sir Charles Oakeley, 22 December 1792. 114  IOR: P/241/37, pp. 511–12, Rt Malcolm to Mathew Yeats, 11 January 1793. 115  IOR: F/4/566/13970, Reports of Sir R. Dick &a relative to the practice stated to have been prevalent of Inveighling away and Selling Slaves within the Division of Dacca [1813, 1816]; F/4/702/19065, Correspondence with the Superintendant of Police relative to the practice of Kidnapping Children from their Parents for the purpose of selling them as Slaves [1818]; F/4/1034/28499, Slavery, Kidnapping and Sale of Children in Tanjore and Tinnevelly [1825].

Carrying Away The Unfortunate From India And Southeast Asia

1439

exported from the Coromandel Coast during the seventeenth century.116 The letter that reported the arrival of 400 slaves on a Portuguese ship at Aceh in October 1646 observed that the “very strong” thirteen-month-long famine then afflicting the region around Nagapatnam had led many people to “give themselves up for slaves to any man that will but feede them, all kinds of provisions, especially graine, beeing att excessive rates.”117 Famine remained a powerful, malevolent ally of slavers during the remainder of the century. Three severe famines plagued Madras between the mid-1640s and the late 1680s, the last of which resulted in the death of 35,000 out of a total population of 300,000.118 Droughts, floods, crop failures, and their attendant famines likewise forced many men and women to sell themselves or their children into bondage during the eighteenth century. Serious famines befell South India in 1718–19 and 1728– 36, while ten “subsistence crises,” varying in intensity from temporary dearth to major famine, occurred in the region between 1747 and 1798.119 The famine that racked Bengal in 1770 killed an estimated 10 million people, one-third of the province’s population. Not unsurprisingly, in such situations men and women readily sold themselves or their children into slavery in desperate attempts to stay alive.120 The famine that swept central Bengal in 1785 after severe flooding is an illustrative case in point. Matthew Day reported from Dacca that because of this catastrophe, “many hundred, I may say thousands, of unhappy Writches [sic] are now laying on the Banks of the Burrumpoter [Brahmaputra], some in the Agonies of death, and others emaciated by Famine, with hardly Strength to crawl along imploring the Assistance of Passengers.” The scarcity of grain and the high price of what grain could be found, Day continued, had reduced the district’s poor and lower classes “to the lowest Pitch of Misery and Distress,” with no other option than to sell their children, “many hundreds” of whom were immediately sent to Calcutta and its environs for subsequent shipment elsewhere.121 The famine that ravaged the Northern Circars in the early 1790s likewise prompted people to sell themselves or their dependents 116  Basu, “Notes on Slave Trade,” 22, 27; Arasaratnam, “Slave Trade,” 199, 203–4; Vink, “ ‘The World’s Oldest Trade.’ ” 117  Foster, The English Factories in India, 1646–1650, 54, n. 2. 118  B. V. Narayanswami Naidu, “Famines in the City of Madras,” Madras Tercentenary Commemoration, 75–76; Ravi Ahuja, “State Formation and ‘Famine Policy’ in Early Colonial South India,” Indian Economic and Social History Review 39, no. 4 (2002): 361. 119  Ahuja, “State Formation,” 361, 366. 120  Ibid, 352. See also David Arnold, “Hunger in the Garden of Plenty: The Bengal Famine of 1770,” in Dreadful Visitations: Confronting Natural Catastrophe in the Age of Enlightenment, ed. Alessa Johns (New York: Routledge, 1999), 81–111. 121  See n. 8 above.

1440

Allen

not only to Andrew Perry’s dubash but also to merchants from other parts of the subcontinent willing to maintain them,122 while a famine in western India drove large numbers of people into slavery in Bombay in 1790 and again in 1803.123 The passage of time did nothing to alleviate this problem. In July 1815, Judge W. T. Smith reported that the famine that stalked the Bengali districts of Dinajpur, Purnia, Rajashahi, and Rangpur from June through October 1814 had aggravated the long-established practice of selling children “and even of men and women selling themselves” in times of severe economic distress.124 Contemporary sources shed little light on the specific geographical provenance of the men, women, and children exported from the subcontinent. Slaves of South Asian origin in the Mascarenes were often described simply as indien, although the somewhat less generic bengali and malabar were also frequently used descriptors. Contemporary observers likewise noted the presence of “Talingas” (Telegus) among the islands’ Indian bondmen and -women,125 a descriptor that also appears occasionally in other sources, such as slave censuses. Unfortunately, it is impossible to determine the proportion of Bengalis, Malabars, and Talingas among the islands’ Indian slaves. Early nineteenth-century reports by British district magistrates provide additional, albeit indirect, information about the regions that may have supplied the late eighteenth-century export trade. In addition to districts such as Dacca, where famine generated large numbers of slaves, at least in the short term, these reports suggest that the districts of Buckergunge (Bakerganj), Chittagong, and Sylhet may have been relatively consistent sources of supply for the Bengali trade.126 Sylhet figures prominently in these accounts; in 1813, R. K. Dick noted that while it was impossible to calculate the number of slaves now being exported from the district each year, the number of such exports appeared to be “much less considerable than formerly.”127 Three years later, 122  IOR: P/241/17, p. 640, [Petition] To the Worshipfull James Taylor Esqre [by] Merchants &c of foreign Countries of Southward [March 1790]; P/241/37, pp. 891–92, Petition of Southern Cholia merchants now at Madras to the Hon. Sir Charles Oakeley [22 February 1793]. 123  Jeanette Pinto, Slavery in Portuguese India, 1510–1842 (Bombay: Himalaya Publishing House, 1992), 99. 124  IOR: F/4/578/14078, pp. 21–23 [Report of W. T. Smith, 2nd Judge of the Court of Circuit of Moorshedabad, 5 July 1815]. 125  Milbert, Voyage pittoresque, vol. 2, 169; Raymond Decary, Les voyages du chirugien Avine à l’Île de France et dans la mer des Indes au début du XIXe siècle (Paris: G. Durassié, 1961), 17. 126  IOR: F/4/566/13970, p. 19, S. Bird and G. Hartwell to M. H. Turnbull, 26 April 1816; F/4/566/13970, p. 31, J. Hayes to H. Walters, 10 February 1816. 127  IOR: F/4/566/13970, p. 11, Sir R. K. Dick, 2nd Judge, Dacca Court of Circuit, to M. H. Turnbull, 12 March 1813.

Carrying Away The Unfortunate From India And Southeast Asia

1441

J. W. Sage also emphasized Sylhet’s importance as a conduit through which children from the neighboring Assamese tribal territories of Cachar and Jaintia passed before being sold in other places.128 Reports during 1811 about the movement of enslaved Nepalese children into districts in what is now the state of Uttar Pradesh underscores the fact that tribal hill peoples were vulnerable to enslavement and possible exportation.129 A detailed inquiry during 1819 into slavery in the Madras Presidency suggests that, as in Bengal, some areas in southern India supplied slaves for the export trade with greater regularity than did others. The extent to which districts may have done so depended in part on the rights attached to “slave” caste status. These rights varied from region to region, with local law and custom often limiting masters’ ability to sell their bondmen or send them away from a particular locale.130 Officials in Salem, Coimbatore, Tanjore, Tinnevelly (Tirunelveli), South Arcot, Chingleput (Chengalpattu), and Canara (Kanara) reported accordingly that the sale of slaves away from their homes was an uncommon occurrence in their districts.131 In other districts, however, masters had a freer hand in disposing of their servile dependents. Such was the case in Trichinopoly, where C. M. Lushington observed that while “Pullers” (Pulayas)132 were usually sold with the land they worked, these laborers were also frequently sold independently of the land.133 As noted earlier, Nagapatnam, the region’s principal port, exported significant numbers of slaves during the seventeenth 128  IOR: F/4/566/13970, pp. 32–33, J. W. Sage to Shearman Bird et al., 12 February 1816. 129  IOR: F/4/369/9221, Relative to the Measures adopted for the prevention of the Trade in Children carried on between the Upper Provinces and the Territories of Napaul. 130  On agrestic servitude in southern India, see Dharma Kumar, Land and Caste in South India: Agricultural Labour in the Madras Presidency during the Nineteenth Century (Cambridge: Cambridge University Press, 1965), 34–63. 131  IOR: F/4/919/25850, p. 53, E. R. Hargrave [Salem] to the Board of Revenue, 14 June 1819; p. 62, J. Sullivan [Coimbatore] to the Board of Revenue, 24 June 1819; pp. 71–72, J. Hepburn [Tanjore] to the Board of Revenue, 30 June 1819; p. 78, J. Cotton [Tinnevelly] to the Board of Revenue, 30 June 1819; pp. 90–91, C. Hyde [South Arcot] to the Board of Revenue, 12 September 1819; p. 113, W. Cooke [Chingleput] to the Board of Revenue, 31 July 1819; p. 162, T. Harris [Canara]to the Board of Revenue, 10 July 1819. 132  K. Saradamoni, Emergence of a Slave Caste: Pulayas of Kerala (New Delhi: People’s Publishing House, 1980). See also Sebastian Joseph, “Slave Labour of Malabar in the Colonial Context,” in Essays in Modern Indian Economic History, ed. Sabyasachi Bhattacharya (Delhi: Munshiram Manoharlal, 1987), 46–54. 133  IOR: F/4/919/25850, p. 134, C. M. Lushington, Collector of Trichinopoly, to A. D. Campbell, Secretary to the Board of Revenue, 1 July 1819. It should be noted that Lushington also contradicts himself on this point.

1442

Allen

century.134 Lushington also noted the “indiscriminate” sale of men, women, and children in Malabar, a report seconded by that district’s collector, who observed that local slaves were frequently transferred from one owner to another by sale, mortgage, or hire.135 The extent to which individuals of “slave” caste status may have been exported by Europeans or indigenous merchants during the seventeenth and eighteenth centuries is impossible to ascertain. Information about other aspects of this export trade also tends to be sparse. Famines depressed local slave prices, often dramatically so, as large numbers of desperate people sold themselves or their dependents into slavery in attempts to keep body and soul together. In 1618, one pagoda could purchase 3 or 4 children around famine-ravaged Tirupapaliyur in southern India.136 Little had changed almost two centuries later, when Judge Smith reported that those victimized by famine in Purnia and surrounding districts in 1814 sold in public markets “for [a] less sum than would purchase a Cat or Dog.”137 Ship manifests shed additional light on slave prices during the early seventeenth century. The cargo of the Nieuw Zeeland, bound from Pulicat to Batavia via Masulipatam in November 1619, included 11 boys and 8 girls valued at 374 florins (approximately 20 florins each), while two and a half years later, the Sampson left Surat with 83 male and female slaves valued at 4,672 florins (approximately 56 florins each).138 Prices could vary widely from one region to another, depending on local conditions and circumstances. In 1622, for instance, slaves who cost no more than 8 rials apiece along the Coromandel Coast in 1622 fetched almost 20 rials each at Surat.139 Price data from the eighteenth century are equally fragmentary. The average value of the 245 slaves, probably Coffrees, shipped from Bombay to Bencoolen in 1765 was put at $40 each.140 Documented sales of domestic slaves provide other price data during the latter part of the eighteenth century. On 22 September 1774, the widow Baptiste sold Jamal, an eleven-yearold boy, to Mr. Lachenay in Chandernagore for Rs 70, while almost three years later the widow Boufet sold Manon, purchased by her husband for Rs 7½ in 134  Vink, “ ‘The World’s Oldest Trade,’ ” 141–42. 135  IOR: F/4/919/25850, p. 174, J. Vaughan [Malabar] to the Board of Revenue, 20 July 1819. Major James Bayley, who surveyed the area between Tanjore and Malabar in 1808–9, also reported that slaves were numerous and “bought & sold at pleasure” in Malabar (BL: Eur. Mss. D 970, Extract from “Observations on Survey in the Year 1808–9”). 136  Prakash, The Dutch Factories, 80. 137  See n. 124 above. 138  Prakash, The Dutch Factories, 120, 194, respectively. 139  Ibid., 203. 140  IOR: G/35/13, fols. 215v–216r (para. 62), Roger Carter et al. to Court of Directors, 20 March 1765.

Carrying Away The Unfortunate From India And Southeast Asia

1443

1770, to Mlle. Louise Lachenay for Rs 20.141 The extent to which these prices corresponded to those paid for slaves exported from coastal comptoirs remains unclear. Somewhat more is known about slave prices in early nineteenth-century southern India. The 1819 Madras Presidency inquiry revealed that a “good” slave cost as much as Rs 50 in Coimbatore,142 while the value of male and female slaves in Tinnevelly ranged from 15 chuckrums to Rs 3 and 5 chuckrums to Rs 3, respectively.143 Slaves in Trichinopoly commanded between 5 and 10 pagodas, depending on their age and qualifications.144 Strong young men in Canara usually cost between Rs 12 and Rs 26, while a strong young woman fetched from Rs 12 to Rs 24.145 Other evidence suggests that child slaves could be purchased for one-quarter to one-third of what adults commanded, if not less.146 Once again, the extent to which these figures may be indicative of late eighteenth-century export prices cannot be ascertained. Like those shipped from Africa to the Americas and into the wider Indian Ocean world, slaves exported from the Indian subcontinent were mostly young men and women. Information about the demographic structure of individual cargoes remains scarce. While some of the 130 Bengali slaves loaded on the Friendship at Chandernagore in 1789 were reported to be as old as twenty, the majority ranged from eight to sixteen years of age, with the youngest being only five or six years old.147 The 1827 Mauritian slave census, our only source of demographic information about the island’s Indian bondmen and women, suggests that the composition of Indian cargoes arriving in the Mascarenes generally matched that of French slavers crossing the Atlantic during the eighteenth century.148 The census enumerated 1,459 males and 892 females among the island’s Indian slaves, all of whom, not unsurprisingly, were reported to be 141  BL: Eur. Mss. F 193/45. 142  IOR: F/4/919/25850, p. 62, J. Sullivan [Coimbatore] to the Board of Revenue, 24 June 1819. 143  IOR: F/4/919/25850, p. 82, J. Cotton [Tinnevelly] to the Board of Revenue, 30 June 1819. The value of the chuckrum vis-à-vis other currencies varied through time. In 1711, one pagoda was worth approximately 1.8 chuckrums; in 1813, one chuckrum equaled 4/9 pagoda (Yule and Burnell, Hobson-Jobson, 217). 144  IOR: F/4/919/25850, p. 134, C. M. Lushington [Trichinopoly] to A. D. Campbell, 1 July 1819. 145  IOR: F/4/919/25850, p. 161, T. Harris [Canara] to the Board of Revenue, 10 July 1819. 146  Joseph, “Slave Labour in Malabar,” 49. 147  IOR: P/3/46, p. 12, Affidavit of John Peters respecting Captain Peter Horrebow, 1 June 1789. See also PP 1828 XXIV [125], 14–16, Extract, Bengal Public Consultations, 3d June 1789. 148  TSTD2 reports that males comprised an average of 64.6 percent of French transatlantic voyages during the eighteenth century, while children accounted for an average of 20.9 percent of such cargoes. See also David Geggus, “Sex Ratio, Age and Ethnicity in

1444

Allen

sixteen years of age or older.149 These data point to a 2:1 male-to-female ratio among Indian slaves exported to the Mascarenes, a ratio supported by censuses in 1785, 1788, and 1809, which reveal that males outnumbered females in the island’s total slave population by approximately 2:1.150 These censuses also reveal that children, usually defined as boys and girls less than fifteen years of age, comprised one-fifth of the island’s chattel population during the 1780s, with boys outnumbering girls by a ratio of 3:2. In 1809, children comprised 23 percent of all slaves on the island, with boys outnumbering girls by a margin of 54:46.

Trading “Malays”

Europeans transported enslaved men, women, and children from Southeast Asia as well as India into the wider Indian Ocean world. Although the export trades from these two regions shared some features in common, they also differed in ways that further illustrate the complexities of European slave trading in the Mare Indicum between the early sixteenth and mid-nineteenth centuries. Perhaps the most salient difference is the number of exports beyond each of these catchment areas. While the VOC shipped at least 26,000 to 38,000 and perhaps as many as 100,000 Indian slaves to the Cape Colony, Ceylon, and Indonesia during the seventeenth century, and French ships left India for Mauritius and Réunion during the eighteenth century with as many as 29,000 Indian men, women, and children on board, substantially fewer “Malay” slaves reached European settlements and colonies in the wider Indian Ocean world. Of the 63,000 slaves imported into South Africa from the midseventeenth century to the end of the eighteenth, approximately 14,300 (22.7 percent) are estimated to have been of Southeast Asian origin compared to the 16,300 of Indian origin.151 Contemporary observers confirm that the Île de France housed “Malay” slaves at the beginning of the nineteenth century, but the passing nature of such references indicate that only small numbers of such

the Atlantic Slave Trade: Data from French Shipping and Plantation Records,” Journal of African History 30, no. 1 (1989): 23–44. 149  MNA: ID 14, Return of the slave population of Mauritius [1827]. 150  R. R. Kuczynski, Demographic Survey of the British Colonial Empire, vol. 2 (London: Oxford University Press, 1949), 760–61. 151  Shell, Children of Bondage, 40–41. On these slaves’ more specific place of origin, see Worden, “Indian Ocean Slavery,” 31.

Carrying Away The Unfortunate From India And Southeast Asia

1445

slaves resided in the colony.152 Men and women from Bali and Nias numbered among the 107,000 or more slaves imported illegally into the Mascarenes between 1811 and the mid-1830s, but various sources suggest that no more than 3,000 to 3,550 slaves of Southeast Asian origin reached Mauritius and Réunion during this era.153 While Indonesian slaves reached Ceylon, the Mascarenes, and South Africa between the seventeenth and early nineteenth centuries and may also have arrived at the East India Company’s Indian comptoirs during the 1680s,154 the great majority of these men, women, and children remained in Southeast Asia, where they comprised the overwhelming majority of European-owned slaves after the 1660s, when India ceased being the VOC’s most important source of chattel labor. More than 90 percent of the slaves at Kota Ambon in eastern Indonesia during the 1690s, for example, were of Southeast Asian origin.155 Although slaves labored on VOC spice plantations in the Moluccas, most worked as domestic servants and in urban centers, where they loaded and unloaded ships, constructed and maintained buildings and other facilities, and served as skilled craftsmen and artisans. Slave women also worked as prostitutes or became their masters’ concubines.156 This pattern of domestic, 152  Milbert, Voyage pittoresque, vol. 2, 162, 174; Decary, Les voyages du chirugien Avine, 17. 153  Allen, “The Mascarene Slave-Trade,” 41. The 1827 Mauritian slave census reported only 239 Malays (131 males, 108 females), all between sixteen and sixty-one years old, among the island’s 69,000 chattel residents (MNA: ID 14, Return of the slave population of Mauritius [1827]). 154  In 1683, the East India Company frigate Formosa arrived at Surat from Batavia with a consignment of slaves, presumably of Southeast Asian origin, which remained unsold at both Bombay and Surat for unknown reasons (IOR: E/3/43, No. 5001, J. Chilet et al. at Swally Marine to Court of Directors, 30 December 1683; Sir Charles Fawcett, The English Factories in India, vol. 3 [n.s.], Bombay, Surat, and Malabar Coast [Oxford: Clarendon Press, 1954], 325). 155  Gerrit J. Knaap, “A City of Migrants: Kota Ambon at the End of the Seventeenth Century,” Indonesia 51 (1991): 119. 156   On relations between European men and indigenous/slave women, see: Barbara Watson Andaya, “From Temporary Wife to Prostitute: Sexuality and Economic Change in Early Modern Southeast Asia,” Journal of Women’s History 9, no. 4 (1998): n-34; Hendrik E. Niemeijer, “Slavery, Ethnicity and the Economic Independence of Women in Seventeenth-Century Batavia,” in Other Pasts: Women, Gender and History in Early Modern Southeast Asia, ed. Barbara Watson Andaya (Honolulu: Center for Southeast Asian Studies, University of Hawai’i at Mânoa, 2000), 174–94; Dhiravat na Pombejra, “VOC Employees and Their Relationships with Mon and Siamese Women: A Case Study of Osoet Pegua,” in Andaya, Other Pasts, 195–214; Eric Jones, Wives, Slaves, and Concubines: A History of the Female Underclass in Dutch Asia (DeKalb: Northern Illinois University Press, 2010).

1446

Map 4

Allen

Southeast Asia.

Carrying Away The Unfortunate From India And Southeast Asia

1447

urban slavery was consistent with local custom and practice,157 and highlights that Europeans in Southeast Asia interacted with and conformed to already existing systems of slavery rather than imposing their own system of servitude, as they did in the Americas and South Africa.158 Involuntary bondage was well established in Southeast Asia long before Europeans arrived in the region, the local socioeconomic significance of which varied widely.159 Slaves comprised only 2.5 to 5.5 percent of Malacca’s population in the early sixteenth century but played a crucial role in the port’s functioning.160 Slavery remained of limited economic importance in precolonial west Java at the beginning of the eighteenth century.161 In other areas, however, slaves comprised a greater proportion of local populations that ranged from 10 percent among certain groups in Sarawak to about 15 percent on Nias, and as high as 30 percent among the Batak and Toraja on Sulawesi.162 As in India, Southeast Asian men and women were enslaved for various reasons, including judicial punishment, kidnapping, piracy, and war,163 while children were vulnerable to sale by impoverished parents or relatives who

157  Remco Raben, “Cities and the Slave Trade in Early-Modern Southeast Asia,” in Linking Destinies: Trade, Towns and Kin in Asian History, ed. Peter Boomgaard, Dick Kooiman, and Henk Schulte Nordholt (Leiden: KITLV Press, 2008), 121. 158  Anthony Reid, “Introduction: Slavery and Bondage in Southeast Asian History,” in Reid, Slavery, Bondage and Dependency, 14–15. 159  Lasker, Human Bondage, passim; K. Endicott, “The Effects of Slave Raiding on the Aborigines of the Malay Peninsula,” in Reid, Slavery, Bondage and Dependency, 216– 41; M. Aung Thwin, “Athi, Kyun-Taw, Hpayà-Kyun: Varieties of Commendation and Dependence in Pre-Colonial Burma,” in Reid, Slavery, Bondage and Dependency, 64– 89; Andrew Turton, “Thai Institutions of Slavery,” in Formes extrêmes de dépendance: Contributions à l’étude de l’esclavage en Asie du Sud-Es, ed. Georges Condominas (Paris: Éditions de l’École des Hautes Études en Sciences Sociales, 1998), 411–12. 160  Luis Filipe F. R. Thomaz, “L’esclavage à Malacca au XVIe siècle, d’après les sources portugaises,” in Condominas, Formes extrêmes, 360; P.-Y. Manguin, “Manpower and Labour Categories in Early Sixteenth Century Malacca,” in Reid, Slavery, Bondage, and Dependency, 209. 161  M. Hoadley, “Slavery, Bondage and Dependency in Pre-Colonial Java: The CirebonPriangan Region, 1700,” in Reid, Slavery, Bondage and Dependency, 93. 162  Anthony Reid, “ ‘Closed’ and ‘Open’ Slaves Systems in Pre-Colonial Southeast Asia,” in Reid, Slavery, Bondage and Dependency, 161. 163  Sir Thomas Stamford Raffles, The History of Java (London: John Murray, 1830), vol. 1, 260. See also Reid, “‘Closed’ and ‘Open’ Slave Systems,” 158.

1448

Allen

wished to avoid the cost and trouble of caring for orphaned boys and girls.164 The most common reason for enslavement, however, was debt,165 and debt bondage supplied many of the men, women, and children funneled into the region’s extensive slave-trading networks. Sumatran slaves reached Malacca,166 while Nias supplied bondmen and -women to Sumatra’s west coast and Aceh.167 Java exported slaves to Brunei on Borneo’s northwest coast, Malacca, Pasai on Sumatra’s north coast, and Siam, the largest regional importer of slaves during the sixteenth century.168 Bali and neighboring islands exported an estimated 100,000 to 150,000 slaves between 1620 and 1830.169 Slaves from other islands in the Indonesian and Philippine archipelagos, such as Alor, Buton, Manggarai, Mindanao, Sulu, Tawi-Tawi, and Timor, reached Makassar, from whence they were reexported, together with Bugi and Torajan slaves from Sulawesi, to Aceh, Banjarmasin, Jambi, Palembang, Sukadena, and as far away as Ayudhya in Siam.170 As Portuguese reports circa 1516 about Balinese slaves attest,171 Europeans quickly became aware of indigenous slave trading networks following their arrival in the region. These networks continued to elicit European interest 164  BL: Eur. Mss. Mack Private 82, pp. 83–84, Some particulars on the Island of Bali [by Mr. Knops, 1812]; Eur. Mss. E 108, p. 352, Report of a Journey into the Batak Country by Messrs Burton and Ward, 1824. 165  Reid, “Introduction,” 12; Endicott, “The Effects of Slave Raiding,” 216; Thomaz, “L’esclavage à Malacca,” 361; Bryce Beemer, “Southeast Asian Slavery and Slave-Gathering Warfare as a Vector for Cultural Transmission: The Case of Burma and Thailand,” The Historian 71, no. 3 (2009): 490; Hans Hägerdal, “The Slaves of Timor: Life and Death on the Fringes of an Early Colonial Society,” Itinerario 34, no. 2 (2010): 20. For a fuller discussion of debt slavery, see Lasker, Human Bondage, chap. 3, and Alain Testart, “The Extent and Significance of Debt Slavery,” Revue française de sociologie 43, supplement: An Annual English Selection (2002): 173–204. On debt and bonded labor in the wider Indian Ocean world, see Gwyn Campbell and Alessandro Stanziani, eds., Bonded Labour and Debt in the Indian Ocean World (London: Pickering and Chatto, 2013) 166  Thomaz, “L’esclavage à Malacca,” 361. 167  Edwin M. Loeb, Sumatra: Its History and People (Kuala Lumpur: Oxford University Press, 1972), 129, 230. In the early nineteenth century, Nias reportedly exported approximately 1,500 slaves each year on Achinese and Chinese vessels, the latter of which carried these slaves to Padang and Batavia (IOR: F/4/659/18295, pp. 58–59, John Prince and William Jack to Sir T. Stamford Raffles, 31 December 1820). 168  Anthony Reid, Charting the Shape of Early Modern Southeast Asia (Bangkok: Silkworm Books, 1999), 210; Turton, “Thai Institutions of Slavery,” 417. 169  Vink, “ ‘The World’s Oldest Trade,’ ” 144; Hägerdal, “The Slaves of Timor,” 21. 170  Reid, “ ‘Closed’ and ‘Open’ Slave Systems,” 170; T. Bigalke, “Dynamics of the Torajan Slave Trade in South Sulawesi,” in Reid, Slavery, Bondage and Dependency, 343–44. 171  Hägerdal, “The Slaves of Timor,” 23.

Carrying Away The Unfortunate From India And Southeast Asia

1449

throughout the seventeenth and eighteenth centuries. East India Company officials at Bencoolen, for example, made a point of informing London in 1708 that Nias, from which they had already acquired chattel laborers, had been a source of slaves “long before our time.”172 More than sixty years later, JeanBaptiste Pelon reported that slaves were traded on Timor as they were “everywhere else” in the region.173 As repeated complaints at Bencoolen illustrate, the high cost of local free labor encouraged Europeans to tap regional slave trading networks. Early in 1686, only months after the factory’s establishment, Benjamin Bloome and Joshua Charlton informed London that slaves were “absolutely necessary” to its survival, since local free workers were not only expensive, refusing to work for less than a quarter of a dollar a day, but also lazy and so temperamental that they could not be disciplined.174 Although Bencoolen relied initially on imported Coffrees, high morbidity and mortality rates because the settlement was “so sickly” compelled officials to seek chattel labor from elsewhere in the region.175 In March 1702, Richard Watts and his colleagues informed Madras that since “slaves in Plenty are to be procured at Nayas [sic],” they hoped Governor Thomas Pitt and his council would order a ship to call there en route to Bencoolen to procure as many as might be had.176 Nias slaves apparently reached the factory no later than February 1704.177 Nine years later, the settlement’s 188 slaves included 16 men, 13 women, 6 boys, and 10 girls from Nias, 172  IOR: G/35/6, Richd Skingle et al. to Court of Managers, 2 January 1708. 173  Jean-Baptiste Pelon, Description de Timor occidental et des îles sous domination hollandaise (1771–1778), edited and annotated by Anne Lombard-Jourdan, Cahier d’archipel 34 (2002): 13. 174  IOR: G/35/1, p. 61, Benjamin Bloome and Joshua Charlton, York Fort, to the Governor, Deputy & Committee for Affaires of the Rt Honble Eng: East India Compa …, February 1685/6. 175  Studies of Bencoolen ignore the settlement’s “Malay” slaves. See: Frenise A. Logan, “The British East India Company and African Slavery in Benkulen, Sumatra, 1687–1792,” Journal of Negro History 41, no. 4 (1956): 339–48; Alan Harfield, Bencoolen: A History of the Honourable East India Company’s Garrison on the West Coast of Sumatra (1685–1825) (Barton-on-Sea, Hamps., UK: A and J Partnership, 1995); Shihan de Silva Jayasuriya, “East India Company in Sumatra: Cross-Cultural Interactions,” African and Asian Studies 8, no. 3 (2009): 204–21. 176  IOR: G/35/5, Richard Watts et al. to Thomas Pitt, 28 March 1702. 177  IOR: E/3/95, pp. 584 (para, 11), 588 (para. 24), Orders and Instructions Given by Us the Court of Managers for the United trade to the East Indies to Mr James Crosse Governour of Bencoolen and Chief of all our Affairs on the Island Sumatra, and to Mr Abraham Hoyle second there, July 1706. Reports that men and women had cost $85 to $100 each and children $50 to $60 each aroused suspicions in London that the factory’s managers had sought to line their own pockets while purchasing these slaves.

1450

Allen

as well as 19 “Malays” (10 men, 4 women, 3 boys, 2 girls),178 most, if not all, of whom were probably of Javanese origin.179 Although the archival record confirms the presence of Nias slaves at Bencoolen into the mid-1780s,180 the frequent failure of local slave censuses to report the ethnicity of factory slaves makes it impossible to determine how many “Malays” reached Fort Marlborough during the eighteenth century. Public consultations suggest that relatively few did so during the mid-eighteenth century. Only 12 men and 3 women from Nias and one Malay man are mentioned among the settlement’s 169 adult slaves in 1748,181 while eight years later, the fort’s 357 resident slaves included 23 adult Nias males and no Malays.182 In July 1759, Richard Wyatt, in charge of the substation at Nattal, received permission to purchase “Six Stout Nias Men” on the best terms he could.183 Wyatt subsequently contracted in 1763 to supply Bencoolen with Nias slaves at a rate of $70 for each man and $50 for each woman,184 and in August that year he delivered 4 such slaves.185 The following May, however, Bencoolen’s managers opined that since they were now “fully supplied” with Nias slaves, Wyatt’s contract should be canceled, although, they hastened to add, since “these People are more expert in handicraft Businesses than the Coffrees, the Board should be glad of (30 or 40) Thirty or Forty Young healthy Lads from (12 to 16) Twelve to Sixteen Years of age, but the Price must be adjusted when they are brought here.”186 Wyatt’s contract was finally annulled, probably in March 1765, although the factory’s managers noted that since slaves of “this Cast when Young 178  IOR: G/35/8, fol. 225, An Account of Slaves Remaining Septr Primo 1713. Three of the men and 2 of the women from Nias were described as disabled, as were 5 Malay men and a Malay woman. The remainder of the factory’s slave population consisted of 64 Coffrees (20 men, 32 women, 3 boys, 9 girls) and 60 Malabars (33 men, 16 women, 2 boys, 9 girls). 179  The 1712 slave census recorded Javanese as well as Bengalis, Coffrees, and Malabars among the factory’s chattel workers (IOR: G/35/8, fol. 62, York Fort Bencoolen Mo Septembr 1712 An Acct of the Rt Honbe Company Slaves where they are & how employ[ed]). 180  IOR: G/35/89, fol. 402, Fort Marlborough Cons., 10 November 1786. 181  IOR: G/35/9, fols. 177r–180r, [No. 1175], List of the Honble Company’s Slaves at their Settlement of Fort Marlbro the 20th Jany 1747/8. The ethnicity of the fort’s 32 boy and 36 girl slaves was not reported. 182  IOR: G/35/10, fols. 121r–124v, [No. 479], List of Slaves & Transports at Fort Marlborough and its Dependencies Pmo January 1755, and fol. 126r, Return of Slaves at Fort Marlborough and its Dependencies for Janry 1755 [compiled by Thomas Combes, Paymaster]. 183  IOR: G/35/70, p. 215, Fort Marlborough Cons., 9 July 1759. 184  IOR: G/35/13, fols. 28v–29r (para. 42), Roger Carter, Heny Idell, and Robert Nairne to Court of Directors, 22 July 1763. 185  IOR: G/35/71, Fort Marlborough Cons., 18 August 1763. 186  IOR: G/35/72, p. 76, Fort Marlborough Cons., 17 May 1764.

Carrying Away The Unfortunate From India And Southeast Asia

1451

are more easily trained to Handicraft Business than any other,” Wyatt had been informed that they would still be glad to purchase 40 or 50 boys from twelve to sixteen years of age.187 The fort’s consultations provide additional details about its enslaved “Malay” residents. After smallpox killed a majority of the settlement’s bondmen in 1704, officials informed London that had they not been supplied with slaves from Banjarmasin on Borneo’s south coast, “many of which were Debtors whose contracted time had Expired,” they would have forced to rely upon much more expensive local free labor.188 Adult Nias males were executed for murdering their masters and company personnel in 1709, 1738, and 1755,189 while an unnamed Nias man found guilty in 1746 of “villainous intentions” against the officer-in-charge at Laye was transported to St. Helena together with See Chinto, his Malay accomplice, and Lymarro, a Coffree slave, also from Laye.190 Factory records also noted the birth of a Nias slave girl in 1707191 and that “Nias people” purchased the freedom of Aneckee, an aged and infirm countryman, in 1732 for $50; of Peenore, an aged and infirm countrywoman, in 1735 for $25; and of an unnamed four-year-old child in 1741 for $15.192 The sense of community implicit in these actions remained intact forty-five years later, when the fort’s managers observed that Nias slave women rarely reported the births of their newborn children, who they gave to manumitted relatives and friends to be nursed and raised.193 While the British, French, and Portuguese all traded Southeast Asian slaves, the VOC, private Dutch traders, and commercial interests in Batavia handled the great majority of those who passed through or into European hands. Slaves constituted almost one-half of Batavia’s population in 1673, a figure that climbed to 59 percent in 1679 and then to 61 percent by 1749.194 Bondmen and -women accounted for 52.3 percent of Kota Ambon’s population in 1694, a figure that increases to 60.3 percent if the town’s “unsettled” free residents

187  IOR: G/35/13, fol. 215v (para. 60), Roger Carter et al. to Court of Directors, 20 March 1765. 188  IOR: G/35/6, Richd Skingle et al. to Court of Managers, 2 January 1708. 189  IOR: G/35/56, Cons, of 25 September 1709 and 27 September 1709; G/35/62, Fort Marlborough Cons. of 9 February 1737/8 (p. 28) and 20 April 1738 (p. 730); G/35/10A, p. 78, John Pybus et al. to Court of Directors, 10 January 1755. 190  IOR: G/35/64, p. 341, Fort Marlborough Cons., 25 February 1745/6. 191  IOR: G/35/56, fol. 169, York Fort diary, 29 July 1707. 192  IOR: G/35/61, Fort Marlborough Cons. of 13 November 1732 and 29 January 1734/5; G/35/63, Fort Marlborough Cons., 21 February 1740/1. 193  IOR: G/35/89, fol. 402, Fort Marlborough Cons., 10 November 1786. 194  Hägerdal, “The Slaves of Timor,” 21; Andaya, “From Temporary Wife,” 25.

1452

Allen

are excluded from consideration.195 Chattel laborers comprised 71 percent of Makassar’s resident population in 1730.196 Overall, slaves constituted 50 percent or more of the population in VOC-administered urban areas between the mid-seventeenth and late eighteenth centuries.197 The archival record’s frequent silence about the number of slaves imported legally into Batavia, the difficulties of gauging the number of illegal imports into the city, and the problems of determining the proportion of slaves handled by the VOC and private Dutch, Chinese, and other Asian traders makes it difficult to determine how many Southeast Asian bondmen and -women reached VOC towns and territories between the 1620s and 1790s. What is clear, however, is that hundreds of thousands did so and that the volume of such imports increased over time. Markus Vink has argued that Dutch administrative centers, colonies, and factories needed to import 3,200 to 5,600 slaves each year during the late seventeenth century to maintain local slave populations, figures that, after allowing for a 20 percent average mortality rate on the high seas, required the acquisition of 4,476 to 7,716 new slaves on a yearly basis.198 Gerrit Knaap’s estimate that the value of private sector slave imports into Batavia averaged Rix$150,000 a year between 1774 and 1777 speak to the volume of this traffic during the latter part of the eighteenth century.199 Remco Raben in turn has argued that Batavia imported fewer than 100 “Malay” slaves each year during much of the 1650s, a number that doubled between the late 1650s and 1669 and then increased to more than 1,000 a year, peaking at 1,584 in 1671. Indonesian vessels delivered 9,809 slaves to the city between 1663 and 1682 (an average of 490 each year), after which Asian imports ranged from 300 to 700 a year until the late 1680s.200

195  Knaap, “A City of Migrants,” 123–24. 196  Andaya, “From Temporary Wife,” 25. 197  Abeyasekere, “Slaves in Batavia,” 286; Raben, “Cities and the Slave Trade,” 125. 198  Vink, “A Work of Compassion?” The demand for chattel labor prompted the VOC to include clauses in agreements with indigenous chiefs and headmen who promised to deliver specified numbers of slaves and other commodities to the company in exchange for its protection and military assistance. Between 1650 and 1675, the company signed such agreements with chiefs and headmen on and near New Guinea, Sulawesi, Sumatra, and Timor (Markus Vink, “Freedom and Slavery: The Dutch Republic, the VOC World, and the Debate over the ‘World’s Oldest Trade,’” Southern African Historical Journal 60, no. 1 [2008]: 42). 199  Gerrit Knaap, “All About Money: Maritime Trade in Makassar and West Java, Around 1775,” Journal of the Economic and Social History of the Orient 49, no. 4 (2006): 497–98. 200  Raben, “Cities and the Slave Trade,” 127–31.

Carrying Away The Unfortunate From India And Southeast Asia

1453

By all accounts, the number of slaves who reached Batavia increased substantially during the eighteenth century. In 1720, for instance, 2,149 slaves (1,164 men, 985 women) were imported legally into the city. Legal imports by private Asian traders averaged 1,500 a year between 1750 and the 1770s, while Jacob Rademacher, one of the city’s Dutch residents, reported in 1779 that 4,000 slaves were reaching Batavia each year. Imports decreased after the 1770s, however, as the city’s population and commercial vitality declined. Overall, some 200,000 to 300,000 slaves are estimated to have arrived in Batavia during the seventeenth and eighteenth centuries, one-half of whom were imported by the VOC, company employees engaging in private trade, and Batavian entrepreneurs, while traders based elsewhere in the region handled the other 50 percent of such imports.201 In his history of Java, Sir Stamford Raffles reported that in 1814 the largest proportion of slaves in Batavia, the city’s environs, and the districts of Semarang and Surabaya came from Bali and Celebes (Sulawesi).202 Modern scholarship confirms these two islands’ prominent role in supplying slaves to Batavia from the mid-seventeenth to the late eighteenth century.203 The structure of Balinese society, including the fragmented nature of political power, the local pattern of private land ownership, the weak ties between local states and villages, and an attendant inability to absorb all of those enslaved on the island generated a regular supply of bondmen and -women for export.204 Twenty or thirty and sometimes as many as 100 Balinese arrived in Batavia each month during the seventeenth century, many of whom were traded by Chinese merchants based in the city.205 A total of 100,000 to 150,000 Balinese are estimated to have been exported as slaves between 1620 and 1830.206 Sulawesi supplied the largest number of “Malay” slaves to reach VOC establishments, especially after the company’s destruction of the Makassar Sultanate (1666–69). Makassar’s prominence as a regional commercial entrepôt207 made 201  Ibid., 135. 202  Raffles, The History of Java, vol. 1, 84. 203  Abeyasekere, “Slaves in Batavia,” 290; H. Sutherland, “Slavery and the Slave Trade in South Sulawesi, 1660s–1800s,” in Reid, Slavery, Bondage and Dependency, 266–67; Raben, “Cities and the Slave Trade,” 129. 204  A. van der Kraan, “Bali: Slavery and Slave Trade,” in Reid, Slavery, Bondage and Dependency, 315–40. 205  BL: Eur. Mss. Mack Private 82, p. 84, Some particulars on the Island of Bali [by Mr. Knops, 1812]; Gerrit Knaap and Heather Sutherland, Monsoon Traders: Skips, Skippers and Commodities in Eighteenth-Century Makassar (Leiden: KITLV Press, 2004), 153. 206  Vink, “ ‘The World’s Oldest Trade,’ ” 144; Hägerdal, “The Slaves of Timor,” 21. 207  See Knaap and Sutherland, Monsoon Traders; Knaap, “All About Money.”

1454

Allen

it not only a market for slaves taken from among the Bugi, Toraja, and other ethnic groups on the island but also a major transshipment point through which bondmen and -women from elsewhere in the Indonesian archipelago reached Batavia. The city’s importance as a slave trading center is suggested by the value of private sector slave trading during the eighteenth century (table 17), and by estimates that 3,000 slaves passed through the city each year during the eighteenth century en route to Batavia.208 An 1816 slave census likewise attests to Sulawesi’s importance as a source of Batavian slaves. More specifically, the census indicates that 43 percent of city’s slaves came from Sulawesi, compared to 19.7 percent from Bali and 13.3 percent from the Lesser Sundas (Bima, Ende, Manggarai, Sumba, Sumbawa, Timor). The census also highlights the extent of the regional trading networks that supplied the city with chattel labor, revealing as it does the presence of slaves from elsewhere in Java (Cirebon, Surabaya), as well as Sumatra (Padang, Palembang), Nias, Borneo (Banjarmasin), the Moluccas (Ambon, Ceram, Ternate), the Banda Islands, and Papua—New Guinea in addition to those from Bali, the Lesser Sundas, and Sulawesi.209 Table 17

Private sector slave trading at Makassar, 1720s–1780s

Period

AAVa (Rix$) % Total major commodities

AAV (Rix$)

% Total major commodities

Slave prices (Rix$)

1720 1740s 1760s 1770s 1780s

11,000 — 20,000 26,500 48,500

15,000 — 30,000 40,000 70,000

33.7 — 22.8 22.5 29.1

— 40b 44–63c — 50

Slave imports

14.1 — 16.3 19.2 21.7

Slave exports

Notes: a Average annual value. b For 1743. c For males and females. Source: Gerrit J. Knaap and Heather Sutherland, Monsoon Traders: Ships, Skippers and Commodities in Eighteenth-Century Makassar (Leiden: KITLV Press, 2004), 125, 230, 235. 208  Sutherland, “Slavery and the Slave Trade,” 270. 209  Abeyasekere, “Slaves in Batavia,” 290–91.

Carrying Away The Unfortunate From India And Southeast Asia



1455

Trading Indians and Malays in Perspective

The information currently at our disposal indicates that British, Dutch, French, and Portuguese ships carried a minimum of 98,000 to 110,000 slaves away from India and Southeast Asia into the wider Indian Ocean world between the sixteenth and early nineteenth centuries. This figure, when added to the estimated 100,000 to 150,000 “Malays” who reached Batavia under Dutch auspices between the mid-seventeenth and eighteenth centuries, suggests that Europeans were directly involved in trafficking at least 198,000 to 260,000 Indian and Southeast Asian slaves within this oceanic basin between 1580 and 1800. If we add another 62,000 to 74,000 possible seventeenth-century Dutch exports from India to Southeast Asia and the 100,000 to 150,000 slaves brought to Batavia by Asian traders during the seventeenth and eighteenth centuries to these figures, Europeans may have been involved, either directly or indirectly, in trafficking 360,000 to 484,000 Indian and Southeast Asian men, women, and children between 1500 and 1830. These estimates are obviously imprecise and will remain so until slave trading in the Indian Ocean becomes the subject of the kind of careful, sustained research that has allowed Philip Curtin’s classic census of the transatlantic trades to be revised, refined, and elaborated upon over the last four decades. Although problematic, these estimates and the small but compelling body of scholarship upon which they rest demonstrate that this European commerce in Asian slaves was not insignificant, a point illustrated by Markus Vink’s comparison of the seventeenth-century Dutch Indian Ocean trade with other trades during the same period. Vink notes that this Dutch trade was only slightly smaller in volume than the trans-Saharan slave trade, was one and a half to three times the size of the Swahili, Red Sea, and Dutch West India Company trades and equaled 15 to 30 percent of the transatlantic trade.210 His assessment, coupled with what we now know about European slave trading in the region during the eighteenth century, highlights the role that this traffic played in the development of the migrant labor trades that emerged within and ultimately beyond the Indian Ocean as this oceanic world became increasingly integrated into the global capitalist economy that emerged during the seventeenth, eighteenth, and nineteenth centuries. Our current lack of knowledge about many features of these Asian trades necessarily limits our ability to assess their impact on indigenous states and societies in anything other than rather general terms. One consequence of the Dutch demand for chattel labor was the replacement of relatively “open” 210  Vink, “ ‘The World’s Oldest Trade,’ ” 168.

1456

Allen

indigenous systems of slavery, in which bondmen and -women could be assimilated into dominant populations, by more restrictive “closed” systems characterized by a greater emphasis on slaves’ racial distinctiveness and the perpetuation of slave status.211 The expansion of regional slave-trading networks is also closely associated with the European demand for chattel labor. Independent Portuguese traders are credited with increasing the volume of slave trading in the northeastern Indian Ocean during the seventeenth century, while the VOC’s demand for slaves encouraged the development of larger, more centralized slave markets at the expense of smaller, fragmented trading systems.212 The revival of the Arakanese state of Mrauk-U during the 1620s is attributed to the need for slaves and rice in the VOC’s Indonesian establishments, while the kingdom’s decline after circa 1664 coincided with the declining demand for these commodities following the company’s destruction of the Makassar Sultanate between 1666 and 1669 and its subsequent ability to procure these commodities from closer sources of supply.213 VOC requirements likewise affected slave trading networks within and beyond the Indonesian archipelago.214 The dramatic increase in slaving in the Moluccas during the seventeenth century has been traced to the Dutch and Spanish demand for chattel labor in urban areas under their control.215 Similar pressures not only reinforced the institution of slavery on Timor, but also encouraged Asian traders during the eighteenth century to carry increasing numbers of slaves from islands such as Flores and Sumba to Makassar, from whence they were reexported to Java.216 To assess these trades’ significance only in such terms, however, runs the risk of ignoring the larger contexts in which they occurred. Studies of Indian merchants note that the general orientation of the subcontinent’s transoceanic trade began to shift away from Southeast Asia and toward the western Indian Ocean during the second half of the eighteenth century as British commercial and political dominance in India increased and the subcontinent 211  Reid, “‘Closed’ and ‘Open’ Slave Systems,” 156, 175. 212  Arasaratnam, “Slave Trade in the Indian Ocean,” 197–99. 213  Charney, “Crisis and Reformation.” See also Sanjay Subrahmanyam, “Slaves and Tyrants: Dutch Tribulations in Seventeenth-Century Mrauk-U,” Journal of Early Modern History 1, no. 3 (1997): 201–53. 214  Raben, “Cities and the Slave Trade,” 133–35. 215  Leonard Y. Andaya, “Local Trade Networks in Maluku in the 16th, 17th, and 18th Centuries,” Cakalele 2, no. 2 (1991): 83. 216  Hägerdal, “The Slaves of Timor,” 24, 26; Knaap and Sutherland, Monsoon Traders, 116–19, 123, 143.

Carrying Away The Unfortunate From India And Southeast Asia

1457

became increasingly incorporated into the modern global economy.217 Evidence of this shift includes increased Indian commercial involvement in southeastern Africa during the latter part of the eighteenth century, activity that included trading in African slaves.218 What, if any, role the traffic in Indian slaves to the Mascarenes played in these developments remains to be ascertained, but the ties between the Mahamays in Goa and French slaving interests in the Mascarenes during the 1770s and 1780s219 and the involvement of Mozambican-based Gujarati merchants such as Assan Valy and Sobhachand Sowchand in shipping African slaves to Mauritius and Réunion after 1780 are tantalizingly suggestive. The exportation of tens of thousands of Indian slaves to the Mascarenes during the eighteenth century, especially after 1770, assumes even greater significance if this traffic is viewed as the institutional precursor of the “new system of slavery”220 that scattered some 2.2 million African, Chinese, Indian, Japanese, Javanese, Melanesian, and other non-European indentured workers across and beyond a colonial plantation world that reached from the Caribbean to East and South Africa, the Mascarenes, Southeast Asia, Australasia, and the 217  Ashin Das Gupta, “India and the Indian Ocean in the Eighteenth Century,” in India and the Indian Ocean, 1500–1800, ed. Ashin Das Gupta and M. N. Pearson (Calcutta: Oxford University Press, 1987), 131–61; Immanuel Wallerstein, “The Incorporation of the Indian Subcontinent into the Capitalist World-Economy,” in The Indian Ocean: Explorations in History, Commerce and Politics, ed. Satish Chandra (New Delhi: Sage Publications, 1987), 222–53; Prakash, European Commercial Enterprise, 268–314. 218  Pedro Machado, “A Forgotten Corner of the Indian Ocean: Gujarati Merchants, Portuguese India and the Mozambique Slave Trade,” in Campbell, The Structure of Slavery, 18–20, 22, 25. This involvement would intensify during the nineteenth century. See: Lawrence J. Sakarai, “Indian Merchants in East Africa. Part I: The Triangular Trade and the Slave Economy,” Slavery and Abolition 1, no. 3 (1980): 292–328; Lawrence J. Sakarai, “Indian Merchants in East Africa. Part II: British Imperialism and the Transformation of the Slave Economy,” Slavery and Abolition 2, no. 1 (1981): 2–30; William Gervase Clarence-Smith, “The Economics of the Indian Ocean and Red Sea Slave Trades in the 19th Century: An Overview,” Slavery and Abolition 9, no. 3 (1988): 11–12; Luis Frederico Dias Atunes, “The Trade Activities of the Banyans in Mozambique: Private Indian Dynamics in the Panel of the Portuguese State Economy (1686–1777),” in Mathew, Mariners, Merchants, 301–31; Michael N. Pearson, “Indians in East Africa: The Early Modern Period,” in Mukherjee and Subramanian, Politics and Trade, 227–49. 219  de Souza, “French Slave Trading,” 123–25. 220  Modern proponents of the argument that the indentured labor system was essentially a “new system of slavery” follow Hugh Tinker, A New System of Slavery: The Export of Indian Labour Overseas, 1830–1920 (London: Oxford University Press, 1974; 2nd ed., London: Hansib, 1993).

1458

Allen

Central and South Pacific between the mid-1830s and the 1930s. More than 1.5 million of these laborers came from India, 900,000 of whom labored in British and French colonies washed by the Indian Ocean. Mauritius has long been regarded as the crucial test case for the use of free contractual labor in the postemancipation plantation world,221 and more than 452,000 indentured Indians reached the island between 1834 and 1910, while another 75,600 landed on Réunion between the 1840s and 1880s.222 The regional and global significance of this system becomes even more evident if the 1,187,000 Indians recruited to work on short-term, usually verbal contracts in Malaya and the tens of thousands of Tamils from South India who labored on Ceylonese estates under similar terms are included in this postemancipation labor diaspora.223 The indentured labor system’s origins are usually traced generally to the desire of early nineteenth-century British abolitionists to end slavery by demonstrating the superiority of “free” over slave labor and, more specifically, to the rapidly expanding Mauritian and Réunionnais sugar industries’ need for agricultural labor at a time when slave amelioration policies and suppression of the illegal slave trade threatened the long-term viability of the islands’ workforce.224 Indentured Indian labor was an attractive option for a number of reasons: the subcontinent’s geographical proximity, its seemingly inexhaustible supply of inexpensive agricultural labor, and the existence of a government 221   I. M. Cumpston, Indians Overseas in British Territories, 1834–1854 (London: Oxford University Press, 1953), 85. 222   Kernial Singh Sandhu, Indians in Malaya: Some Aspects of Their Immigration and Settlement (1786–1957) (Cambridge: Cambridge University Press, 1969), 306–9; David Northrup, Indentured Labor in the Age of Imperialism, 1834–1922 (Cambridge: Cambridge University Press, 1995), 159–60. 223  Sandhu, Indians in Malaya, 306–9. The number of these laborers in nineteenth-century Ceylon is unknown, but the magnitude of this migration is suggested by the 1871 census’s enumeration of 107,284 plantation workers and 86,093 other Indians on the island. The 1881 census reported 220,576 Indians on the island, a figure that rose to 225,837 in 1891 (Patrick Peebles, The Plantation Tamils of Ceylon [London: Leicester University Press, 2001], 45–46, 50). See also Roland Wenzlhuemer, “Indian Labour Immigration and British Labour Policy in Nineteenth-Century Ceylon,” Modern Asian Studies 41, no. 3 (2007): 575–602. 224  Huguette Ly-Tio-Fane Pineo, La diaspora chinoise dans l’océan Indien occidental (Aixen-Provence: Association des Chercheurs de l’Océan Indien, Institut d’Histoire des Pays d’Outre-Mer, GRECO—Océan Indien, 1981), 262; Huguette Ly-Tio-Fane Pineo, Lured Away: The Life History of Indian Cane Workers in Mauritius (Moka, Mauritius: Mahatma Gandhi Institute Press, 1984), 14–17; Richard B. Allen, “Capital, Illegal Slaves, Indentured Labourers and the Creation of a Sugar Plantation Economy in Mauritius, 1810–60,” Journal of Imperial and Commonwealth History 36, no. 2 (2008): 153–54, 159–60.

Carrying Away The Unfortunate From India And Southeast Asia

1459

committed, at least in principle, to improving the quality of life for millions of impoverished men, women, and children. Mascarene colonists, moreover, had well-developed commercial ties with the subcontinent and substantial experience with Indian slaves. Some 3,000 to 3,200 free Indian laborers reached Réunion from Pondichéry, Karikal, and/or Yanam during the late 1820s,225 a period of renewed British concern about slave exports from the Coromandel and Malabar coasts.226 In addition to being major sources of slave exports, these enclaves were also closely associated with the recruitment of free Indian craftsmen and artisans to work in the Mascarenes during the eighteenth century. In 1729, 95 such workers arrived on the Île de Bourbon from Pondichéry, along with 150 Indian slaves.227 Free Indian artisans from Pondichéry landed on the Île de France as well in 1729,228 and continued to do so until at least the late 1790s.229 What we know about the background of some of the indentured Indians who reached Mauritius during the late 1830s likewise points to the European exportation of Indian slaves establishing the precedent for the indentured labor trade, if not some of the institutional foundations upon which it rested. The Assamese and Himalayan foothills supplied Bengali and northern Indian slave markets with tribal hill people during the 1810s, if not before, and reports in 1825 that famine had forced Assamese to sell themselves into slavery despite

225  Northrup argues that 3,000 such workers arrived from Pondichéry and Karikal (Indentured Labor, 60), while Jacques Weber puts the number at 3,211 from the area around Yanam (“L’émigration indienne à La Réunion: ‘Contraire à la morale’ ou ‘utile à l’humanité’? (1829–1860),” in Esclavage et abolitions dans l’océan Indien (1723–1860), ed. Edmond Maestri [Paris: L’Harmattan, 2002), 309). On Indian immigration to Réunion before 1848, see also Sudel Fuma, “Un conflit de civilisations: Immigrants indiens et société colonial de l’Ile Bourbon, 1828–1848,” in Le movement des idées dans l’océan Indien occidental: Actes de la table ronde de Saint-Denis, 23–28 juin 1982 (St. Denis, Réunion: Nouvelle Imprimerie Dionysienne, 1985), 331–62. 226  IOR: F/4/1034/28499, Slavery, Kidnapping and Sale of Children in Tanjore and Tinnevelly [1825]; F/4/1035/28524, Slavery: Correspondence with the Resident at Mocha relative to the Traffic carried on by a subject of the Judda Government [June 1826]. 227  Gerbeau, “The Indians of the Mascarenes,” 37. 228  Baker, “On the Origins of the First Mauritians,” 193. 229  MNA: A 32, fols. 82, 87, 89–91, Liste des passagers arrivés, 1794–1801. The number of such arrivals during the eighteenth century is unknown, but 190 free Indian workers are known to have reached the Île de France between 1729 and 1735. In 1740, 73 of 137 skilled workers in Port Louis’s port facilities were Indians (Auguste Toussaint, Port Louis, deux siècles d’histoire [Port Louis, Mauritius: La Typographie Moderne, 1936], 33, n. 1).

1460

Allen

governmental attempts to ameliorate their condition230 invariably raises the question of whether any of these men, women, and children ultimately reached labor markets elsewhere in eastern India. The fact that early Indian immigrants to Mauritius included large numbers of dhangars, or tribal hill people, highlights the need to raise this question, all the more so since dhangars accounted for approximately one-third of the 7,000 indentured Indians who landed on the island during 1837–38.231 While many of these men came from the Chota Nagpur plateau in southern Bihar, the presence of hill tribesmen from Assam or elsewhere among these early indentured immigrants cannot be discounted pending further inspection of the archival record. Structural links between the Indian slave and indentured labor trades are suggested in other ways as well. The Madras Presidency’s 1819 inquiry revealed that individuals of “slave” caste status could be sold at will in some parts of southern India, a finding that, given the ease with which labor circulated in the subcontinent,232 suggests that such individuals were among the tens of thousands of slaves exported from India during the seventeenth and eighteenth centuries. More than forty-five years ago, Benedicte Hjejle argued that the recruitment of some indentured Indians during the nineteenth century cannot be understood without reference to indigenous slave systems and that a significant number of the Indian laborers who reached Ceylon between 1843 and 1873 came from among South India’s praedial slave population.233 Mauritian 230  IOR: F/4/1115/29887, pp. 3–4, D. Scott to Mr. Chief Secretary Swinson, 4 March 1825, and pp. 7–8, Mr. Acting Secretary Stirling to D. Scott, 3 April 1825. 231  Marina Carter, Servants, Sirdars and Settlers: Indians in Mauritius, 1834–1874 (Delhi: Oxford University Press, 1995), 104. More than 50,000 Indian tribal peoples, mostly from the modern states of West Bengal, Bihar, and Uttar Pradesh, reportedly reached Mauritius between 1834 and 1870 (Pooja Ramchurn-Jokhun, “Tribal Migration,” in Angajé: The Early Years. Explorations into the History, Society and Culture of Indentured Immigrants and their Descendants in Mauritius, ed. Vijayalakshmi Teelock, Anwar Janoo, Geoffrey Summers, Marc Serge Rivière, and Sooryakanti Nirsimloo-Gayan [Port Louis, Mauritius: Aapravasi Ghat Trust Fund, 2012], 13–14). 232   Ravi Ahuja, “‘Opening up the Country’? Patterns of Circulation and Politics of Communication in Early Colonial Orissa,” Studies in History, n.s., 20, no. 1 (2004): 73–130; Ian J. Kerr, “On the Move: Circulating Labor in Pre-Colonial, Colonial, and Post-Colonial India,” in Coolies, Capital, and Colonialism: Studies in Indian Labour History, ed. Rana P. Behal and Marcel van der Linden [International Review of Social History, Supplement 14] (Cambridge: Cambridge University Press, 2006), 85–109. 233  Benedicte Hjejle, “Slavery and Agricultural Bondage in South India in the Nineteenth Century,” Scandinavian Economic History Review 15, nos. 1–2 (1967): 106. Forty-one percent of Indian emigrants to Ceylon from 1921 through 1930 came from “depressed” castes (R. Jayaraman, “Indian Emigration to Ceylon: Some Aspects of the Historical and Social

Carrying Away The Unfortunate From India And Southeast Asia

1461

immigration registers confirm that indentured laborers who reached the colony from southern India during the late 1830s included individuals of “slave” caste status.234 That these men were, in one sense, simultaneously slave and free is a potent reminder that the winds of abolitionism that began to blow in the Indian Ocean world during the late eighteenth century and did so with increasing force during the early nineteenth century could be, and often were, complex and variable.

Background of the Emigrants,” Indian Economic and Social History Review 4, no. 4 [1967]: 338). See also S. Manickam, Slavery in the Tamil Country: A Historical Overview, rev. ed. (Madras: Christian Literature Society, 1993). 234  MGI: PE 1. The men in question were Palins and Pulayas.

“Closed” and “Open” Slave Systems in Pre-Colonial Southeast Asia Anthony Reid The distinction between two types of slave system—“open” and “closed”— was made by James Watson. He developed it to resolve the confusion between those anthropologists (primarily Africanists) who saw slavery as a process of incorporating “outsiders” into the kin group, and those (primarily Asianists) who argued that an inalienable kinlessness and outsider status was the key to slavery. The “open” system is more prevalent in Southeast Asia, as it is in Africa, as Watson rightly noted, largely because of the relative abundance of land and the primary importance placed on accumulating manpower (Watson 1980, pp. 9–13). In Southeast Asia, however, there are examples of both “closed” and “open” systems, and of others which appear to straddle the border between the two. While the distinction remains an extremely helpful one, the evidence of pre-colonial Southeast Asia enables us to see how both types fitted into a larger pattern. A “closed” system of slavery may be defined as one oriented primarily towards retaining the labour of slaves by reinforcing their distinctiveness from the dominant population. In Southeast Asia this pattern occurs typically in relatively static and self-contained communities practising labour-intensive wetrice agriculture, where commercial exchange and the money economy have made little impact. Although slaves were regarded as property in this context, there was little exchange in practice. An “open” system is one which is acquiring labour through the capture or purchase of slaves, and gradually assimilating them into the dominant group. The most recently acquired slaves are also the most clearly demarcated as such, and their constant recruitment makes it unnecessary to emphasize unduly the servile status of the second and third generation. In pre-colonial Southeast Asia the most striking cases of this pattern are the wealthy mercantile cities and a few other labour-deficient areas.

Source: Reid, Anthony, “ ‘Closed’ and ‘Open’ Slave Systems in Pre-Colonial Southeast Asia,” in Anthony Reid (ed.), Slavery, Bondage, and Dependency in Southeast Asia, St. Lucia: University of Queensland Press, 1983, 156–181.

© koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_047

“ Closed ” and “ Open ” Slave Systems

1463

Although flourishing in different economic conditions, these two forms and a number of transitional ones are part of a common Southeast Asian cultural pattern in which the routes into slavery are almost identical. Moreover the constant flow of slaves from the “closed” to the “open” systems makes it necessary to retain the insistence of all contemporary sources that we are dealing here with the one phenomenon, “slave”, even though the forms that phenomenon takes are naturally very varied. We will review here the evidence of the sixteenth and seventeenth centuries in particular, although in the case of many “closed” systems we must read back from the nineteenth century evidence which is all that is available. This was a period of rapid commercial expansion and urban growth, in which the transfer of labour was undoubtedly unusually marked. Two other features of sixteenth and seventeenth century Southeast Asia ensured that slavery would be unusually prominent: firstly the population was still very low, particularly in relation to the great wealth flowing from the spice trade. A recent estimate has suggested only about fifteen million people in the whole of Southeast Asia at that time, as against about a hundred million each for Europe and for China, and very much higher densities also in Japan and India (McEvedy and Jones 1978). Secondly there was a relatively dispersed and pluralistic political system, both within established states like Malacca and Banten, and in areas which were politically very fragmented, either temporarily (Java) or permanently (Philippines, Sulawesi, etc.). As a result of these two factors, control of men, not land or capital, was both the key and the index to power. The lack of legal and financial institutions made a powerful patron the most useful security for the poor, and bondsmen the most valuable asset for the rich. Because bondsmen were the measure of wealth and power, there was a marked reluctance to waste manpower through killing, either judicial or military (Reid 1980, pp. 242–45). Warfare was waged to obtain manpower, not to lose it, and prisoners were made bondsmen. Judicial execution was rare, and prisons were unknown, though criminals or debtors might be put in the stocks or tied to trees for short periods. This phenomenon was so marked a feature of most parts of Southeast Asia that most outside observers were struck by it, in both what I have called the “closed” and the “open” types of system. Some representative quotes may illustrate the point: Aceh: “As in the rest of Below the Winds [Southeast Asia], the natives reckon high rank and wealth by the quantity of slaves a person owns” (O’Kane 1972, p. 177).

1464

Reid

Banten: “The one who has most [men] is held to be richer and more powerful” (Eerste Schipvaart 1598, I, p. 117). “Their wealth lyeth altogether in Slaves, so that if their Slaves be kild, they are beggared” (Scott 1606, p. 142). Johor: “Each [orangkaya] feared to lose his slaves, which are their only wealth” (Matelief 1608, p. 17). Malacca: “The greater part of their riches are slaves, some men even having 600 and 700 slaves” (Florentine letter 1513, p. 371). Philippines: “The greater part of the wealth of these Indios consists in slaves” (Colin 1660, I, p. 75). “It was the easiest thing in the world for the powerful to multiply their slaves. They therefore possessed them in very large numbers, for among them slaves represent the greatest form of wealth” (Chirino 1604, p. 364; see also quotations below). Batakland: “Almost all the chiefs are poor and needy. The slaves make up their greatest wealth” (Francis 1839, p. 215). Throughout the region there is remarkable similarity about the means by which men became slaves. Two representative sixteenth century sources on the pre-Islamic Archipelago, one internal and the other external, will serve for many. The Bugis law code Latoa recorded: “A person is called a slave (ata) when the four following circumstances arise: first, a person is peddled for sale and bought; second, the person sold says ‘buy me’ and you buy him; third, a person is seized in war and sold; fourth, a person has transgressed the customary law (ade) or the state (kerajaan), he is sold and you buy him”, or in a fifth case, a man can be sold to make good his inability to pay his or his parents’ debts (Mattulada 1976, pp. 278–81). A Spanish inquiry in the Philippines noted: Some are slaves because their fathers and grandfathers were such; others sold themselves … either to make use of the money or to pay their debts; others were captured in war; others became slaves because, being orphans, they were held in that condition for food and expenses; others were sold in times of famine by their fathers, mothers, or brothers; others bear that name because of loans, for interest multiplies rapidly among the Indios and the Moros, and thus a poor man becomes a slave. There are men who become slaves on account of crimes, and failure to pay fines and penalties. Davalos 1584, p. 61; see also Lavezaris 1574, pp. 286–88; Morga 1609, pp. 272–74; Chirino 1604, pp. 363–64

“ Closed ” and “ Open ” Slave Systems

1465

Systematizing these and other sources, we may distinguish four major types of path into bondage: 1. 2. 3. 4.

inheritance or sale capture in war judicial punishment (in lieu of fines) failure to meet debts.

The relative importance of these four routes to slavery differed under different conditions. Inheritance and judicial punishment were the most common sources of slavery in the “closed” systems where a money economy was little developed, while sale and indebtedness were more important in the cities and other areas open to the money economy. Yet it is important to note that all societies recognized the four routes as valid. The prominence of debt and judicial sentence as sources of slavery appears to be distinctive in the Southeast Asian pattern (Nieboer 1910, p. 39; Wilken 1888, p. 608). Since inability to pay a judicial fine may be taken as a variation on inability to pay a debt, we may safely regard debt as the most fundamental source of Southeast Asian slavery. It was reported very widely, from the Batak and Pasumah of Sumatra (Wilken 1888, p. 563–64) to seventeenth century Banten (Jourdain 1615, in East India Company LR, III, p. 274; Pigeaud 1931, p. 44) and sixteenth century Ternate (Jacobs 1971, pp. 127–29), that a creditor could seize a defaulting debtor or some member of his family until the debtor paid. As an old Batak saying put it, “Whoever has a debt pays; if he can make no payment it is paid with the head”—an almost exact equivalent of the dictum of early Roman law, “Qui non habet in aere, luat in corpore” (Wilken 1888, p. 568). The labour of this captive in no sense contributed to diminishing the original debt, however. If the debt continued to be unpaid, the debtor or his surrogate was gradually transformed into a permanent bondsman, and his children after him. One did not need to be poor to become a debt-bondsman, since interest rates were extremely high and commercial ventures uncertain. The poor were the most vulnerable, and the destitute had no alternative but to “sell themselves” to stay alive. Fortunately large-scale famine was rare in Southeast Asia, in comparison with India, China or Europe. However, almost the whole indigenous slave population of nineteenth century Sumbawa (as opposed to imported Endenese) was descended from those who had sold themselves during the famine that followed the severe volcanic eruption of 1814 (Ligtvoet 1876, p. 570; Bik 1928, p. 14). The Undang-undang Melaka (Laws of Malacca)

1466

Reid

took it for granted that in a time of shipwreck or disaster people would ask to be accepted as bondsmen in return for food (Liaw 1976, pp. 116–19). The earliest Spanish chroniclers found it quite extraordinary that: “if a man in some time of need, shelters a relative or a brother in his house, supports him, and provides him with food for a few days, he will consider that man as his slave from that time on” (Legazpi 1569, p. 54; see also Lavezaris 1574, p. 288). The ease with which these bonds were contracted was one of the things which most astonished foreign visitors to Southeast Asia. La Loubère records a saying that Siamese would sell themselves for the pleasure of eating a durian (La Loubère 1691, p. 77), and the Persian embassy similarly noted that “pawning men or women is one of the most common practices in Slam. For a certain sum the Siamese will pawn themselves or their sons or daughters … It is also common for men to pawn themselves and … work for the pawnee as if they were his slave” (O’Kane 1972, pp. 131–32). The Spanish sources on the Philippines are much more detailed, and again this was the phenomenon which most surprised them: “They are all usurers, lending money for interest, and go even to the point of making slaves of their debtors, which is the usual way of obtaining slaves … They are so mercenary that they even make slaves of their own brothers, through usury” (Sande 1576, p. 69; also Colin 1660, I, p. 75; Chirino 1604, p. 363). Much of the literature on slavery makes a distinction between debt-bondsmen and slaves proper, on the grounds that debt-bondsmen had a theoretical right to redeem themselves, or that they could not be sold outside the system. The colonial governments in the nineteenth century themselves emphasized such a distinction, since it enabled them to continue to tolerate a form of slavery after the word itself had become anathema (Schröder 1917, p. 201). However, the bulk of the pre-colonial Southeast Asian evidence makes no distinction. On the contrary, debt is the most common origin for the heritable slaves found in many societies. Moreover, several of the legal codes specifically mention that the debtor can be sold to make good the debt. The Bugis Latoa cited above is one such example, and the (probably sixteenth century) Kutai code is another: “When a free man has many debts here and there then that man is sold and his selling price is divided” (Adatrechtbundels 1937, XXXIX, p. 320). A foreign source says of Banten: “Their laws for debt are so strict that the creditor may take his debtour, his wives, children and slaves, and all that he hath, and sell them for his debt” (Scott 1606, p. 173). When the Undang-undang Laut of Malacca forbids members of a ship’s crew to buy bondsmen or bondswomen and bring them on board without the Captain’s permission, it uses the verb menebus (Winstedt and de Josselin de Jong 1956, p. 44), which Wilkinson translates as “buying out someone’s debt”.

“ Closed ” and “ Open ” Slave Systems

1467

The difference from outright purchase seems again insignificant, since it clearly involved carrying the bondsman away from his native place. The Undangundang Melaka forbade bondsmen to be resold if they had become such through starvation or shipwreck, even if they had agreed in extremis that they could be resold (Liaw 1976, pp. 116–19). That there was a need for this humane provision itself confirms that other types of debt-bondsmen were sold. The everyday terminology of Acehnese, Batak, Malay, Bugis and others distinguished types of bondsmen in terms of their function, but not in terms of whether they had come into that situation through debt or through some other means (especially Brown 1970, p. 19; Mattulada 1976, pp. 31–36, 278–85). Once again, therefore, we must side with Nieboer in accepting “pawns or debtor-slaves” as functionally part of the slave category, even if they may in theory have been temporary members of it (Nieboer 1910, pp. 39–40). Only in Siam and Java do there appear to be legal provisions which distinguish between bondsmen on these grounds, perhaps because the court in both cases had an interest in limiting the numbers of corvéeable subjects lost to private slavery. The old-Javanese code, Kûtâramânawa, declares that a debt-bondsman becomes a permanent slave after two years if the debt is in buffaloes or copperwork, or after five years if it is in slaves or cloth (Wilken 1888, pp. 571–72). Subsequently, as we shall see below, Java was one of the first societies after Islamization to take the view that “slavery”, in contrast to debtbondage, was unacceptable. Thai law similarly insisted that debt-bondsmen had a right to regain their freedom if they paid the debt, though the children born to them in bondage nevertheless remained bonded. The importance of accumulating bondsmen and the role of debt in acquiring them may, then, be taken as general features of Southeast Asian slavery in the pre-colonial period. If we turn to the practical conditions under which they were held, however, we are faced with a bewildering variety of forms. The distinction between “closed” and “open” systems, with which we began, will help us to understand at least the most strikingly different polar types. Of the “closed” slave systems we may take Nias, Toba Batak, Toraja (in the broadest sense) and to some extent Dayak and Kenyah as representative, although for all these we must work with nineteenth century evidence. Each of these societies was essentially stateless, yet governed by very clear ritual divisions between aristocrat, free and slave categories. Access to the labour of slaves was an essential attribute of the dominant aristocratic class. Certain types of labour, such as carrying water, cutting wood, handling the dead (both animals and humans) were typically seen as slave’s work. For the Sa’dan Toraja we have a particularly elaborate division of functions, with different groups of slaves (kaunan) having specified duties. The most fundamental slave role,

1468

Reid

however, was indicated by the traditional label, “Those who are used like buffalo on the sawah” (Nooy-Palm 1979, pp. 86–91). Estimates of the proportion of slaves in the population have been as high as thirty per cent for the Batak and Sa’dan Toraja (Castles 1972, p. 20; NooyPalm 1979, p. 51), to about fifteen per cent for Nias and ten per cent for the Kayan, Kenyah and Melanau of Sarawak (Marschall, personal communication; Rousseau 1980, p. 54; Watson 1980, p. 307). In all cases this probably involves a broad definition of slavery embracing several sub-groups. For the Melanau, Nias and Toba Batak the primary distinction appears to have been between field slaves, who had their own house and some freedom to work for their own livelihood although they were obliged to labour on demand for the master; and house slaves, who had less freedom but a close and even intimate relationship with their masters (Schröder 1917, p. 350; Watson 1980, p. 305; Loeb 1935, pp. 41–42). In Sulawesi, on the other hand, a primary distinction was made between inherited slaves belonging to the whole lineage and slaves acquired in other ways who could more easily be disposed of. The inherited slaves of the Mandar and the Sa’dan Toraja could neither be redeemed nor sold, because their existence was too important to the dominant lineage. Their position was typically like that of household slaves, in an intimate relationship to their master (Friedericy 1933, pp. 106–107; Nooy-Palm 1979, p. 45; Bigalke in this volume). In Poso, on the other hand, Kruyt gives the impression that inherited slaves could be killed or sold, but only with the permission of all family members having rights over them (Kruyt 1911, p. 64). The inherited slave enjoyed the security of knowing that his two most important and expensive ritual needs—marriage and funeral—would be met by his master, though of course on a scale lower than for a freeman. A slave therefore had greater certainty of being married than a poor non-slave (Schröder 1917, p. 351). On the other hand his children might be taken from him at an early age, even six or seven years, to serve as members of the master’s household. In Poso this process was ritualized in the giving of a piece of cloth to the newly delivered mother, as a sign that this child would belong to the giver rather than to any other member of the lineage (Kruyt 1911, pp. 72– 73, 91). The young slave would be socialized in the master’s household into a relationship which was permanently that of a child, intimate but inferior. After marriage the slave occupied his own household and the relationship became less intimate, but at best still warm and dutiful. In several languages, including those in Nias and Sumba, the word for slave was the same as that for child. Nevertheless there seems to have been a theoretical right, and sometimes duty, of the master to kill his slave. The Toraja in general believed a great chief

“ Closed ” and “ Open ” Slave Systems

1469

should be accompanied to the afterlife by slaves to serve him there (Kruyt 1911, p. 64; Wilcox 1949, p. 330). The kin group might also be required to give a life in retribution for a crime committed by one of its members, and in all these societies a slave seems frequently to have been substituted for the guilty party (Kruyt 1911, p. 64; Schröder 1917, pp. 351–52; Freeman 1970, pp. 304–306; Warren 1981, p. 199). Naturally no trusted inherited slave would be used for this purpose, unless he had made himself particularly objectionable, but rather a recently bought or captured slave. The central feature of our “closed” systems of slavery was the ritualized distinctiveness of the slave. The Sa’dan Toraja believed slaves were already a distinct class in the Upperworld before mankind existed on earth (Nooy-Palm 1979, p. 146). In Toraja only slaves ate dogmeat; slaves were forbidden to wear gold or bronze ornaments, to have carving on their houses, or to eat from the same dish as their masters (ibid. pp. 45–47). Marriage, or sexual activity, between a slave man and a free woman was regarded as especially heinous, and the punishment for it was death for both parties—usually through some spectacular form like burning or drowning (ibid. p. 47; Kruyt 1911, p. 70; Schröder 1917, p. 351). These societies prior to the nineteenth century were little penetrated by a money economy, so that there was very little marketing of slaves, and both the supply of and demand for new slaves was extremely limited. The aim of the system was a static one—to keep the slave class in its dependent position over many generations, preventing assimilation despite conditions of considerable intimacy. Although the sale of slaves in “closed” conditions was rare, there appears to have been no institutional obstacle to selling on a massive scale once a strong demand for slaves made itself felt from outside along with the spread of a money economy. We then find extraordinarily high figures for slave exports, like the thousands from Tana Toraja around 1900, from Nias around 1800, or from Bali and South Sulawesi in the eighteenth century (see above, pp. 32–33). Wherever commerce and the use of money spread widely throughout a society without giving rise to any strong state, we may expect to find much more movement in and out of the slave category, with a greater stress on slaves as property which could be exchanged for other goods or cash. This appears to have been the situation in the sixteenth century Philippines, described for us in considerable detail by a number of Spanish writers. We are inclined to describe this society as “transitional” between the “closed” and “open” systems. There was a clear distinction between the bonded class of alipin (Tagalog) or oripun (Visayas) (always translated esclavos by the Spanish) and an intermediate class of “freemen” (or “commoners” or “knights”) known

1470

Reid

as maharlika in Tagalog and timagua in Visayan. These maharlika had only limited obligations to their chief (datu), to help him in war, on voyages, in building his house and on other specified occasions. The means by which one could become an alipin are frequently described in the terms we have used above—inheritance, sale, capture, judicial punishment and debt. However, we read of no distinctions of status or treatment being made between bondsmen in terms of how they came to be such. The important distinction is that which we have already noted between household slaves (alipin sa gigilid) at the beck and call of their master, and field slaves or serfs (alipin namamahay). The latter were settled in their own house and worked a plot of land for their master, sharing the harvest equally with him. They had to assist their master in any major undertaking and “to come to the master’s house whenever he sends for them” (Morga 1609, p. 273; also Plasencia 1589, p. 175; Phelan 1959, pp. 20–21). De Morga, a relatively careful and dispassionate observer, has no doubt that both types of alipin could be and were sold; he gives the price for both, a sa gigilid being worth twice a namamahay: “These slaves constitute the main capital and wealth of the natives of these islands, since they are both very useful and necessary for the working of their farms. Thus they are sold, exchanged and traded, just like any other article of merchandise” (Morga 1609, p. 274; see also Lavezaris 1574, p. 288 and Legazpi 1569, p. 55, for similar views). The good Franciscan Juan de Plasencia, who is primarily concerned to curb what he sees as the terrible abuse of trading in slaves, takes a different view. He insists that alipin namamahay could not be sold: “Their children inherit … and enjoy their property and lands. The children … cannot be made slaves (saguigilir) [sa gigilid] nor can either parents or children be sold. If they should fall by inheritance into the hands of a son of their master who was going to dwell in another village, they could not be taken from their own village and carried with him” (Plasencia 1589, p. 175). Alipin sa gigilid, he conceded, could be sold, though in practice those “born in the house of their masters are rarely, if ever, sold. That is the lot of captives in war, and of those brought up in the harvest fields” (ibid. p. 176). This makes good sense as a description of a relatively “closed” system in stable times. However, Plasencia complained that in practice Filipino masters ignored the distinction between the two types of alipin, and “have adopted the custom of taking away the children of aliping namamahay [alipin namamahay], making use of them as they would of the aliping sa guiguilir, as servants in their households, which is illegal … In this way he becomes a sa guiguilir, and is even sold” (ibid. p. 177).

“ Closed ” and “ Open ” Slave Systems

1471

A master and slave (1.) of Bada, Central Celebes, in 1902. (From Sarasin, Celebes)

The difference between de Morga and Plasencia is thus in part the difference between what was and what ought to have been. In practice there was great fluidity, and bondsmen of various sorts were sold in response to a stronger commercial demand, just as occurred later in Bali, Toraja and Nias. If there was a customary distinction between saleable and non-saleable bondsmen, it had little real force either in ideology or legal institutions. When we move to the flourishing city-states of Southeast Asia in the sixteenth and seventeenth centuries, the property aspect of slavery becomes still more marked. Cities such as Malacca, Aceh, Banten, Ayudhya and Makassar grew into large international entrepots of the spice trade and the China trade, with cosmopolitan populations (and incidentally relatively full descriptions by visiting traders). Unlike the “closed” or “transitional” systems described above, we also find here valuable sources in the written law codes, essential for regulating the property relations between so many foreign and local traders. In all these codes, slaves feature as the most important single form of property, one which had to be regulated with some care. These cities had a notably pluralistic character in the period of their most rapid growth (Reid 1979; 1980, pp. 247–48). Since there was considerable international mobility on the part of large merchants, they were attracted only

1472

Reid

to cities which offered them a modicum of security as well as the opportunity to trade. Such cities were frequently far from orderly, however, and the surest security for the rich and powerful was their armed following, including all the slaves they could afford to acquire. In sixteenth century Aceh, for example, the rich orangkaya (merchant-aristocrats) “had beautiful, large, solid houses, with cannons at their doors, and a large number of slaves, both as guards and servants” (Beaulieu 1666, p. 111). Slaves were important not only for protection but for display, since the great men seldom stirred outside their urban compounds without a large retinue, and also as a labour force for cargo-handling, retailing, shipping, building and small-scale manufacture. The possession of slaves was considered very important in freeing the master from manual labour and thus marking him as a substantial citizen. In Malacca, “You will not find a native Malay who will carry on his back his own or any man’s property, however much you may offer him for doing so” (De Barros, Decada II, 6, i, cited Crawfurd 1856, p. 404). In Aceh, “the poorer sort, not having a slave of their own, will yet hire one to carry a mess [mas] worth of rice for them,… scorning to do it themselves” (Dampier 1699, p. 94). In Makassar, to say nya ata (he has bondsmen) was to designate a member of the gentry, who “believe it a piece of indecency … to till the ground, or follow any Mechanick Art” (Gervaise 1700, p. 94; see also Chabot 1950, p. 117). When Admiral Speelman, the destroyer of Makassar, was arguing optimistically that the place could still prosper, he prophesied that if all went well even the “ordinary people” would be able to save a little each year “which will eventually serve to buy a male or female slave in order to be free from labour themselves” (Speelman 1670, p. 67). The same outside observers who remarked on the ubiquity of slaves in the maritime city-states were equally surprised at the leniency with which they were usually treated. “There is nothing of rigour used by the Master to his Slave, except it be to the very meanest” (Dampier 1699, p. 98; see also Bowring 1856, p. 193; Legazpi 1569, pp. 54, 56; La Loubère 1691, p. 77). Many slaves lived in the large urban compounds of the orangkaya, but others had their own houses in the suburbs. They ate often with their masters, dressed like the majority of the population, and were permitted in many cases to bear arms and go to war (which Romans would not have approved at all). Many were free to conduct trade on their own as well as their master’s behalf, and thus to improve their lot. A group of Portuguese slaves in Aceh, for example, had done well for themselves as expert bleeders of the sick, and thus had “obtained the means to live well … and beyond that to pay other men to perform the task of labour ordered by the King” (Beaulieu 1666, p. 61). In the same way, we have clear evidence

“ Closed ” and “ Open ” Slave Systems

1473

that successful slaves were able to purchase slaves of their own to perform heavy labour, in Dutch Batavia (see below) as well as the Islamic city-states (Dampier 1699, p. 98; Warren 1981, p. 30). Since a common feature of these “open” slave systems was the expansion and social mobility of flourishing cities, slavery functioned largely as a means for incorporating labour into the system. The most menial work must have tended to fall to the lot of the latest generation of slaves imported into the city, whereas subsequent generations had various options to assimilate. Intermarriage appears to have been relatively common, for we have numerous regulations about the status of the children of such unions between slave and non-slave, even when the latter was the female (Liaw 1976, pp. 92–93; Morga 1609, p. 273; Adatrechtbundels 1939, XXXI, p. 183). The slaves were quickly incorporated into the dominant Islam of the cities. Many were only too ready to identify with the culture of the more sophisticated and powerful city, especially women whose lot tended to be much easier, even as slaves, than in the animist “closed” systems (Low, in Lasker 1950, pp. 290–91; Bigalke, in this volume). Some slaves were formally freed in return for services or at the death of their masters. A Makassar code provides for cases where the heirs of the master dispute the validity of testaments promising such manumission (Adatrechtbundels 1919, XVII, pp. 169–70; see also Friedericy 1933, pp. 107–108). It was probably much more common, however, for slavery gradually to transform itself into a looser sort of dependency with the turnover of generations. In none of these trading cities of the sixteenth and seventeenth centuries can we identify a class of independent urban artisans or labourers, free to work for wages or not to work at all. If we compare the wage rates given in European sources with the cost of a day’s rice, we find a very high labour cost, equivalent to about thirty times the rice requirement in Ayudhya and fifteen times that in Banten (see Smith 1974, p. 316; Van Vliet 1640, p. 96; Eerste Schipvaart 1598, pp. 118, 129). This was not a free market wage paid to the worker, but the cost of hiring bondsmen from a master: “It is their custom to rent slaves. They pay the slave a sum of money, which he gives to his master, and then they use the slave that day for whatever work they wish” (O’Kane 1972, pp. 177–78; see also Gervaise 1700, p. 82). The Undangundang Melaka similarly gives many examples of the legal implications of “hiring” (mengupah) or “borrowing” (meminjam) slaves, but none of any other type of labour contract (Liaw 1976, pp. 88–93, 162–63). Labour, then, was performed on the basis of obligation. Because there was really no “free” wage labourer category to escape to, slaves who were manumitted reportedly sold themselves again at once (Bowring 1856, p. 193, on Siam). Everybody appeared to be in some kind of bondage relationship: “neither can

1474

Reid

a Stranger easily know who is a Slave and who not amongst them; for they are all, in a manner, Slaves to one another” (Dampier 1699, p. 98, on Aceh). The means to escape from manual labour was not manumission but acquiring the wherewithal to buy a slave of one’s own. As the bondsman’s economic position rose he would typically not escape his obligation so much as transform it into a more agreeable form, perhaps in money, perhaps in simple loyalty. The introduction into these city-states of written legal codes, on Islamic or Indian models, gave to slaves a theoretically precise legal value, usually less than half that of a freeman. Even the smaller fines which had to be paid for killing, injuring or raping a slave were seen as compensation to the owner for damage to his property, rather than to the slave (Liaw 1976, pp. 74–77; Scott 1606, p. 94; Adatrechtbundels 1919, XVII, pp. 151–60; Adatrechtbundels 1937, XXXIX, pp. 302–304). Despite these theoretical distinctions, however, the horizontal cleavage between slave and free seemed in practice an indistinct one, and certainly less vital a means of self-identification than the vertical bonds which joined each bondsman to his master. Despite the apparent numerical dominance of bondsmen (to take the broader term) in these commercial cities, we know of only one slave revolt, in Patani in 1613. In this case, as in most of the classic slave revolts, ethnic difference had a role to play, since the slaves were overwhelmingly Javanese and the masters Malay (Floris 1616, pp. 94–95). Other reported slave revolts in the literature appear in reality to be rebellions by one whole people (typically hill people) against tribute or servitude to another (Kruyt 1911, p. 75; Turton 1980, p. 286; information from B. A. Batson). If many of the definitions of property rights in slaves came from the law of Islam, the more fundamental influence of Islamization on Southeast Asian slavery was probably the creation of the dividing line between Muslim “insiders” and non-Muslim “outsiders”—a line which tended to override that between slave and free. The shari’a is animated by a more emphatic definition of slaves as property than appears to have been traditional in Southeast Asia, but it also forbids the enslavement of fellow Muslims. The effect of this in the Southeast Asian Islamic city-states appears to have been that slaves in the stricter Islamic sense of abdi (Arabic ’abd) were identified primarily with firstgeneration imported slaves, typically non-Islamic, at least in origin. The most interesting evidence on the impact of Islam comes from the Bugis area of South Sulawesi. One Wajo chronicle provides some detail on the teaching of the apostle Sulaemana, who introduced the new faith to Wajo from Makassar about 1610. Certain former practices beloved by the Bugis— pork, palm-wine, adultery, interest—were absolutely forbidden to Muslims, said Sulaemana. In addition, for the truly righteous, “God would reward those who freed their slaves” when the slaves had followed then master into Islam.

“ Closed ” and “ Open ” Slave Systems

1475

This, however, applied only to “inherited slaves, and not to debt-bondsmen” (Noorduyn 1955, p. 101). From the very beginnings of Islam, therefore, debtbondage must have presented itself as an important loop-hole to avoid too fundamental a disruption of the labour system. The question was undoubtedly of great importance. A major war was fought in South Sulawesi in 1643 with a similar mixture of motives to the United States Civil War, though a different result. The ruler of the Bugis state of Bone took a very literal view of the demands of the new religion, and insisted that all slaves who were not born in that state should immediately be released and paid wages. This demand was unacceptable to many Bone notables, including the ruler’s mother, who fled to Goa (Makassar) to seek redress. Goa asked the ruler of Bone whether he was adopting this purist position “by command of the Prophet or on the basis of traditional custom, or according to his own will and pleasure?” When the raja left this letter unanswered, Goa declared war on him, resulting in the first complete subjection of Bugis Bone to Makassarese Goa, and presumably also in the retention of a relatively traditional bondage system in South Sulawesi (Noorduyn 1955, pp. 116–17; Speelman 1670, I, p. 53). Islam nevertheless became the primary boundary marker between the semi-assimilated second-generation bondsman and the internationally marketed slave. The supply of new slaves for the big maritime cities now had to be provided chiefly by non-Muslims. Some non-Muslim societies continued to sell their slaves voluntarily to outsiders—Nias, Batakland, Toraja, Bali and the still-Hindu ports of Java (Sunda before 1520 and Balambangan before 1600). In addition, petty warfare, which at best could be interpreted as genuine jihad to extend the faith and at worst as ruthless raiding for helpless human prey, continued on all the boundaries of Islamic communities. Almost every nonMuslim people within reach, from Malayan aborigines and Vietnamese in the west to Papuans and Timorese in the east, were subject to such raids. Speelman’s description of the trade of Makassar in the 1660s demonstrates how the slave trade operated in the eastern part of the Archipelago. Slaves were listed among the principal imports to Makassar from Manggarai, Timor, Tanimbar, Alor, Buton, Mindanao, Tawi-Tawi, Sulu and Brunei. These, together with Toraja slaves from the Sulawesi interior, were then re-exported to the major trading cities—Batavia, Palembang, Jambi, Aceh, Sukadena and Banjarmasin are mentioned as destinations. In the case of Banjarmasin, Speelman specified that “male and female slaves fitted for labour in the pepper gardens” were exported (Noorduyn 1983, p. 112). As an indication of how much further the Southeast Asian slave trade extended, Makassar ships were known to take cargoes of Toraja slaves as far as Ayudhya (Gervaise 1700, pp. 81–82); in the other direction a slave born in Siam became well known to the Spanish in

1476

Reid

the course of serving successive Brunei masters, and eventually followed one of these masters to Manila (Sisian 1579, pp. 179–200). Some of the slave suppliers listed by Speelman, notably Sulu, Brunei and Buton, exported little else but slaves. As was reported of Buton in 1613, “this is all their trade, to take slaves and sell them, for those of Boutoune are thought to be the best warriors” (Jourdain 1617, p. 292; compare Dampier 1697, p. 307). Sulu, the later role of which has been well described by James Warren (1981), was already notorious as a slaving centre in 1578; “The Joloans [of Sulu], as is well known, are open pirates, whose only ambition is to steal, and to assault men in order to sell them elsewhere” (Sande 1578, p. 176). This was no more than the continuation, on a more organized and outward-directed basis, of a pattern of raiding for bondsmen which had been a feature of much of the Archipelago for centuries before Islam. At the height of their prosperity, cities like Malacca, Aceh, Banten and Makassar could afford to import slaves on a considerable scale. “These slaves are drawn from every nation and race”, as was said of Aceh (O’Kane 1972, p. 178). Captive women were used as concubines or presented to retainers as wives, and their children became part of the dominant Muslim society. Slavery was an important means whereby animistic interior peoples were absorbed into the dominant Islam of the city and coast. As long as the trading cities of Southeast Asia were economically expanding on a pluralistic political base, slavery and the buying and exchanging of slaves were marked features of their social systems. Although we hear more about the international trade in slaves than about the internal marketing system, there undoubtedly was buying and selling. As was reported of Banten: “If they want to sell one of these [slaves] she is taken from one house to another, and someone bids enough for her … and some … they do not sell, and they are their best slaves” (Eerste Schipvaart 1598, I, pp. 129–30). Slaves appear to have been involved in almost every type of work in these cities. They are most widely reported in domestic functions, in construction and other manual labour, and as traders in the market. As Dampier said of Aceh, the money-changers and sellers in the market were predominantly slaves, “and you scarce trade with any other” (Dampier 1699, p. 98). Craftsmen, writers, translators and others with special skills were especially prized as slaves, and were often able to earn a substantial livelihood for themselves in addition to their work for their master. Weaving appears frequently to have been the job of slave women, especially, we may assume, when it was done for commercial purposes (Eerste Schipvaart 1598, I, p. 129). Malay and Pegu slave seamen are reliably reported (Bouchon 1979, p. 135), and slaves were also fishermen, at least in Aceh and Banten (Dampier 1699, p. 98; Eerste Schipvaart 1598, I, p. 129), We

“ Closed ” and “ Open ” Slave Systems

1477

also have numerous reports of slaves doing agricultural work in the hinterland of the trading cities, and some specify bought slaves being put to work on the master’s land (Dampier 1699, p. 91; Lockyer 1711, p. 54; La Loubère 1691, p. 77, and the case of the Banjarmasin pepper gardens cited above). We need not doubt, however, that the further away the fields of these “slaves” were from the master’s house, the more they resembled in practice serfs or even tenants. We have no Southeast Asian examples of agricultural estates or latifundia worked by slaves and centrally managed over a significant period. Slaves as private property were the primary asset of the oligarchs or orangkaya of the maritime city-states, and thus the principal feature of a pluralistic political order. Private slaves were seen as non-citizens, not subject to the normal corvée or tribute to the king (Terwiel, in this volume; Friedericy 1933, p. 106). Strong rulers therefore sought to extend the category of such “royal slaves” (hamba raja), and reduce the role of private slaves. This category of royal slaves was usually critical in enabling a dynasty to rise above other lineages and form a significant state. We have striking examples in the orang laut of Malacca and the Bajau of Makassar, each of which had a very close and loyal relationship to the ruling house and became indispensable to it as naval forces and providers of sea produce. Although both these communities had some features of pariah outcastes, they are not true slaves as we have defined them above. Similarly, whole villages or ethnic groups became designated as hamba raja through conquest or some communal crime, without thereby becoming alienable as property. Most Southeast Asian trading kingdoms sanctioned the confiscation by the ruler of the dependants and bondsmen of citizens who died before having settled their property on an heir. Ambitious or absolutist rulers like Ekathotsarot and Prasat Thong in Ayudhya, Mahmud in Malacca, and Iskandar Muda in Aceh were accused of executing wealthy subjects deliberately so as to take possession of their bondsmen. The number of royal slaves could therefore become very large. The Portuguese rounded up fifteen hundred of them after their capture of Malacca and assigned them to public works as the Sultan had done (Alboquerque, III, p. 135). In Aceh, too, forts and other public buildings were built by royal slaves, who nevertheless had to find their own subsistence by working for half their time on their own account (Beaulieu 1666, p. 108). If these manual labourers were the lowest sort of hamba raja, there were others whose closeness to the king made them among the most powerful in the realm. Mansur Shah of Malacca reportedly raised one of his slaves, originally a non-Muslim from Palembang, to be his Laksamana or Admiral (Pires 1515, p. 249). The heroes and guardian spirits of Chiengmai are two royal slaves who reputedly saved the kingdom in about 1400 (Vijewardene 1980, pp. 146–47; and see Bowring 1856,

1478

Reid

pp. 189–91 for Siam). Legal penalties against killing or injuring the hamba raja were especially high (Liaw 1976, pp. 72–75; Abdullah 1838, pp. 124–25). The balance between royal and private slaves was an index of the extent of political pluralism in these Southeast Asian city-states. If rulers succeeded in virtually eliminating “private slavery” in favour of universal obligations to the throne, as seemed to be happening in seventeenth century Mataram, and temporarily also in Ayudhya and Aceh, then of course the parallels with slavery as we know it in other times and places disappear completely. A striking case of the influence of a strong ruling house, allied with Islam, on the system of slavery, is again provided by South Sulawesi. In the sixteenth century, this region was one of the best and cheapest sources of slaves for export (Pelras 1981, pp. 158–59). During the period of Makassar’s greatness (c.1600–60), however, we hear nothing of Makassar and Bugis slaves being sold outside, although non-Muslims from other regions were marketed in Makassar. One of the conditions imposed for the establishment of an English factory in Makassar was that the English must buy no slaves “of our religion”, Islam (Articles of Agreement). After the fall of Makassar to the Dutch, South Sulawesi was once again politically fragmented and became in the eighteenth century the major seller of slaves to the VOC (Pelras 1981, pp. 158–59; Sutherland in this volume). Another factor which greatly modified the system of slavery was the impact of Muslims and Christians becoming the major buyers of slaves in the Archipelago. Both religions tended to soften the ritualized demarcation between slave and free of the “closed” systems, replacing it by a religious boundary in the Islamic case and a racial boundary in the Christian case. But whereas the Islamic city-states tended to assimilate their second-generation slaves, the Europeans brought with them a pejorative concept of slavery which obliged them to maintain a great social distance between European master and Asian slave. The English factory in Banten appealed for African slaves because, not surprisingly, the slaves they had bought locally kept running away (Willoughby 1635, fol. 115V–116). Although the Dutch in particular eventually adapted to a more Southeast Asian domestic style, neither they nor the English seriously attempted to Christianize their slaves, most of whom eventually became Muslim (Kumar 1976, p. 21; Raffles 1835, I, p. 104). Although the Europeans quickly became the largest slave buyers of Southeast Asia, at least in the two most populous islands of the Archipelago, Java and Luzon, they appear also to have stimulated the replacement of slavery by other forms of bondage. In Java, which had provided most of the exportable slaves for Malacca, Patani and Brunei in the sixteenth century, slave export almost disappeared in the course of the seventeenth century. The emergence of a strong

“ Closed ” and “ Open ” Slave Systems

1479

Two men of Kageroa (Bada), Central Celebes, with their slave (left), in 1902. (From Sarasin, Celebes.)

Styles of dress of the people of Banda. (Dutch woodcut, 1602). The figures on the extreme left and extreme right are identified as female and male slaves.

Mataram monarchy, and the other general factors mentioned below, must have played a part in this change. However, if we can believe de Haan (1922, I, p. 452), there was also an element of deliberate Dutch policy. The purchase of Javanese slaves for Batavia was explicitly forbidden by the VOC, in the belief that otherwise the colonial capital would have been overwhelmed with Javanese slaves sold by their chiefs, bringing the great danger of a “slave revolt” by a single ethnic group. With the major urban demand for “slaves” thus removed, the Javanese bondage system was encouraged to define itself in other ways. The Spanish in the Philippines had nobler motives, but shared the same conviction that slavery was a demeaning category. Believing that the “slavery” they found in the Philippines was incompatible with the new Christian man they wished to create there, they attempted to suppress it. Instead, the very considerable labour requirements of the Spanish government and church were acquired by a system of tribute and corvée. It is reasonable to surmise that this produced enormous dislocation of the social and political system, rather than

1480

Reid

any lightening of the burdens of ordinary Filipinos (Phelan 1959, pp. 93–104). One of the first major revolts against Spanish rule in Luzon was inspired in part by the resentment of the former elite of the Manila region at the loss of their bondsmen (Vera 1589, pp. 95–111). There seems reason to conclude that the decline in Filipino productivity noted by several acute observers in the wake of the Spanish conquest (Morga 1609, pp. 295–96; Plasencia 1589, pp. 178–79) was caused by a crisis in the whole system of production. As Europeans became the wealthiest urban element in Southeast Asia and therefore the biggest slave buyers, the “open” system of slavery tended to give way to one marked by racial distinctiveness. Some assimilation continued to take place, but less into the dominant slave-owning group (even here, of course, it occurred through concubinage) than into a new ethnic identity as the urban freed slave—a large part of the original kaum Betawi. In other respects it would be a mistake to exaggerate the difference between the European slaveowning pattern in Southeast Asia and the older urban pattern out of which it grew. Some slaves continued to have a relatively intimate relationship with the master’s household, as nursemaids and domestic help of all kinds. As the most valuable form of property they continued to be used as pawns and security for loans on a large scale. As before, in addition to productive functions they served to display the high status of their owner. Perhaps the most strikingly “Southeast Asian” feature of the slavery practised by the Dutch in Batavia was that successful slaves could still own their own houses, shops and slaves (de Haan 1922, I, pp. 454–57; Iwao 1970, pp. 5–9). It is very clear that slavery was an important element in the production system, in social stratification and in population movement in Southeast Asia over many centuries. The forms it took in different conditions were very varied, however. The ubiquity of slaves in Southeast Asian cities of the sixteenth and seventeenth centuries, which so much struck European and other observers, was probably a relatively short-lived phenomenon. When the Europeans first observed it, the system was already much altered, by rapid commercialization and urban expansion as well as by Islam, from the “closed” systems which may once have been much more widespread. The Europeans themselves began to alter the system almost as soon as they began to observe it. There were other reasons besides Islam and Christianity, however, which brought about a steady replacement of slavery by other forms of bondage. Among the most important was the impoverishment of the indigenous trading cities from the seventeenth century. Slaves in large numbers could only be maintained within the large urban compounds of the orangkaya as long as this dominant class was drawing very large profits from trade. The buying of newly

“ Closed ” and “ Open ” Slave Systems

1481

enslaved “outsiders” was a particularly expensive way of making up a labour force when cheaper forms of bondage were available. As the seventeenth and eighteenth centuries progressed, these internationally marketed slaves went increasingly to Europeans and Chinese alone. There were cheaper forms of bondsmen available in the cities even to Europeans, once they began to understand the societies of the hinterland. A German in Semarang in 1772 reported that he acquired a debt-bondsman as his manservant for only 25 Rijksdaalders, whereas a slave proper would have cost him 150. Functionally there was little difference (Dutch and Javanese appear to have agreed that debt-bondsmen should not be sold overseas), but by this stage it was a case of preferring the local system to what had come to be seen as a foreign one. As the indigenous cities declined in size and wealth and their export trade was diverted in colonial channels, there was a marked shift towards rice agriculture and economic self-sufficiency (Reid 1975, pp. 52–55; 1979, pp. 40–47). Bondsmen were put on the land, where their position within a generation or two resembled that of serfs, unlikely to be alienated without the land. The same process had occurred during the decline of the Roman Empire, as a consequence of a similar urban impoverishment (Bloch 1975, pp. 3–7). The swing back towards agricultural self-sufficiency was accompanied by a reassertion of the ruler’s dominance over the social and economic life of each state or community, and probably by a gradual rise in population. In these circumstances the rulers’ attempts to curb “private” slavery and render everybody liable to corvée were largely successful. The most densely populated areas began, moreover, to experience a marked decline in the value of manpower as against land, official position, or other forms of capital. Crawfurd (1820, I, p. 248) was partly right in suggesting, as reasons why Javanese alone appeared not to bother to take captives in war, nor to buy and sell slaves, that “the numbers and servility of the population of Java have, among them, rendered slaves of little value”. Two features of Nieboer’s definition of slavery were therefore eroded, if not eliminated, in the sevententh and eighteenth centuries. The decline of indigenous commerce and urban wealth diminished the property and exchangeability aspect of slaves, while the role of the major world religions, combined with the rulers’ desire to see their subjects as uniformly liable to corvée, weakened the legally and ritually subordinate status of slaves. The third aspect of our definition, compulsory labour, continued to be a marked feature of Southeast Asian societies, as did the tight vertical bond between those who performed it and their chiefs, masters, creditors or patrons.

1482

Reid

References Abdullah, Munshi. 1838. Kisah Pelayaran Abdullah, ed. Kassim Ahmad. Kuala Lumpur: Oxford University Press, 1962. Adatrechtbundels. 45 vols. ’s-Gravenhage: M. Nijhoff, 1910–55. Alboquerque, B. de. 1875–84. The Commentaries of the Great Afonso Dalboquerque, ed. W. de Gray Birch. 4 vols. London: Hakluyt Society. Anderson, J. 1826. Mission to the East Coast of Sumatra in 1823. Edinburgh/London: Blackwood. Reprinted Kuala Lumpur, 1971. Articles of Agreement. Articles of Agreement between the Kings of Goa and Tallo and the English Company, 1624. India Office Library G/10/1, fols. 35–36. Beaulieu, A. de. 1666. Memoires du Voyage aux Indes Orientales du General Beaulieu, dressés par luy-mesme. In Relations de divers voyages curieux, ed. M. Thevenot, vol. II. Paris: Cramoisy. Bik, A. J. E. A., ed. 1928. Dagverhaal eener Reis, gedaan in het jaar 1824, tot nadere Verkenning der Eilanden Kefing, Goram, Groot en Klein Kei en de Aroe-eilanden door A. J. Bik. Leiden: A. W. Sijthoff. Bloch, M. 1975. Slavery and Serfdom in the Middle Ages. Selected essays by Marc Bloch, trans. W. R. Beer. Berkeley: University of California Press. Bouchon, G. 1979. Les Premières Voyages Portugais à Pasai et à Pegou (1512–1520). Archipel 18: 127–57. Bowring, Sir J. 1856. The Kingdom and People of Siam. 2 vols. Reprinted Kuala Lumpur: Oxford University Press, 1969. Brown, D. E. 1970. Brunei: The Structure and History of a Bornean Malay Sultanate. Brunei: Brunei Museum. Castles, L. 1972. The Political Life of a Sumatran Residency: Tapanuli 1915–1940. Ph.D. dissertation, Yale University. Chabot, H. Th. 1950. Verwantschap, Stand en Sexe in Zuid-Celebes. Groningen/Jakarta: J. B. Wolters. Chirino, P. 1604. Relacion de las Islas Filipinas/The Philippines in 1600, trans. Ramon Echevarria, 1969. Manila: Historical Conservation Society. Colin, F. 1660. Labor Evangelica, ministerios apostolicos de los obreros de la Compania de Jesus, fundacion y progressos de su Provincia en las Islas Filipinas, ed. Pablo Pastells, 1900–1902. 3 vols. Barcelona: n.p. Crawfurd, J. 1820. History of the Indian Archipelago. 3 vols. Edinburgh: Constable. Crawfurd, J. 1856. A Descriptive Dictionary of the Indian Islands and Adjacent Countries. London. Reprinted Kuala Lumpur, 1971. Dampier, W. 1697. A New Voyage Round the World, ed. A. Gray, 1927. London: Argonaut Press.

“ Closed ” and “ Open ” Slave Systems

1483

Dampier, W. 1699. Voyages and Discoveries, ed. C. Wilkinson, 1931. London: Argonaut Press. Davalos, M. 1584. Melchior Davalos to King Felipe 11, 3 July 1584. In The Philippine Islands 1493–1803 VI, ed. E. H. Blair and J. A. Robertson, pp. 54–64. Cleveland, 1903. Henceforth B&R. Eerste Schipvaart. 1598. De Eerste Schipvaart der Nederlanders naar Oost-Indie Onder Cornelis de Houtman (1595–97), ed. G. P. Rouffaer and J. W. Ijzerman. 3 vols. ’s-Gravenhage: Linschoten-Vereniging, 1915–29. Florentine letter. 1513. Lettera … scripta in Lisbona e mandata a fra Zuambatista in Firenze a’ di 31 genaro 1513. In Angelo de Gubernatis, Storia dei Viaggiatori Italiani Nelle Indie Orientali. Livorno: Franc. Vigo, 1875. Floris, P. 1616. Peter Floris, his Voyage to the East Indies. In the “Globe”, ed. W. H. Moreland. London: Hakluyt Society, 1934. Francis. 1839. Oude Gegevens over de Bataks. Adatrechtbundels 18: 212–16. Freeman, J. D. 1970. Report on the Iban. London: Athlone Press. Friedericy, H. J. 1933. De Standen bij de Boegineezen en Makassaren. Ph.D. University of Leiden. Gervaise, N. 1700. An Historical Description of the Kingdom of Macasar. London: Gregg. Reprinted 1971. Haan, F. de. 1922. Oud-Batavia. 3 vols. Batavia: G. Kolff. Iwao, Seiichi. 1970. Japanese Emigrants in Batavia during the 17th Century. Acta Asiatica 18: 1–25. Jacobs, H., ed. 1971. A Treatise on the Moluccas (c.1544). Probably the Preliminary Version of Antonio Galvao’s Lost Historia das Molucas. Rome. Jesuit Historical Institute. Jourdain, J. 1905. The Journal of John Jourdain, 1608–1617, ed. William Foster. Cambridge: Hakluyt Society. Kruyt, A. C. 1911. De Slavernij in Posso (Midden-Celebes). Onze Eeuw 2:61–97. Kumar, A. 1976. Surapati: Man and Legend. A study of three Babad traditions. Leiden: Brill. Lasker, B. 1950. Human Bondage in Southeast Asia. Chapel Hill: University of North Carolina Press. Lavezaris, G. de. 1574. Slavery among the natives. B & R III, pp. 286–88. Legazpi, M. L. de. 1569. Relation of the Filipinas Islands, and of the character and condition of their inhabitants. Cebu, 7 July 1569. B & R III, pp. 54–61. Liaw, Yock Fang. 1976. Undang-undang Melaka: The laws of Melaka. Bibliotheca Indonesica 13. The Hague: Nijhoff. Ligtvoet, A. 1876. Aanteekeningen betreffende den economischen toestand en de ethnographie van het rijk van Sumbawa. TBG 25: 555–70. Lockyer, C. 1711. An Account of the Trade in India. London.

1484

Reid

Loeb, E. M. 1935. Sumatra: Its History and People. Vienna: Institute für Volkerkunde. Loubère, S. de la. 1691. A New Historical Relation of the Kingdom of Siam. Reprinted Kuala Lumpur: Oxford University Press, 1969. McEvedy, C. and Jones, R. 1978. Atlas of World Population History. Harmondsworth: Penguin. Matelief, C. 1608. Historische Verhael Vande treffelijcke Reyse, gedaen naer de OostIndien ende China, met elf Schepen, door den Manhaften Admirael Cornelius Matelief de longe. In Begin ende Voortgang Vande Vereenigde Neederlantsche Geoctroyeerde Oost-Indische Compagnie, ed. I, Commelin. Amsterdam, 1646. Vol. II of 1974 reprint. Mattulada. 1976. Latoa: Satu Lukisan Analitis terhadap Antropologi-Politik Orang Bugis. Ph.D. dissertation, University of Indonesia. Morga, A. de. 1609. Sucesos de las Islas Filipinas, trans. J. A. Cummins. Cambridge: Hakluyt Society, 1971. Nieboer, H. J. 1910. Slavery as an Industrial System: Ethnological Researches. 2nd ed. The Hague: Nijhoff. Noorduyn, J. 1955, Een Achttiende-eeuwse Kroniek van Wadjo. ’s-Gravenhage: Nijhoff. Noorduyn, J. 1983. De Handelsrelaties van het Makassaarse Rijk volgens de Notitie van Cornelis Speelman (1669). Nederlandse Historische Bronnen, vol, 3. ’s-Gravenhage: Nijhoff for Nederlands Historisch Genootschap, 1983, 97–123. Nooy-Palm, H. 1979. The Sa’dan Toraja. A Study of their Social Life and Religion, vol. I. The Hague: Nijhoff. O’Kane, J. 1972. The Ship of Sulaiman, trans. from the Persian of ibn Muhammad Ibrahim (1688). London: Routledge and Kegan Paul. Pelras, C. 1981, Célèbes-Sud avant l’Islam selon les premiers témoignages étrangers. Archipel 21: 153–84. Phelan, J. L. 1959. The Hispanization of the Philippines. Spanish Aims and Filipino Responses, 1565–1700. Madison: University of Wisconsin Press. Pigeaud, T. G. Th. 1931. Oendang-oendang van Banten (1690–1692), Vergeleken met eenige andere oendang-oendang. Adatrechtbundels 34: 39–85. Pires, T. 1515. The Suma Oriental of Tome Pires, trans. Armando Cortesao. 2 vols. London: Hakluyt Society, 1944. Plasencia, J. de. 1589. Customs of the Tagalogs. 21 October 1589. B & R VII, pp. 173–85. Raffles, Sophia. 1835. Memoir of the Life and Public Services of Sir Thomas Stamford Raffles. New ed. 2 vols. London: James Duncan. Reid, A. 1975. Trade and the Problem of Royal Power in Aceh, c.1550–1700. In PreColonial State Systems in Southeast Asia, ed. A. Reid and L. Castles, pp. 45–55. Kuala Lumpur: Malaysian Branch of the Royal Asiatic Society. Reid, A. 1979. Trade and State Power in 16th and 17th Century Southeast Asia. Proceedings of the Seventh IAHA Conference, Bangkok, 22–26 August 1977 1: 391–419. Bangkok: International Association of Historians of Asia.

“ Closed ” and “ Open ” Slave Systems

1485

Reid, A. 1980. The Structure of Cities in Southeast Asia, Fifteenth to Seventeenth Centuries. JSEAS 11, 2: 235–50. Rousseau, J. 1980. Iban Inequality. BKI 136: 52–63. Sande, F. de. 1576. Relation of the Filipinas Islands. Manila, 7 June 1576. B & R IV. Sande, F. de. 1578. Relation and Description of the Phelipinas Islands. Manila, 6 February 1578, B & R IV. Schröder, E. E. W. Gs. 1917. Nias. Ethnographische, Geographische en Historische Aanteekeningen en Studiën. 2 vols. Leiden: Brill. Scott, E. 1606. An exact Discourse of the Subtilties, Fashions, Policies, Religion, and Ceremonies of the East Indians, as well Chyneses as Javans, there abyding and dweling. Reprinted in The Voyage of Sir Henry Middleton to the Moluccas 1604–1606, ed. William Foster, pp. 81–176. London: Hakluyt Society, 1943. Sisian. 1579. Deposition of Sisian. Brunei, 22 March 1579. B & R IV, pp. 197–200. Smith, G. V. 1974. The Dutch East India Company in the Kingdom of Ayutthaya, 1604– 1694. Ph.D. dissertation, Northern Illinois University. Speelman, C. 1670. Notitie Dienende voor eenen Korten Tijd en tot nader last van de Hooge Regering op Batavia voor den ondercoopman Jan van Oppijnen. Typescript copy held in KITLV, Leiden. Turton, A. 1980. Thai Institutions of Slavery. In Asian and African Systems of Slavery, ed. J. L. Watson. Berkeley: University of California Press. Vera, S. de. 1589. Conspiracy against the Spaniards (testimony of Governor Santiago de Vera and others). May–July 1589. B & R VII, pp. 95–111. Vijewardene, G. 1980. Monks, Mediums, Cities and Sema. Typescript awaiting publication, generously made available. Vliet, J. van. 1640. The Short History of the Kings of Siam, trans. Leonard Andaya, 1975. Bangkok: Siam Society. Warren, J. F. 1981. The Sulu Zone 1768–1898: The Dynamics of External Trade, Slavery, and Ethnicity in the Transformation of a Southeast Asian Maritime State. Singapore: Singapore University Press. Watson, J. L., ed. 1980, Asian and African Systems of Slavery. Berkeley: University of California. Wilcox, H. 1949, White Stranger: Six Moons in Celebes. London: Collins. Wilken, G. A. 1888. Het pandrecht bij de volken van den Indischen Archipel. BKI 37: 555–609. Willoughby, G. 1635. G. Willoughby (Banten) to East India Company, 31 January 1635, India Office Library E/3/15, f. 115v. Winstedt, R. and de Josselin de Jong, P. E., eds. 1956. The Maritime Laws of Malacca. JMBRAS 29, 3: 22–59.

The Structure of Slavery in the Sulu Zone in the Late Eighteenth and Nineteenth Centuries James Francis Warren Introduction At the end of the eighteenth century, the world economic system came to dominate the Sulu Sultanate and its surrounding areas. The island of Jolo became an entrepôt and the Sulu Sultanate flourished. This prosperity was made possible by economic interconnections between British India, South-East Asia and China. At that time in England, tea replaced ale as the national drink. Most of this aromatic tea was cultivated in the mountains of Fujian, China. The south Fukienese people who controlled the Amoy trading network on the South China coast had maintained contact with the Sulu Sultanate and eastern parts of the archipelago since the beginning of the eighteenth century. Because of this contact, the dominant ethnic group, the Taosug, had an initial outlet for their marine and forest products, especially trepang, pearls and birds’ nest, which the junks carried on the return trip to China. When the British realized that they could participate in the Sino-Sulu trade, they also sought these Sulu zone exports.1 Among the richest sources of such products were the shores and wilderness along the northeast coast of Borneo. Kenyah and Kayan speakers of Borneo’s east coast tropical forests and the Subanun of Mindanao’s southern plateau were increasingly drawn into trade with the Taosug. They provided enormous amounts of wax, camphor and birds’ nest in exchange for salt, textiles and other manufactured goods. At first, trepang was traded for cloth, clothing, iron and other metals; later, for gunpowder, musket and cannon. Guns, gunpowder and other imported products, including textiles and opium became very important to Sulu. Source: Warren, James Francis, “The Structure of Slavery in the Sulu Zone in the Late Eighteenth and Nineteenth Centuries,” Slavery and Abolition, 24(2) (2003): 111–128. © Taylor & Francis Ltd, reprinted by permission of the publisher Taylor & Francis Ltd, http://www.tandfonline.com. 1  See J. F. Warren, The Sulu Zone, 1768–1898, The Dynamics of External Trade, Slavery, and Ethnicity in the Transformation of a Southeast Asian Maritime State (Singapore: Singapore University Press, 1981); J. F. Warren, The Sulu Zone, The World Capitalist Economy and the Historical Imagination (Amsterdam: VU University Press, 1998). © koninklijke brill nv, leiden, ���8 | doi ��.��63/9789004346611_048

The Structure of Slavery in the Sulu Zone

1487

The efforts of ambitious datus, Taosug aristocrats, to participate in this burgeoning world-capitalist economy swelled the flow of global–regional trade and forced up the demand for additional labour, which was met by Iranun and Samal Balangingi slave raiders of the Sulu zone. From the end of the eighteenth to the middle of the nineteenth centuries, South-East Asia felt the full force of these slave raiders who earned a reputation as daring, fierce marauders who jeopardized the maritime trade routes of South-East Asia and dominated the capture and transport of slaves to the Sulu Sultanate.2 Tens of thousands of captives from across South-East Asia were seized by these sea raiders and put to work in the zone’s fisheries, in the sultan’s birds’ nest caves, or in the cultivation of rice and transport of goods to markets. More than anything else it was this source and use of labour power that was to give Sulu its distinctive predatory character in the eyes of nineteenth-century Europeans. It was powerful economic forces, energized by world commerce demands for a range of products, that pushed the Taosug aristocracy to acquire increasing numbers of slaves. In order to trade, it was necessary for the Taosug to have something to give in exchange and the only way to obtain commodities was to secure more slaves by means of long-distance maritime raiding. In the early nineteenth century, the rate of growth of the sultanate’s population had not kept pace with its expanding international trade economy. Since the labour of slaves made global–regional trade possible, slavery rose markedly from this time and became the dominant mode of production.

Slavery in the Sulu Sultanate

Slavery in the Sulu Sultanate was not as rigidly defined an institution as in the West where it was historically synonymous with property. Slavery in Sulu, as in other areas of South-East Asia, was primarily a property relation but not exclusively so. In this context, slavery must be understood to imply several possible statuses of ‘acquired persons’ who were sometimes forcefully transferred from one society to another.3 Taosug drew a distinction between chattel-slaves

2  Warren, The Sulu Zone, 1768–1898, pp. 147–211. 3  See A. Reid (ed.), Slavery, Bondage and Dependency in Southeast Asia (St. Lucia: University of Queensland Press, 1983); J. L. Watson (ed.), Asian and African Systems of Slavery (Oxford: Basil Blackwell, 1980); J. F. Warren, ‘Slavery in Southeast Asia’, in S. Drescher and S. L. Engerman (eds.), A Historical Guide to World Slavery (New York: Oxford University Press, 1998), pp. 80–87; on the problem of defining slavery see E. R. Leach, ‘Caste, Class and Slavery: The Taxonomic Problem’, in A. de Reuck and J. Knight (eds.), Caste and Race: Comparative Approaches

1488

Warren

(banyaga, bisaya, ipun or ammas) and bond-slaves (kiapangdilihan). Banyaga were either the victims or the offspring of victims of slave raids; Kiapangdilihan were commoner Taosug whose servility was the direct result of personal debt.4 Between these two categories of slave there was a continuum of status and privilege in which social position was determined by factors often independent of their servile status. In Sulu, banyaga could have family roles as husband or wife. They could own property, and they often filled a variety of political and economic roles— as bureaucrats, farmers and maritime raiders, concubines and traders—by virtue of which they were entitled to certain rights and privileges also accorded to non-slave members of the community. As such, slavery was a means of incorporating people into the Taosug social system. Banyaga were enrolled to provide political support for datus and to labour in the fields and fisheries to maintain the expansive redistributional economy. They were predominantly Visayan, Tagalog, Minahassan and Bugis speakers although almost every major ethnic group of insular South-East Asia was represented among their ranks. While some inherited their banyaga status, others were fulfilling tax or debt obligations. However, all banyaga or their descendants had been seized by slave raiders and retailed in communities throughout the Sulu chain.5 Capture in raiding was the principal mode of recruitment, but debt and fine obligations among the Taosug themselves provided a significant number of bond-slaves. Convictions for criminal offences such as stealing and acts of sexual impropriety, particularly adultery, were punishable by heavy fines.6 Debts (London: J. and A. Churchill, 1967); see also, S. Miers and I. Kopytoff (eds.), Slavery in Africa Historical and Anthropological Perspectives (Madison: University of Wisconsin Press, 1977). 4  M. Mednick, ‘Encampment of the Lake: The Social Organisation of a Moslem Philippine (Moro) People’ (PhD thesis, University of Chicago, 1965), pp. 60–61; T. Kiefer, ‘The Taosug Polity and the Sultanate of Sulu: A Segmentary State in the Southern Philippines’, Sulu Studies, I (1972), p. 30; T. M. McKenna, Muslim Rulers and Rebels (Berkeley: University of California Press, 1998) pp. 48–51. 5  Diary of William Pryer, 25 Nov. 1878; 26 June 1879, Colonial Office, Public Records Office, London (herafter CO-PRO), 874/68. 6  B H-PCL (Beyer-Holleman Collection, Philippine Custormary Law), The Library of Congress, Washington DC. (hereafter LC), Vol. VI, paper 162, No. 16; A. Gunther, ‘Correspondence, Papers and Reports Relating to the Sulu Moros’, Jolo and Manila (1901–1903), pp. 10–12; paper 162, vol. VI, No. 25, Emerson B. Christie, ‘The Non Christian Tribes of the Northern Half of the Zamboanga Peninsula’ (1903) p. 87; paper 163, Vol. VI, No. 28; L. W. N. Kennon, D. P Barrows, J. Pershing and C. Smith, ‘Census Report Relating to the District of Lanao Mindanao’(1903), p. 4; N. Saleeby, Studies in Moro History, Law and Religion (Manila: Bureau of Public Printing, 1905) pp. 92–3.

The Structure of Slavery in the Sulu Zone

1489

were also incurred by gambling and pawning. Inability to pay or offer some form of security reduced people to the status of kiapangdilihan. Kiapangdilihan were an integral part of a creditor’s following, but with a lower status than commoners who had voluntarily attached themselves to a leader. The creditor, while claiming rights over the economic services of kiapangdilihan was, in theory, not allowed to harm the person physically. In return for food, clothing and shelter, a kiapangdilihan was obliged to work for the creditor but these services did not generally count towards repayment of the debt. Many kiapangdilihan became dependents for life and their families could remain obligated for several generations. Debt bondage, as an economic institution in Sulu, was most fully developed at the end of the nineteenth century, when the Taosug could no longer rely on Balangingi raids to supply sufficient numbers of banyaga for their retinues. At this time, the amount of tribute ordinarily collected from clients was increased and fines in the legal codes were made prohibitive.7 In the Sultanate of Sulu, the legal position of a banyaga was determined by the Sulu code—a body of law codified from custom, precedent and Islamic law.8 In theory, as defined in the Taosug codes, a banyaga had no legal personality,9 and was left absolutely in the power and possession of their owner. A banyaga could not hold property and could be transferred, bought, or sold at will. An owner held the power of life and death over a banyaga who could be punished for the slightest infraction of the law. Punishments were much more severe for banyaga than members of other social classes and are exemplified in the scale of penalties and fines in the codes for the offences of murder, adultery, theft and inheritance. If a male banyaga had sexual intercourse with a free woman, he could either be killed outright or be severely punished and become the property of the woman’s husband or family.10 On the other hand, if a free man had sexual relations with a married female slave he need only pay a fine of twenty lengths of cotton cloth.11 Less severe penalties for adultery between banyaga derived from their inferior social status. 7  Scott to Governor, 30 June 1904, H. L. Scott Papers, Container 55, LC; Otis to Bates, 11 July 1899, U.S. Senate Documents, 9; Otis to Bates, 11 July 1899, U.S. Senate Documents, 9, document 136, p. 15, LC; Saleeby, Studies in Moro History, p. 94. 8  Saleeby, Studies in Moro History, pp. 65, 81, 89. 9  Ibid., pp. 76, 86–7. 10  Ibid, pp. 71, 83, 93; No. 139, El Gobernador de Zamboanga a Gobernador Capitan General (hereafter GCG), 12 Feb. 1845, Philippine National Archives, Manila (hereafter PNA), Mindanao/Sulu 1836–97. 11  Saleeby, Studies in Moro History, p. 93.

1490

Warren

Although these laws provide institutional opinion on the debasement of people, banyaga were often socially and economically indistinguishable from commoners and were, in some respects, more secure. The actual life situation of many banyaga, as revealed in their testimonies, contradicted their legal status as a group. Banyaga were permitted to purchase their freedom and assume a new status and ethnicity; the children of a female banyaga and a freeman inherited the status of their father; some banyaga could bear arms; any slave could own property, which reverted to their owner at death. The basic difference between slavery among the Taosug and slavery as it was generally understood in the West was the variability of social distance that existed between slave and owner. William Pryer stated that on the east coast of Borneo the relation was that of follower and lord rather than slave and master.12 The power and wealth of a datu was commensurate with the number of slaves he owned which in turn was reflected in the number of people willing to seek security within a particular datu’s settlement in return for services. Banyaga were often well-clothed, carried fine kris, and were entrusted to undertake long journeys for their owners.13 The personal and economic ties of slaves in the Sulu Sultanate ‘provided a sense of security which bound them to their masters and gave them an identity and reason to labour’.14 A slave’s individual status was enhanced by the prominent status of an acquiring owner. Alternatively, when a powerful slaveholder suffered a serious loss of prestige, the slaves’ worth as human beings also diminished. An owner was constrained to feed and clothe the slaves or give them sufficient opportunity to earn a living; otherwise the slaves might demand to be sold.15 It appears to have been a common practice in the Sulu Sultanate to allow a banyaga to change owners rather than risk desertion. Nevertheless there are also statements of fugitive slaves and other reports which present a much less ‘benign’ view of the owner–slave relationship. In principle, the master’s ownership and authority were absolute and unbounded. A banyaga could suffer physical cruelty, be put to death, sold, bartered or given away if it served the owner’s interests.16 Banyaga who repeatedly tried to escape were 12  Diary of William Pryer, 14 March 1878, CO-PRO, 874/68. 13  W. Pryer, ‘Notes on Northeastern Borneo and the Sulu Islands’, Royal Geographical Society Proceedings, 5 (1883), pp. 92–3. 14  J. K. Reynolds, ‘Towards an Account of Sulu and its Bornean Dependencies, 1700–1878’ (M.A. dissertation, University of Wisconsin, 1970), p. 81. 15  Pryer, ‘Notes on Northeastern Borneo’, p. 92. 16  T. Forrest, A Voyage to New Guinea and the Moluccas from Balambangan: Including an Account of Magindano, Sooloo and other Islands (London: G. Scott, 1779), p. 330.

The Structure of Slavery in the Sulu Zone

1491

put to death, but it was far more common for them to be disposed of—in most cases to the Segai-i of east Borneo.17 While there is evidence of contrasting degrees of benevolence and material deprivation in the sultanate, the fact that power was defined in terms of the number of banyaga a datu possessed, and that banyaga were able to flee to another datu or attempt to escape to Zamboanga or Menado, were important constraining considerations. A purely antagonistic relation would little benefit an owner’s self-interest since the successful exploitation of Sulu’s natural produce hinged on the large-scale organization of the goodwill and cooperation of slaves and their dependents. While an owner was liable to neglect or mistreat a banyaga who was remiss, statements and travel accounts of observers reveal that some slaves, especially those with knowledge and skills, had good relations with their owner and were not easily distinguished from other followers. Towards the end of the eighteenth century, the impact of the West’s commercial intrusion into China was a watershed in the formation of the Sulu state. At this time, slaves who were valuable for the variety of their labours came to play a more avowedly important role in Sulu. Banyaga were used in trading ventures and diplomatic negotiations. They were slave raiders, concubines, wet nurses, tutors, craft workers, peasants and fishers,18 and there was a clear division of labour between the work of male and female banyaga. Heavy work was performed generally by male slaves—clearing virgin forest, ploughing, harvesting timber, building and maintaining boats, and hauling water.19 Male banyaga also laboured in the fisheries, manufactured salt, accompanied their owners on trading expeditions and, when occasion demanded, sailed as crew on Balangingi raiding prahus. The crew of such vessels invariably included banyaga, who were previously captured from coastal settlements and from fishing and trading boats, because of their seafaring skills and local knowledge of target areas, and to replace members of the raiding communities lost to battle, storm and illness on the high seas. Among the major tasks of female banyaga were sowing and weeding in rice farming, pounding and threshing 17  Witti to Treacher, Nov. 1881, CO-PRO, 874/229. 18  J. Hunt, ‘Some Particulars Relating to Sulo in the Archipelago of Felicia’, in J. H. Moor (ed.), Notices of the Indian Archipelago and Adjacent Countries (London: Cass, 1967), p. 37; A. J. F. Jansen, ‘Aanteekeningen omtrent Sollok en de Solloksche Zeeroovers’, Tijdschrift voor Indische Taal-Land en Volkenkunde, vitgegeven door het (Koninklijk) Bataviaasch Genootschap van Kunsten en Wetenschapen (herafter TITLV) 7 (1855), p. 224. 19  Statements of Mariano de la Cruz and Francisco Gregorio in Expediente 12 (4 Oct. 1836), PNA, Mindanao/Sulu 1803–1890; El Gobierno Politico y Militar del Zamboanga a GCG, 9 June 1847, PNA, Mindanao/Sulu 1838–85.

1492

Warren

of rice, and gathering and preparation of strand products.20 Female banyaga were also included in the entourage of their female owners as attendants, and some were the concubines of leading datus. Slavery played an essential role in the economic and military organization of the Sulu Sultanate. Banyaga were encouraged to participate in the economic life of the state through a system of incentives, which some used to advantage thus enabling them to rise in the social hierarchy. J. Hunt noted that the Taosug employed slaves in their prahus not only as crew, but also as traders.21 In the 1830s, slaves regularly traded from Jolo to Balangingi and Palawan on behalf of their owners.22 Those who exhibited initiative and energy were employed in trading excursions to the northeast coast of Borneo. Noble women, by virtue of their station were constrained from leaving their quarters to barter goods. Nevertheless, aristocratic women used banyaga to assist them in their business activities—primarily local marketing.23 Indeed, by the mid-nineteenth century, some of the leading local traders in Sulu were women, and slave hawkers were an important source of their wealth. Ordinarily, the vending of cloth, vegetables, and other trade goods in villages, at the open market, or to foreign vessels was done by banyaga. On the arrival of a European vessel, it was common for Taosug women to send one or two Spanish-speaking slaves into the roadstead in small canoes.24 The boats carried fruits, vegetables, coils of tali lanun (cheap local rope of excellent quality), weapons and curiosities. In 1834, an American sailor described the boats of slave vendors at Siassi as being full of ‘poultry, eggs, coconuts, bananas,

20  Extracts from Singapore Free Press, 6 April 1847, Public Records Office, London (hereafter PRO), Admiralty 125/133; BH-PCL, Vol. VI, paper 162, No. 26; F. P. Williamson, ‘The Moros between Buluan and Punta Flecha’ (1903), p. 103; Mednick, ‘Encampment of the Lake’, p. 62. 21  Hunt, ‘Some Particulars Relating to Sulo’, p. 37. 22  Statements of Alex Quijano, Francisco Sacarias and Domingo Francisco in Expediente 12 (4 Oct. 1836), PNA; statement of Juan Florentino in Relacion jurada de los dos individuos cautivos venidos en la corbetta de guerra Francesa Salina procedente de Sumalasan en la Archipelago de Jolo, PNA, Piratas 3; Treacher to Sir Rutherford Alcock, 3 July 1884, COPRO, 874/237. 23  Pryer, ‘Notes on Northeastern Borneo’, p. 93. 24  Statements of Anastacio Carillo and Ignacio Valero in Expediente 12 (4 Oct. 1836), PNA; R. P. Robinson, Journal kept aboard the Vicennes, 4 Feb. 1842, Department of the Navy, The United States National Archive, Washington D.C. (hereafter DN-USNA); S. St. John, Life in the Forests of the Far East, Vol. 2 (London: Smith Elder & Company, 1862), p. 193.

The Structure of Slavery in the Sulu Zone

1493

turtles, monkeys, and parrots’.25 The banyaga were instructed to barter a specified minimum amount of produce by evening in exchange for which they accepted cups and saucers, scissors, buttons, nails, empty bottles, tobacco and opium from European sailors. Those banyaga employed in agriculture contributed towards the community sustenance, thus freeing datus from subsistence pursuits, and allowing them to pursue trading and maritime raiding.26 Small, dispersed farming communities of banyaga dotted the interior of the larger fertile islands—especially Jolo, Tapul and Pata. Banyaga were allotted a farm plot and a bamboo hut large enough to accommodate a single family, which were scattered over large tracts of land. In some cases these agricultural settlements were homogeneous in language and religion, as exemplified by the fact that most banyaga who were settled in Datu Tahel’s community in the hinterland of Jolo Island were Bugis and ‘Malay’. Other communities were settled almost entirely by Visayans. Agricultural slaves were expected to provide for their own wants from the fields they tilled. They were obligated to remit a fixed minimum portion of produce to their owner through the agency of the village head, who could be of slave or non-slave origin. Farming was their major economic obligation, but datus also demanded that villagers near the coast collect trepang and pearl shell, for which they received barter goods. All were liable to be called upon for military services.27 While, for farm slaves, opportunities for social advancement were few, to many Filipinos the Taosug system of sharecropping and reciprocity was a less harsh form of exploitation than the poverty and social injustice they had previously endured under the tributo system of the Spanish Government. The prosperity of the Sulu Sultanate depended to a large extent on the labour of banyaga who crewed the raiding prahus. The banyaga augmented the strength of client communities that specialized in slaving through active participation in maritime raids. In this way, owners, who exercized an enforceable claim on the wealth their banyaga received when hired out on raiding vessels, were enriched.28 Undoubtedly datus were constrained to reward such 25  T. J. Jacobs, Scenes, Incidents and Adventures in the Pacific Ocean, or the Islands of the Australasian Seas, during the Cruise of the Clipper Margaret Oakley (New York: Harper and Bros., 1844), p. 335. 26  Farren to Palmerston, 17 Jan. 1851, CO-PRO, 144/8; Corbett to the Sec. of the Admiralty, 6 Oct. 1862, Foreign Office, PRO, 71/1. 27  Statements of Vincente Remigio, Francisco Octoberino and Juan Saballa in Expediente 12 (4 Oct. 1846), PNA, Mindanao/Sulu 1803–1890. 28  Statement of Alex Quijano in Expediente 12 (4 Oct. 1836), PNA, Mindanao/Sulu 1803–1890.

1494

Warren

banyaga, who might otherwise have been reluctant to participate in such hazardous undertakings. Nevertheless, raiding seems to have provided slaves with opportunities for modest social advance, especially if they showed a talent for fighting. Differences in wealth, status and privilege among banyaga were reflected in the diversity of occupations they pursued. Obviously, banyaga of unusual talent or considerable ability and training who could not readily be replaced commanded more prestige and rewards than those with few skills. In prominent trading centres like Jolo, Parang and Bual, where agriculture was of secondary importance, talented banyaga included bureaucrats, tribute collectors, artisans, musicians, scribes and commercial agents among their numbers. Banyaga recruited by the sultan’s officeholders enjoyed considerable power and prestige. The interests of these banyaga, by virtue of their elevation to political office, lay unquestionably with the sultan, and they made loyal followers. Banyaga played leading roles as bureaucrats on the Samal islands, acted as tribute collectors throughout the zone, and staffed tariff stations on Bornean rivers.29 The sultan also made use of banyaga to exercise control over dependent groups on the northeast coast of Borneo. As Taosug trade became more complex, and the political problems posed by the West grew, so literacy skills became important. The use of written documents was no longer confined to records of the genealogy of the sultan, the appointment of officials and the collection of tribute and legal fees. After 1768, writing was required for diplomatic and trade correspondence with the Spanish, Dutch and English; for recording grants of land and the terms of treaties of various sorts with the West; and to keep track of the accounts of datus’ commercial enterprises. Paradoxically, few Taosug aristocrats could read and write and banyaga with education who could serve as scribes, interpreters and language tutors were much sought after.30 As Charles Wilkes commented, ‘Few if any of the Sooloos can write or read, though many talk Spanish. Their accounts are all kept by the slaves. Those who can read and write are, in consequence, highly 29  Statement of Matias de la Cruz and Francisco Sacarias in Expediente 12 (4 Oct. 1836), PNA; Verklaring van Chrishaan Soerma, 10 Aug. 1846, Arsip Nasional Republik Indonesia, Jakarta (hereafter ANRI), Menado 50; W. Pryer, ‘Diary of a Trip up the Kinabatangan’, p. 119, CO-PRO, 874/68. 30  W. Briskoe, Journal kept aboard the Relief and the Vicennes, Vol.2 (5 Feb. 1842), DNUSNA; Diario de mi Comision a Jolo en el Vapor Magallenes (19 March 1848), PNA, Mindanao/ Sulu unclassified bundle; Jansen, ‘Aanteekeningen omtrent Sollok en de Solloksche Zeeroovers’, p. 214.

The Structure of Slavery in the Sulu Zone

1495

prized.’31 Most of the scribes were male slaves drawn from different parts of the Malay world, but female Filipino slaves and the occasional deserter from Zamboanga all served as the sultan’s secretary at different times.32 While most other slave specialists—artisans and craftspeople—were more or less expendable, the skills of the educated banyaga could not easily be replicated by others and were considered indispensable to the datus. Banyaga who could speak or write one or more foreign languages were employed as trading agents by datus, enabling them to amass considerable personal wealth—which reverted to the owner on their death.33 The number of slave artisans—goldsmiths, silversmiths, blacksmiths and weavers—was never large and comprised a fraction of the total slave population. Gifted banyaga were full-time artisans transforming raw materials, brought by trade or tribute, into jewellery, tools, weapons and armour. Less talented others pursued their occupations on a part-time basis. Not surprisingly, the arbitrary distribution of banyaga left some talents wasted. Jose Ruedas, a silversmith, spent three years as a fisher and gatherer of pearl shell before being exchanged for a bundle of cotton cloth at Jolo, where he resumed his craft.34 While some banyaga found their skills superfluous in a particular island’s economy, others appear to have had the opportunity to acquire training in critical occupations—especially as blacksmiths and armourers.35 It is clear from the accounts of Thomas Forrest, J. Hunt, Dumont D’Urville, and Charles Wilkes that slaves were called upon to recite Visayan poetry or to perform instrumental music and song recitals for religious festivals and when Europeans visited Jolo.36 Under such circumstances, there was ample opportunity for banyaga with musical talent to improve their material condition. Furthermore, some datus played the flute, violin or guitar and all were fond of Spanish songs and dances. Filipino slaves could and did act as both music 31  C. Wilkes, ‘Jolo and the Sulus’, in E. H. Blair and J. A. Robertson (eds.), The Philippine Islands, XLIII (Cleveland: A. H. Clark Company, 1903–1919), pp. 128–92; pp. 160–61 (the first citation refers to the reprinting of the entire account of the American naval expeditions (Charles Wilkes) visit to Sulu). 32  D. Patricio de la Escosura, Memoria sobre Filipinas y Jolo redactada en 1863 y 1864 (Madrid: Manuel G. Hernandez, 1882), p. 371; Diary of William Pryer (8 March 1879), CO-PRO, 874/68. 33  Wilkes, ‘Jolo and the Sulus’, p. 166; Dumont D’Urville, Voyage au pole sud et dans l’Oceanie sur les Corvettes l’Astrolobe et la Zelee (Paris: Gide et J. Baudry, 1841–54), Vol. 7, p. 170. 34  Statement of Jose Ruedas in Expediente 12 (4 Oct. 1836), PNA, p. 32. 35  Statement of Gabriel Francisco in Expediente 12 (4 Oct. 1836), PNA, p. 32. 36  Forrest, Voyage to New Guinea, p. 330; Hunt, ‘Some Particulars Relating to Sulu’, p. 40; D’Urville, Voyage au pole sud, Vol. 7, pp. 308, 313; Wilkes, ‘Jolo and the Sulus’, p. 165.

1496

Warren

instructors to datus and entertainers at night. Banyaga who were talented musicians could expect to be transferred from one office-holder to the next.37 Banyaga with medical knowledge were perhaps most scarce. The experience of Captain C. Z. Pieters among the Balangingi illustrates that slaves who claimed even a rudimentary understanding of medicine and medical practice could enjoy a privileged position. As he wrote, One day my master and his wife asked me to what kind of work I was accustomed. I said … that my former master had only employed me in looking after his goods, accounts and dollars, and giving medicine to sick people. When they learnt this from me, my master went and told everyone that he had a slave who could cure all kinds of sickness. The consequence of this was that on the following day many persons came and asked me to tell them … what was the matter with them.38 As concubines (sandil), women achieved a high status. Concubinage was an important part of the traditional social structure in Sulu. It was a key means of producing wealth and maintaining rank since concubines could be offered to men as mistresses or wives. In theory there was no limit to the number of concubines that a sultan or datu could acquire for his retinue.39 When Forrest visited Cotabato, the sultan had one wife and fourteen or fifteen concubines.40 In Sulu, datus rarely had more than a few concubines. Many of them were kiapangdilihan; others were purchased; some were given to the sultan and leading datus as potential wives in the hope of forging political alliances through the kinship bond.41 Concubines had a recognized status and one who bore a child to her master could expect to be manumitted. The child of a concubine could not inherit the status of the father but was given a lesser title and incorporated into his following.42 Indicative of their higher status was the fact that some Filipino women willingly became concubines.

37  Diary of William Pryer (25 Nov. 1878), CO-PRO, 874/68. 38  C. Z. Pieters, ‘Adventures of C. Z. Pieters among the Pirates of Magindanao’, in Journal of the Indian Archipelago and Eastern Asia (1858), p. 305. 39  Artura Garin y Sociats, ‘Memoria sobre el Archipielago de Jolo’, in Boletin del la Sociedad geografica de Madrid, 10 (1881), pp. 110–33, 161–97. 40  Forrest, Voyage to New Guinea, p. 247. 41  Escosura, Memoria sobre Filipinas y Jolo, p. 371; Kiefer, ‘The Taosug Polity’, p. 375. 42  Diary of William Pryer (3 Feb. 1879), CO-PRO, 874/68; BH-PCL, 1, Paper 160, No. 1, Major Charles Livingston, ‘Constabulary Monograph of the Province of Sulu’ (1915) p. 14.

The Structure of Slavery in the Sulu Zone

1497

The boundaries between segments of the retainer class (commoners who voluntarily attached themselves to datus) and slaves were not clear-cut. There was considerable mobility for talented members of dependent status groups within and between all classes and strata. Among the hierarchy of banyaga, those who functioned as bureaucrats, craftsmen, scribes and concubines often had a greater degree of power and privilege than Taosug commoners. For most slaves, however, opportunities for advance were modest. Dramatic advance depended largely on historical circumstance, good fortune, skills and personal character, and were associated with occupations that provided access to wealth and power. There is some evidence illustrating that in rare instances banyaga of remarkable talent rose to the rank of orang kaya, a man of some means, and datu as proteges of their owners.43

Manumission, Ransom or Escape

Manumission was an important feature of the Taosug social system and was commonly practised in the Sulu Sultanate when freed slaves were merged into the general population, by assuming a new ethnicity and status.44 For banyaga, conversion and/or marriage were prerequisites to manumission. The process of manumission in the Taosug social system occurred primarily among banyaga who were in close contact with their owners, and tended to be gradual with an implicit understanding of cultural incorporation.45 A Tagalog or Visayan who altered ethnic identity by becoming a Muslim and was manumitted found a new range of opportunities open as a ‘Taosug’ and a free person. It is not clear whether captured Muslim slaves could be manumitted and then rise in status more easily than renegade Filipinos. However, groups, like the Bugis, who took pride in their cultural heritage, found the process of assimilation more difficult. Manumission was easier for women than men. It was not necessary for a female Filipino slave to renounce her faith to be manumitted. Whether or not she adopted Islam, her children were raised as Muslims and were absorbed into Sulu society. Marriage played an important role in the Sulu social system. 43  José Montero y Vidal, Historia de la Pirateria Malayo-Mahometana en Mindanao Jolo y Borneo (Madrid: M. Tello, 1888), p. 69; Leonard Wood Papers, LC, container 3, Diary of Leonard Wood (18 Aug. 1903). 44  M. Mednick, ‘Some Problems of Moro History and Political Organisation’, Philippine Sociological Review, 5 (1957), p. 48. 45  Statement of Francisco Enriquez in Escosura, Memoria sobre Filipinas y Jolo, p. 373.

1498

Warren

The provision of wives was a basis on which followings were built and retained. For the upwardly mobile captive, an arranged marriage was closely linked to status advancement or manumission. In the Sulu Sultanate, banyaga could purchase freedom.46 Indeed, the steady leakage of manumitted slaves increased datus’ political authority and prestige by swelling their retainer ranks. The likelihood of manumission was essentially a function of occupation. Banyaga who provided immediate and indispensable services to their owners, who served in their households or on their trading prahus, had better chances of manumission than those who laboured in the fields or fisheries. The Iranun and Balangingi sometimes permitted people they seized to redeem themselves by ransom. Often this was done immediately after their seizure, but banyaga were also ransomed from the Taosug, sometimes after having spent a considerable period in captivity. When away on cruises, slave raiders went out of their way to take captives for whom ransoms might be obtained—and a banyaga was always entrusted to fetch the ransom. Züñiga observed that only certain types of persons could be ransomed. By 1800, there was a standard scale of ransom fees. A friar was valued at about 2000 pesos, a European at 300 pesos, and a male Filipino slave at 30 to 50 pesos. Lascars taken from English ships were ransomed at 100 pesos each.47 Given the magnitude of slavery, it was inconceivable for the Church to ransom Filipino slaves generally, but villages were sometimes expected to raise the money for the ransom of friars, and were left impoverished as a result.48 Since it was not a widespread practice for Malay royalty to obtain the release of their kindred by negotiation and ransom,49 the heavy ransoms offered for friars meant that they were especially coveted by the slave raiders.50 The ransoming of priests and other Europeans at Jolo was a common practice in the eighteenth century, and the sultan took an active part in such negotiations, especially in cases that involved the Governor of Zamboanga.51 By 1830, however, the Sultan of Sulu no longer condoned the seizure or ransom of priests. He did not consider it 46  O. J. W. Scott and L. C. Brown, ‘Ethnography of the Magindanaos of Parang’, BH-PCL, VI, Paper 163, No.34 (1908), p. 16; Garin y Sociats, ‘Memoria sobre el Archiepelago de Jolo’, p. 171. 47  Robinson, Journal (31 Jan. 1842) and (4 Feb. 1842). 48  No.226, AGI (Archivo de Indias, Seville), Filipinas 492, p. 22. 49  Extracts from Singapore Free Press (6 April 1847), PRO, Admiralty 125/133. 50  Memoir of Sooloo, 101, Orme Collection, Records of the East India Company, India Office Library, London (herafter EICL), 67, p. 128. 51  Sultan Muyamad Alimudin a Gobernador D. Juan de Mir, 13 May 1781, PNA, Mindanao/ Sulu 1769–1898; V. Barrantes, Guerras Piraticas de Filipinas Contra Mindanaos y Joloanos (Madrid: Manuel Hernandez, 1878), p. 162.

The Structure of Slavery in the Sulu Zone

1499

advantageous from the standpoint of trade and politics to permit Europeans, particularly priests, to be brought to Jolo where their presence could become a cause célèbre and jeopardize relations with Western powers. This became a source of some of the resentment and conflict that developed between the Taosug and Balangingi after 1835. The Balangingi refused to yield the religious and Europeans among their slaves to the sultan, to be turned over to the Zamboanga authorities, and defiantly pressed for their ransom. Before 1850, European slaves in Jolo were regularly redeemed by ships’ captains who traded there. Some ships’ captains took advantage of the fact that they held out the sole opportunity of redemption for Filipinos. Merchant– traders considered them a cheap, convenient source of labour, and redeemed Filipino slaves had to work the ransom price off in passage. As crew members, their indenture was apt to last up to a year and take them increasingly further away from their villages.52 For many banyaga escape remained their central ambition. This was particularly the case of Filipino men who had been separated from their families, or who had clung to their faith.53 Age was also a factor. Slaves who had been seized in their youth were more easily assimilated to the Taosug social system,54 while banyaga who retained memories of another home were more prone to escape. It was during the early years of captivity that the desire to escape was greatest. From a total of 50 slaves who escaped to the Spanish fleet visiting Jolo in 1836, 38 had been in captivity fewer than six years; 7 for between six and eight years; and the remainder for more than a decade. In 1847, the number of newly acquired Filipino slaves who either escaped or were ransomed to the captain of the brigantine Cometa was even higher; 43 out of the 45 slaves had been in captivity fewer than six years. Of this number, 25 had been in captivity less than a year, 7 for two years, 8 for three years, and 3 for five years. Similarly, between 1845 and 1849, of the 26 Dutch subjects who escaped to Menado by small sailboat, 17 had been in captivity for fewer than three years.55

52  Statement of Ignacio Ambrocio (29 Aug. 1850), PNA, Piratas 3; Extracts from the Journal of the barque Osprey in Col. Cavanagh to the Sec. of the Government of India (28 Jan. 1863), FO, 71/1; No. 12 (31 May 1845), Algemeen Rijksarchief, The Hague, Kolonien 2669 and No. 31 (24 March 1847), Kolonien 2692. 53  See the statements in ‘Relacion jurada de los individuos cautivos venidos en la Fragata de guerra Inglesa Samarang’ (15 March 1845), PNA, Piratas 3; Hunt, ‘Some Particulars Relating to Sulo’, p. 50. 54  Jansen, ‘Aanteekeningen omtrent Sollok en de Solloksche Zeeroovers’, p. 225. 55  Expediente 12 (4 Oct. 1836), PNA; Relacion jurada de los cuarenta y cinco cautivos venidos de Jolo sobre el Bergantin Espanol Cometa (18 March 1847), PNA, Piratas 3; Verklaringen van ontvlugten personen uit de handen der Zeeroovers van 1845–49, ANRI, Menado 37.

1500

Warren

The Taosug system of bondage and dependency was such that coercive control was difficult to apply. One to two hundred banyaga fled annually to foreign vessels at Jolo, to the interior of an island in the Jolo archipelago, or to Zamboanga and Menado.56 Once the decision was made to escape, a banyaga often sought out another slave from the same ethnolinguistic group or province with whom to make plans. This banyaga could belong to the same owner or live in a contiguous settlement. Obviously, common heritage was important in maintaining secrecy and the cooperation necessary for a successful escape.57 This seems to have been especially so for banyaga who attempted the long and difficult journey to Menado. Escape from Jolo and neighbouring islands to Zamboanga was relatively easy. Zamboanga’s proximity to, and commercial dependence on, Jolo offered banyaga ample opportunities to reach the presidio on trading prahus or by small canoe. Prospects for escape diminished the further away from Jolo a banyaga was sold.58 This was especially the case for banyaga retailed on the northeast coast of Borneo where some waited 18 years before a chance to flee presented itself in the form of a punitive expedition or the visit of a European trading boat.59 Escape by small boat to the Visayas or Menado was a hazardous undertaking. Unpredictable winds and strong currents could take a small canoe far from its destination, or it could break up and sink in choppy seas. The voyage to Menado took at least ten days and required an outriggered sailing boat large enough for at least two people, but such attempts often required three to five banyaga. The difficult nature of the voyage to northern Celebes is apparent from the small number of slaves who were successful. From 1845 to 1849, those slaves to successfully reach Menado were: 1855, fifteen; 1857, twenty-

56  GCG a Presidente del Consejo de Ministro, 9 Dec. 1858, Archivo Historico Nacional, Madrid, Ultramar 5184; Hunt, ‘Some Particulars Relating to Sulo’, p. 50. 57  Statements of Pedro Flores and Pedro Isidoro in Relacion jurada de los cuarenta y cinco cautivos venidos de Jolo sobre el Bergantin Espanol, Cometa (19 March 1847), PNA, Piratas 3; Statements of Francisco Anastacio and Jacinto Pedro in Relacion jurada de los individuos cautivos venidos en la Fragata de guerra Inglesa Samarang (15 March 1845), PNA, Piratas 3; ‘Berigten omtrent den Zeeroof in den Nederlandsch-Indischen Archipel, 1857’, TITLV, 18 (1868–72), pp. 440, 445; ‘Berigten omtrent den Zeeroof in den NederlandschIndischen Archipel, 1858’, TITLV, 20 (1873), p. 304. 58  Statements of Domingo Apolinario and Antonio Juan in Relacion jurada de los cincos cautivos venidos en la falua de la division de la isla del corregidor (23 Aug. 1845), PNA, Piratas 3. 59  Statements in Relacion de los cuatros cautivos venidos en el Navio Ingles de Guerra Agincourt, procedente de la Isla Borneo (11 Dec. 1845), PNA, Piratas 3.

The Structure of Slavery in the Sulu Zone

1501

one; and 1858, seven.60 Women were less frequently in a position to escape, and when they did it was usually in the company of their husbands. A far easier mode of escape for banyaga in Jolo, and one which could be accomplished alone, was to swim or row out to a visiting European vessel. The usual plan was ‘to sneak alongside at night, cling hold of the chain plates, kick the canoe adrift, which they had in all probability stolen, and then make a noise until helped up the side’.61 The timing of the venture was crucial to its success. Banyaga escaped to merchant vessels only at the close of the trading season, as their chances of being granted asylum were greater when their presence was less apt to become an issue and disrupt trading.62 Despite such precautions, banyaga sometimes were turned away or given by traders—European and Chinese—to another datu as a ‘gift’.63 After the 1830s, when Spanish warships began to frequent the area, larger numbers of banyaga risked escape. In contrast to the uncertainties of escape to trading vessels, fugitives were assured protection on European warships. In 1836, more than fifty Filipino slaves from every island group in the archipelago were taken on board the vessels of the Spanish fleet that visited Jolo.64 The longer the war vessels remained in the Jolo roadstead, the more banyaga risked escape. When datus became aware that banyaga had gained freedom by setting foot on a warship, they began to take precautions. Newly acquired banyaga in particular were herded together and locked up at night, or marched into the interior until the departure of a squadron. In 1836, all small canoes were taken off the beaches and sentinels patrolled the shore after dark. Banyaga who failed in a bid to escape could be in danger of being beaten or killed, though it was more usual for them to be sold.65 60  Verklaringen van ontvlugten personen uit de handen der Zeeroovers van 1845, 1849, ANRI, Menado 37; Politick Verslag der Residentie Manado over het jaaren 1855, 1857, 1858, ANRI, Menado 166. 61  H. Keppel, A Visit to the Indian Archipelago in H.M.S. Maeander (London: Richard Bentley, 1853), p. 69. 62  See log kept aboard the brig Leonidas (entry: 31 Aug. 1836), 656/1835A, The Philips Library, Salem Peabody Museum, Salem, Mass., USA. 63  Statement of Juan Santiago in Expediente 12 (4 Oct. 1836), PNA, p. 22. 64  El Gobernador de Zamboanga a GCG, 13 Feb. 1846, PNA, Mindanao/Sulu 1836–97, p. 20; extract from the Straits Times (11 July 1848), in Governor of the Straits Settlements to the Sec. of the Government of India, 5 July 1848, EICL, F/4 2331 (121954), pp. 13–14; No.5, El Gobernador Politico y Militar de Zamboanga a GCG, 26 March 1853, AHN, Ultramar 5172. 65  Statements of Vincente Remigio, Anastacio Carillo, Ignacio Valero, Juan Apolonio, Francisco Sereno, Juan de la Cruz in Expediente 12 (4 Oct. 1836), PNA; Keppel, A Visit to the Indian Archipelago, pp. 68–9; Escosura, Memoria Sobre Filipinas y Jolo, p. 374; Statement

1502

Warren

Few fates can have been as cruelly deceptive as the one experienced by the banyaga who escaped to Zamboanga. Most banyaga who fled one form of servitude were forced to remain in Zamboanga and enter a ‘second captivity’. It was a standard practice of the Governors of the presidio to delay the return of Filipinos to their villages for years at a time in order to exploit their labour.66 Denied any opportunity to practice a trade in Zamboanga, Filipino fugitives from the Sulu archipelago were integrated into the lowest stratum of Zamboanga society as a residual source of labour power—along with criminals, deserters and deportados. These escapees were forced to labour on the fortifications, manufacture salt, collect firewood, and tend the carabao. They also toile