Greek Slave Systems in their Eastern Mediterranean Context, c.800-146 BC 9780198769941, 0198769946

Citation preview

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

G R E E K S L A V E S YS T E M S IN TH E I R E A S T E R N M E D I T E R R A N E A N C O N T E X T , C . 800– 1 4 6 B C

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Greek Slave Systems in their Eastern Mediterranean Context, c.800–146 BC D A V I D M . LE W I S

1

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © David M. Lewis 2018 The moral rights of the author have been asserted First Edition published in 2018 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2018934541 ISBN 978–0–19–876994–1 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Preface The origins of this book lie in a doctoral dissertation undertaken at the University of Durham. Prof. Edward M. Harris, who suggested to me that a comparison between Greek and Near Eastern forms of slavery might be a fruitful PhD project and acted as my doctoral supervisor, naturally stands at the head of my list of academic debts. As any of Edward’s students will confirm, his support extends well beyond the call of duty, and I have profited immensely from his teaching, advice, and remarkable generosity over the past decade. The influence of Edward’s research into slavery upon my work should be clearly visible here: the legal methodology set out in his 2002 study of debt bondage served as the inspiration for my approach in Chapter 1, whilst his 2012 study of archaic Greek slavery underpins much of Chapter 5. As well as reining in my inveterate tendency towards distraction by minor projects, he read the whole book in draft, helped eliminate many mistakes, and made many useful suggestions besides. Financial support for my doctoral research was generously provided by the Arts and Humanities Research Council. I would also like to thank Prof. George Boys-Stones for finding me much-needed funding towards the start of my research. Prof. Nino Luraghi and Dr Paola Ceccarelli examined the thesis; my warm thanks to both of them for their valuable advice on restructuring and amending it for publication as a monograph. The focus on the slave supply in Chapter 14 in particular grew out of one of Nino’s suggestions; naturally, he should not be held accountable for its final shape. An early career fellowship at the University of Edinburgh (2013–15) provided the funds and time to develop portions of the book that were not adequately treated in the thesis. Dr Ulrike Roth proved an excellent mentor and has read and commented on large portions of this book; the working group on ancient slavery that she set up for staff and students also provided an invaluable sounding board for a variety of ideas. Of that group, I have learned a great deal in particular from Dr Sara Zanovello. I should also like to thank Prof. Douglas Cairns and Dr Juan Lewis (to my lasting regret, not an immediate relative) for providing advice on the theoretical backbone of ideas that I develop in Chapter 2. Douglas also read over the Appendix, and helped to improve it in several respects. This book was completed during a semester of institutional leave granted by the University of Nottingham, which provided invaluable breathing space at a time when I needed it most.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

vi

Preface

I owe a great deal to my history teachers at school and university, in particular Edward Brittain, Rodney Jones, Joe Wilson, Carol Hewitt, and Prof. P. J. Rhodes. Here too I should mention Prof. Robert Hayward, whose Hebrew classes were splendid fun as well as rigorous and illuminating. Of my fellow students at Durham, two in particular must be mentioned: Donald Murray, whose infectious love of the ancient Near East spilled over into many fondly remembered conversations in a variety of pubs; and Mirko Canevaro: one could not find a finer interlocutor with whom to discuss Greek history. My time in Edinburgh alongside Mirko between 2013 and 2016 (the latter year as a Leverhulme Early Career Fellow) was a pleasure; he has read large chunks of the book and constantly offered advice and encouragement. Other classical scholars generously gave of their time to comment on sections of this book, and helped to improve it: Ben Clapperton, Dr Jakub Filonik, Prof. Amy Richlin, Prof. Alain Bresson, Dr Kostas Vlassopoulos, Prof. Stephen Hodkinson, Dr Geoffrey Kron, Prof. Sylvian Fachard, and Dr Maeve McHugh. Despite our several disagreements (see Chapter 2), Prof. Rachel Zelnick-Abramovitz graciously read and commented on this part of the book, and proposed some changes that have improved the analysis. Prof. Noel Lenski kindly shared some forthcoming work, namely two chapters that he penned for his co-edited volume with Catherine Cameron, What is a Slave Society? (Cambridge: Cambridge University Press, 2018); it was gratifying to see that, though working separately, we were thinking along similar lines on several issues. I have never mastered the art of giving colourful titles to anything I have written, but Prof. Patrick Finglass convinced me to tweak this book’s title to its current one, more accurate and concise than its rather prolix predecessor. Finally, the three anonymous readers for Oxford University Press provided a number of useful suggestions, for which I owe them thanks. Aspects of my research that found their way into this book were presented in papers at a variety of locations over the past decade: Athens, Warsaw, Berlin, Philadelphia, Boulder, Princeton, San Francisco, Durham, Nottingham, Exeter, Edinburgh, Liverpool, Stevenage, and Oxford. My thanks for the various invitations to speak, and to the audiences for their questions, observations, and criticisms. I would like to thank the staff at the university libraries at Durham, Edinburgh, and Nottingham; in particular, my thanks to their interlibrary loans departments for their speed, as well as their forbearance of my many requests for obscure works. I am also grateful to Prof. J. Deissler for allowing me to consult several unpublished German translations of Russian articles on slavery produced by the Forschungen zur Antiken Sklaverei at the Akademie der Wissenschaften und der Literatur, Mainz (and to quote from one on p. 297 n. 9), and to Jason Porter for facilitating access to these. I hope that the reader will indulge me regarding the breadth of this study, which is perhaps overly ambitious for a first book. The scholar who engages in

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Preface

vii

cross-cultural comparison, stepping beyond the boundaries of their own field, is something like a guest in a stranger’s house; and potentially an unwelcome one at that. That is not to denigrate the exercise of cross-cultural comparison itself, but to recognize that an etiquette must be observed, and that to avoid epistemic trespassing some acknowledgement must be made of the limits of one’s specialist knowledge.¹ I am very far indeed from having mastered the individual subjects treated in the regional studies in Part III of the book, and do not pretend to true expertise in these fields; where I engage in debate, it is largely to side with the views of one expert against another. This section of the book is thoroughly indebted to the work of many scholars of the ancient Near East (though I have consistently aimed to get as close to the evidence as possible; and have exercised some methodological judgements of a more general sort that any historian should be well placed to make). One possible solution to the problem of the book’s breadth of scope could have been the involvement of collaborators; but on reflection I thought it better for a single mind to compass the subject, rather than a number of scholars with potentially discordant approaches. That said, I have been fortunate enough to have drawn on the advice and feedback of two of the finest scholars in this field: Prof. Heather Baker and Dr Cornelia Wunsch kindly read Chapters 10 and 11 respectively and saved me from various gaffes (naturally, neither of them is implicated in the book’s overall thesis). Furthermore, the third reader of the volume for Oxford University Press, a Near Eastern specialist, provided valuable feedback on Chapter 9. My overall aim has not been to break new ground in the study of Near Eastern slavery, a task for which others are better qualified, but to bring the findings of the last few generations of orientalists into closer dialogue with my own field, Greek history. Charlotte Loveridge and Georgina Leighton at Oxford University Press provided support along the way as I turned my research into a monograph. My thanks for their help, and for their forbearance of my stated deadlines, which, like many Scottish mountains, turned out to have a string of further summits lurking beyond the initial pinnacle. I would also like to thank Suryajeet Mullick from the production team and Donald Watt for his meticulous copy-editing of the manuscript. As for personal debts owed beyond my academic colleagues, I would need another two pages to list them fully, and hope my other friends and teachers don’t feel put out by the omission of a full roster. But I would like to mention my family, to which this book is dedicated with affection: the Tonies, Shaunmol, Daniel, and Ben; and especially my mother and father. Thank you for your support and patience. *

* *

¹ The attitude expressed in Davies (2015): 301–2, it seems to me, provides a model approach in this regard.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

viii

Preface

Translations of Greek texts are my own unless otherwise noted; for Hebrew texts I have used the JPS Tanakh with some of my own modifications. Ancient sources are referenced as per OCD⁴. Regarding the spelling of Semitic words, I have followed the system of Lawrence (1935): 25–6. DML Nottingham November 2017

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Contents List of Abbreviations

xi

Introduction and Brief History of the Issue

1

PART I. PROLEGOMENA 1. Ownership and the Articulation of Slave Status in Greek and Near Eastern Legal Practice

25

2. The Riddle of Freedom

57

3. Status Distinctions in Greece and the Ancient Near East

81

4. Slave Societies, Societies with Slaves: Capturing the Relative Importance of Slavery to Ancient Economies

93

PART II. EPICHORIC SLAVE SYSTEMS OF THE GREEK WORLD 5. The Archaic Greek World

107

6. Helotic Slavery in Classical Sparta

125

7. Classical Crete

147

8. Classical Attica

167

PART III. SLAVE SYSTEMS OF THE WIDER EASTERN MEDITERRANEAN WORLD 9. Iron Age II Israel

199

10. Assyria: The Eighth–Seventh Centuries BC

223

11. Babylonia: The Seventh–Fifth Centuries BC

235

12. The Persian Empire

247

13. Punic Carthage

259

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

x

Contents PART IV. WHY SLAVERY?

14. Differentials in the Magnitude of Slaveholding: Towards an Understanding of Regional Variation

269

Appendix. The Meaning of oiketes in Classical Greek

295

Bibliography Index locorum General Index

307 351 362

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

List of Abbreviations BDB

Brown, Driver and Briggs (1906)

BE 8

Clay (1908)

BMCR

Bryn Mawr Classical Review

CAD

The Chicago Assyrian Dictionary (21 vols, 1956–2010)

Camb.

Strassmeier (1890a)

CIS

Corpus Inscriptionum Semiticarum

C.Ptol.Sklav. Scholl (1990) Cyr.

Strassmeier (1890b)

Dar.

Strassmeier (1897)

D-K

Diels and Kranz (1952)

FGrHist

Jacoby (1923–58)

FHG

Müller (1841–70)

Gigon

Gigon (1960–87)

Hell.Oxy.

Hellenica Oxyrhynchia

IC

Inscriptiones Creticae

IG

Inscriptiones Graecae

IGT

Koerner (1993)

K-A

Kassel and Austin (1983–2001)

Kock

Kock (1880)

LNB

Neo-Babylonian Laws (Roth 1995: 143–9)

LSCG

Sokolowski (1969)

LSCG Suppl. Sokolowski (1962) LSJ⁹

Liddell, Scott, and Jones (1940)

Mich.

Moore (1939)

Nbk.

Strassmeier (1889a)

Nbn.

Strassmeier (1889b)

NRVU

San Nicolò and Ungnad (1929–35)

OCD

The Oxford Classical Dictionary

PF

Hallock (1969)

Pfa

Hallock (1978)

PNA 2

The Prosopography of the Neo-Assyrian Empire (Helsinki, 1998–2011)

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

xii

List of Abbreviations

PTT

Cameron (1948)

R&O

Rhodes and Osborne (2003)

Rose

Rose (1886)

SAA V

Lanfranchi and Parpola (1990)

SAA VI

Kwasman and Parpola (1991)

SAA X

Parpola (1993)

SAA XI

Fales and Postgate (1995)

SAA XIV

Mattila (2002)

SAAB V

Fales and Jakob-Rost (1991)

SAAB IX

Deller, Fales, and Jakob-Rost (1995)

SEG

Supplementum Epigraphicum Graecum

TAD

Porten and Yardeni (1987–99)

TCL 12

Conteneau (1927)

TCL 13

Conteneau (1929)

TGF

Nauck (1889)

Thalheim

Thalheim (1913)

TLG

Thesaurus Linguae Graecae

TMH 2/3

Krückmann (1933)

WDSP

D. Gropp et al. (2001)

West

West (1989–92)

YOS 6

Dougherty (1920)

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction and Brief History of the Issue This book presents a case for radically reassessing our picture of ancient Greek slavery in its contemporary Mediterranean context. The standard view in classical scholarship holds that Greece (by which Athens is usually meant) and Rome were the only ‘genuine slave societies’ of the ancient world. Failing to attain to this dubious honour, other ancient slave systems (often labelled ‘societies with slaves’) thus recede into the background, with twofold consequences: first, a stark, black-and-white picture emerges, with Greece and Rome as apparently stand-alone exempla of slavery practised on a large scale in antiquity; second, this picture provides historians with a ready-made excuse to ignore what went on among the neighbouring ‘societies with slaves’. Greece and Rome are presented as historical rarities, ranking among only a handful of societies in which slaves made up a noticeable proportion of the overall population and whose labour underpinned the dominant position of elites. Such rarities, of course, are usually more interesting to the historian than the comparatively prosaic ‘societies with slaves’: the former require models of development and elaborate theories to explain why they evolved so differently from the norm, whereas the latter, as historically commonplace, require no explanation at all. Greece and Rome can thus be conceptually detached and studied in isolation from their broader historical contexts. The basic parameters of this approach owe much to the highly influential work of M. I. Finley, who wrote that: The pre-Greek world—the world of the Sumerians, Babylonians, Egyptians and Assyrians; and I cannot refrain from adding the Myceneans—was, in a very profound sense, a world without free men, in the sense which the west has come to understand the concept. It was equally a world in which chattel slavery played no role of any consequence. That, too, was a Greek discovery.¹

Finley here drew a sharp line in geographical and chronological terms. Geographically, a clear divide in terms of economic practices separates the classical world from that of the ancient orient. The Greeks (and then Romans) ¹ Finley (1981): 114–15.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

2

Greek Slave Systems

incorporated slavery into economic life to a high degree; Near Eastern societies, according to Finley, failed to do the same. Chronologically, these eastern societies are seen—even if only tacitly—in terms of an earlier stage of development, one that was superseded by the more advanced societies of the classical world. They are pigeonholed as the ‘pre-Greek world’. That Finley can do so implies the presupposition of a high degree of historical continuity— even stagnation—in his view of the Near Eastern world, an assumption that the Assyrians were little different in Homer’s day from what they had been a thousand years before, or that the Babylonian contemporaries of Pericles were still stranded in an earlier stage of economic organization from which the Greeks had since progressed.² In recent years, a growing number of classical scholars have expressed dissatisfaction with claims of this sort. For example, in the field of economic history the notion that the classical world and Near East were hermetically sealed zones characterized by completely different economic practices has been rejected (which is not to downplay the fact that some significant differences did exist).³ The sweeping language of ‘modes of production’ has been all but abandoned, and today’s scholars are far more interested in documenting and explaining regional economies in all their complexity (including connections with surrounding regions) than branding vast swathes of territory as characterized by this or that mode of production.⁴ Furthermore, a plethora of studies has been devoted to the cultural cross-pollination between the Greek world and the Near East: scholars now readily admit that myth, literature, science, and art in Greece all drew on Near Eastern materials and ideas.⁵ Other studies have shown that many of the institutional features of the Greek world that were once thought distinctive—such as city-state culture or democratic government—have Near Eastern antecedents.⁶ Yet the field of ancient slavery has remained relatively impermeable to these developments, with recent studies working within the same basic horizons as those of scholars fifty years ago. The recent Cambridge World History of Slavery, Volume I: The Ancient Mediterranean World is a fitting testament ² Cf. Dalley (2013): 55: Their [i.e. Vere Gordon Childe and M. I. Finley] persuasive prose suggested that there was no need to look for inventors or inventions, because pre-Greek civilisation in the Near East was static. They envisaged a society where the prestige of the elite was the sole motivation for great works and the accumulation of wealth, so that efficiency and productivity were alien concepts. Agriculture and irrigation occupied the efforts of the workforce, which had no incentive for improving technical skills, leading to a stagnant society in which financial credit had not yet evolved. ³ Davies (2001): 13–14; Morris, Saller, and Scheidel (2007): 8–9; Descat (2011): 207; Manning and Oliver (2017); cf. Bresson (2016): xxv for key differences. ⁴ Cf. Harper (2011a): 29; 147. ⁵ These developments are usefully reviewed in Vlassopoulos (2013): 1–31. ⁶ See Fleming (2004); Hansen (2006a): 7–23; Vlassopoulos (2007a).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction

3

to the degree to which Finley’s views have set the agenda for scholarship up to the present day.⁷ Although its constituent studies are, on the whole, of a very fine standard, the scope of the book is shackled to an old-fashioned paradigm that places Greece (especially Athens) and Rome at the centre of inquiry and relegates other societies to the sidelines, often not treating them at all.⁸ Even if, as a matter of fairness, we leave aside those regions of the Near East that did not adjoin the Mediterranean Sea, the focus on Greece and Rome when compared to other regions remains seriously lopsided. Of its 509 substantive pages of text, only two are devoted to Iron Age Israel (as we shall see in Chapter 9, there is rather more to say about the subject than that). Egypt receives a page and a half for its earlier periods, but somewhat more attention for the periods in which it fell under Ptolemaic and Roman rule; most remarkably, Carthage’s large slave system receives only a single footnote in the entire volume.⁹ This is not simply a function of the uneven distribution of evidence in relation to different parts of the ancient world. It represents a specific set of interests (although probably not an overt agenda) shaped by over a century of scholarship that has, in essence, become fossilized. This set of traditional interests takes centre stage, whilst other topics (for which there is often much evidence and excellent scholarly literature) receive mere bit parts. Norman Davies has highlighted a similar problem in early modern European historiography, whose practitioners have tended to focus overwhelmingly on the histories of those polities that have remained in existence to the present day, often paying scant attention to the histories of those that have died out; against this historiographical trend, his Vanished Kingdoms provides a much-needed antidote. Regarding this trend in modern historiography, he writes: Our mental maps are . . . inevitably deformed. Our brains can only form a picture from the data that circulates at any given time; and the available data is created by present-day powers, by prevailing fashions and by accepted wisdom. If we continue to neglect other areas of the past, the blank spaces in our minds are ⁷ Bradley and Cartledge (2011). In 1987 Momigliano wrote that ‘one is led to believe that at least in the near future the debate will continue on the lines indicated by Finley’ (Momigliano 1987: 5). As is often the case, Momigliano was correct. Of Finley’s three main areas of study (archaic Greece, the ancient economy, and slavery), the first two have seen extensive revision: see, e.g., van Wees (1992) and Bresson (2016) for a sense of how scholars have engaged with and moved beyond Finley’s work. But there has simply not been any major, book-length revisionist engagement with Finley’s ideas on Greek slavery. This volume aims in part to fill that gap; and my focus on Finley is due above all to an appreciation of the enormous influence of his work. One cannot simply declare that, thirty years on, we have moved past Finley’s ideas: we owe it to him, should we wish to construct a different view, to engage critically with his work. ⁸ For a judicious assessment of this volume, see Vlassopoulos (2012a). As he rightly notes (2012a: 879) ‘the volume’s panorama illustrates better than ever the limits of the paradigm within which the study of ancient slavery has been operating in the last fifty years’. ⁹ Bradley and Cartledge (2011): 14 n. 57. The volume does include a short chapter on the Near East by D. C. Snell, but the amount of material Snell had to cover in so small a space allowed for only superficial coverage.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

4

Greek Slave Systems reinforced, and we pile more and more knowledge into those compartments of which we are already aware. Partial knowledge becomes ever more partial, and ignorance becomes self-perpetuating.¹⁰

The time is now ripe for a change of perspective. Whilst much important work has been written on Greek slavery, this work has, like that which Davies criticizes in early modern European history, tended to enrich certain compartmentalized fields without affecting the overall shape of the paradigm. One aim of this book is simply to present a great deal of the evidence and scholarship relating to non-Greek forms of ancient slavery, in an effort to stimulate the kind of data circulation to which Davies refers.¹¹ The present book is not, however, intended as a comprehensive comparative study of every aspect of slavery in the Greek world and the ancient Near East, nor is it designed as a wholesale substitution for the traditional view sketched above. (Much of that work that I would define as ‘traditional’ is admirable and valuable, and it is for the most part not its content that requires replacement, so much as the larger framework within which it is conceptualized.) As for comparison between Greece and the Near East, there is much more work to be done on this issue than can be achieved in one volume, and what is presented here is merely aimed at addressing several of what I take to be the most pressing issues, as well as presenting an alternative view that stands outside the dominant paradigm. For example, I have chosen not to cover every last region conceivable, but to focus on specific case studies that are particularly illuminating from a comparative viewpoint. I ask the reader to take this selection of case studies more in the spirit of exploratory test pits than a comprehensive excavation. Likewise, the differing length of the chapters at times reflects the amount of available material; but the space devoted to different problems more generally reflects the state of these questions in current research: where good treatments by other scholars are available, I have aimed at being succinct. Where they are not, I have devoted somewhat more space to specific problems. So whilst there is sufficient evidence to write several books on the role of slavery in the economy of classical Attica, I have chosen to focus on two key areas: the use of slave labour by the elite, and the debate over slavery in sub-elite farming. I hope that the reader will forgive any unevenness. Another choice I have made was not to extend the chronological sweep of the book backwards beyond 1000 BC, partially because of reasons of space, and partially because I wish to focus on societies that are roughly contemporary with Greek societies of the archaic, classical, and Hellenistic periods.¹² My analysis also focuses largely on what we might call the ‘private sector’,

¹⁰ Davies (2011): 4. ¹¹ Cf. Harper (2011a): 18–19. ¹² There is certainly much of interest in this earlier period: see, e.g., Tenney (2011); Seri (2013).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction

5

especially the elite (thus ignoring royal and institutional slaveholdings). I have made this choice in part because of the need for brevity, but mainly because the traditional view of classical scholarship from the past fifty years has based so much of its case for Graeco-Roman exceptionalism on the importance of slavery to elites. This view can and should be shown to be erroneous on its own terms. So, whilst there remains much to do, I offer this volume as a first step towards a more integrated approach to ancient slavery. I divide the book into four parts. Part I (Prolegomena) sets out the tools required for the comparative work undertaken in Parts II and III. To embark on an economic comparison of the sort pursued in the subsequent two parts of the book, we must first establish that we are comparing the same basic phenomenon, though embedded in different socio-economic milieux: in other words, we must be sure that when we speak of slavery in this or that ancient society, we refer to fundamentally the same institution. Much work published over the past half century has, however, made such a comparison exceedingly (and perhaps unnecessarily) complicated. First, an influential strand of sociological research has attacked the legal dimension of slavery and questioned its centrality to the institution; but much of this has been based on a flawed understanding of the nature of the problem, especially in terms of a basic misunderstanding of what the terms ‘property’ and ‘ownership’ mean in a concrete and comparative sense. Second, cultural differences between Greece and the Near East relative to the value of ‘freedom’ have (rightly) been noted and much commented on, but the exact nature of the differences concerned and the practical consequences they entail have been seriously misconstrued. Part I, then, is dedicated to a more precise articulation of (i) the legal similarities and differences relative to slave status in Greek and Near Eastern law; (ii) the cultural, social, and economic consequences of differing conceptions of ‘freedom’ between the two regions; and (iii) the issue of status distinctions viewed in comparative perspective. The principal focus of this book lies in studying the economic role of slave labour in ancient Greek societies. Problems with the existing orthodox views fall in two areas: first, the role and evolution of slave labour within the societies of the Greek world and, second, the alleged uniqueness of Greek systems of slave labour when viewed in the broader context of the Eastern Mediterranean and Near Eastern world (these problems are treated consecutively in Parts II and III of this book). Part II focuses on the Greek world. In terms of the evolution of slavery in ancient Greece, most scholars hold that the progress (if we can call it that) from ‘societies with slaves’ to fully fledged ‘slave societies’ occurred in the sixth century BC. Whilst a variety of scenarios have been proposed to explain this change, few scholars have questioned the view championed by Eduard Meyer (and later by Finley) that slavery played a relatively unimportant role in the economy of the Homeric period. As we shall see in Chapter 5, there are

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

6

Greek Slave Systems

compelling reasons to abandon this view and to propose an alternative picture of the development of Greek slavery between the early archaic period and the fifth century BC. As for the role of slave labour in classical Greece, many scholars still cling to the belief that slavery barely existed in regions such as Sparta and Crete, where dependent labour is commonly described in terms of serfdom. This taxonomic choice has meant that in some general studies of Greek slavery, Sparta, Crete, Thessaly, and certain other regions are ignored altogether. As Chapters 6 and 7 will show, there are cogent reasons to reject this view: whilst dependent labour was structured in a very different manner in these regions from that of their Athenian counterparts (whose slave system is analysed in Chapter 8), Sparta’s helots and Crete’s dependent populations were comprised of privately owned slaves, and it is Sparta, not Athens, that should be seen as the most extreme example of a ‘slave society’ in classical Greece.¹³ Part III of the book sets these Greek regional studies in a broader geographical context and seeks to question the validity of Finley’s view (quoted above, and central to the basic research paradigm that has informed the past halfcentury of investigation) that the slaveholding practices of the Greeks were wholly different from those of their eastern neighbours. It treats a series of case studies in what I have (rather loosely) termed the ‘Eastern Mediterranean world’. Chapter 9 studies one of the most methodologically challenging ancient Mediterranean societies, Israel during the Iron Age II period (that is, the period of the monarchies in Israel and Judah). Here, I argue that the legal understanding of slavery in ancient Israel was substantively similar to that of the Greeks and Romans, and that the famous ‘slave laws’ of the Torah do not adumbrate a special Israelite slave status, but rather aim to create a form of indentured labour prescribed for Israelites that was designed to prevent vulnerable members of the ethnic Israelite community from becoming slaves. I also present a case for viewing slavery as an important source of wealth for elites in Israel and Judah during the monarchical period. Chapters 10 and 11 turn towards Mesopotamia and engage with recent research on slavery during the first millennium BC. The former chapter examines slavery in the archival sources of eighth–seventh century Assyria and shows that Assyrian elites were heavily involved in slaveholding on a scale that was on a par with, if not greater than, many Greeks of the classical era. The latter examines the role of slaves in the economy of Babylonia during the Neo-Babylonian and Persian periods, and shows the contribution of slaves to ¹³ I should have liked to include a further chapter on the Thessalian penestai, but that would have required expanding the scope of the book too far. On this topic, Ducat (1994) presents a characteristically detailed study with close attention to textual matters. More remains to be said, however, on the origins and economic praxis of the system (to which one might apply the approaches of Luraghi 2002 and Hodkinson 2008 respectively), as well as integrating it with other aspects of the socio-economic and political history of Thessaly.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction

7

the activities of Babylonian entrepreneurial families, whose diversified economic activities relied heavily on slave agents in the brokering of business deals and the organization of commercial activities. In both Assyria and Babylonia, however, slavery did not develop so far that we might call them ‘slave societies’; and this lesser economic importance of slavery requires explanation, a task that is treated in part in these chapters and in part in Chapter 14. Chapter 12 analyses several of the far-scattered regions of the Persian Empire in which slave labour is visible as an economically significant venture: Anatolia (as described in several classical sources), Egypt (where the letters of the fifth-century Satrap Aršama provide important evidence for slavery on elite estates), and Fars, the Persian heartland, where the numerous Persepolis Fortification Tablets—along with several key classical texts—attest to a large population of forced labourers, a mixed workforce whose statuses (as we shall see) comprised a combination of enslaved war captives and corvée labourers. Chapter 13 is an exception to the other societies studied in Part III of this book insofar as it studies a slave system located in the western Mediterranean, that of Carthage; but this civilization was firmly eastern in its origins and culture, and its system of slave labour was organized on a massive scale. Part IV contains a concluding chapter that draws together the numerous strands spun in Parts II and III, and attempts to make some progress towards a general framework for understanding why slavery became a significant economic force in some ancient Mediterranean regional economies and not others. Finally, I include an appendix on the word oiketes in classical Greek, for two reasons: first, it is routinely mistranslated by Hellenists, most of a far finer calibre than myself, because of basic errors made in the entries to our standard lexicons; and second, its correct translation affects two important debates on slavery in classical Attica, one on farming, the other on demography, with wideranging implications for how we understand the fabric of Athenian society.

R E G I O N A L P E R S PE C T I V E S I owe the term ‘slave system’ used in this book to Westermann (1955), though he did not define quite what he meant by the designation. The term is specifically defined in the introduction to a valuable recent volume edited by C. Katsari and E. Dal Lago, but it seems to me that their definition is essentially a variant of the term ‘slave society’ with some minor alterations, a cosmetic rebranding of a well-worn category.¹⁴ The term as I employ it here has a different sense and a different purpose. It is not used as an index of the ¹⁴ See Dal Lago and Katsari (2008a): 3–5.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

8

Greek Slave Systems

importance of slavery to a given society economically, culturally, or otherwise; it covers what Finley would have called ‘slaveholding societies’ as well as ‘slave societies’. My usage is regional and essentially legal in orientation. Recent scholarship has vindicated the utility of utilizing regional units of analysis in the study of ancient societies.¹⁵ In the case of slavery, we must remember that the classical Greek world comprised over a thousand poleis, and these polities each created their own laws tailored to their own needs, relating inter alia to the institution of slavery. The legal contours of slavery in the societies of the Greek world, though similar in some fundamental respects, differ in their details from polis to polis and from region to region; these differences in detail can often be explained in terms of local variables and concerns. When we compare Attic slavery to slavery in Sparta or Crete, a keen understanding of local conditions in these individual regions helps to explain what otherwise might be seen as peculiarities or eccentricities in their slave laws. As this study will show, in most cases these ‘eccentricities’ are nothing of the sort, but pragmatic responses to local, region-specific imperatives. As a comparative unit of analysis, then, the notion of ‘slave system’ allows us to treat as a single unit the legal manifestation of slavery in a given politicaljuridical region. It has the heuristic virtue of allowing us to explain slave law and the organization of slave labour in terms of local concerns, and the resulting picture can easily be held up next to other slave systems in order for the reader to discern similarities and differences in practice. In Part II of this book especially, I aim to work towards a view of Greek slavery as a patchwork of epichoric slave systems, viewing our evidence for slavery as embedded in distinctive regional contexts, rather than settling for the standard ‘serfdom’ versus ‘chattel slavery’ dichotomy common in many studies. Above all, this category helps us to move beyond crude generalizations such as ‘Greek society’ or ‘Near Eastern society’. Of course, I make no claim for external reality for this category; it is just as much a modern tool of analysis as ‘slave society’: it is nomos, not physis; and an etic, not emic, tool at that. Nor do I seek to supplant the category of ‘slave society’ with my own one: both are heuristic tools designed for different tasks, both can be used profitably, and neither can reveal the full picture in and of itself.

A N OTE ON TERMINOLOGY This book does not utilize several common scholarly terms employed in many studies of slavery, and this departure from the norm requires justification. ¹⁵ e.g. Horden and Purcell (2000); Vlassopoulos (2011b).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction

9

Perhaps the most notable omission is ‘chattel slavery’. In the context of the Greek world, this term is normally restricted to the commercialized slave systems of the Aegean, such as those of Corinth, Chios, Aegina, and (especially) Athens. These systems relied upon extensive trade networks to import non-Greek slaves in great numbers, particularly from Thrace, the Black Sea, Anatolia, and Syria. The term ‘chattel slavery’ is generally denied to the slave systems of Sparta, Thessaly, and Crete, and only sometimes applied to the Homeric world. But this usage is quite misleading. In Continental scholarship, more appropriate terms are employed to describe slavery on the Attic model: the French equivalent of ‘chattel slavery’ is esclavage marchandise, the German equivalent (though somewhat less common) Kaufsklaverei. As these terms rightly imply, the distinctive attribute of such systems lies in the fact that they drew upon commerce for their supply of slaves. Yet the term ‘chattel slavery’ represents something quite different: ‘chattel’ is a common law term for movable property, and is used to make a distinction between movable and real property, roughly along the same lines as that between res mobiles and res immobiles in Roman law. The term then means something like ‘property slavery’. Two basic problems with the term ‘chattel slavery’ thus emerge. First, like some other popular academic expressions (e.g. ‘lived experience’), it is a pleonasm. Slaves are by their very definition property, and there is no need to add a modifier to hammer the point home. Second, slaves in those Greek slave systems usually denied the label ‘chattel slavery’ were (as we shall see in Chapters 6 and 7) property just as much as Athenian slaves, though they were obtained via different means and governed by different laws. It makes no sense to deny them a term that simply means ‘property’ when most scholars only use the term to mean ‘commercially sourced’. One could try to coin a neologistic term closer to the French or German equivalents—for example, ‘commercial slavery’—but this raises problems of its own. As Jean Allain and Kevin Bales have pointed out in relation to the term ‘debt slavery’, the qualifying term ‘debt’ does not denote a qualitatively different sort of slavery, but merely the manner by which the enslaved person gained his or her status.¹⁶ A distinction must be made, therefore, between the routes into slavery and the condition itself. I rather doubt the utility of extending this principle to other supply strategies, and clogging up academic discourse with terms like ‘birth slavery’, ‘judicial slavery’, ‘piracy slavery’, and so on. It is much more useful to stick to the term ‘slavery’ pure and simple.

¹⁶ Allain and Bales (2012): 6: ‘put simply, it is illogical to name the mechanism of acquisition of a person as an essential component in defining whether a person is in slavery. Slavery is a status or condition, not the means by which a person is removed into that state or condition of control.’

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

10

Greek Slave Systems

The ‘debt slavery’ example highlights another common problem, and that is the conflation of debt bondage with enslavement for debt. The two conditions are qualitatively quite different. The first term describes a form of indentured service in which the bondsman works to pay off an obligation. His or her labour continues only until the debt has been repaid, and the master is not their owner.¹⁷ Enslavement for debt, however, transforms the debtor into a slave, who can be bought, sold, inherited, and so on; and the creditor who enslaves him upon default becomes his owner.¹⁸ As we shall see in Chapter 9, the ‘slave laws’ of the Torah in the Hebrew Bible begin with a standard scenario of enslavement due to debt or poverty, and then proceed to alter that person’s condition radically, limiting it to temporary indentured service. Therefore, I have also avoided the term ‘manumission’ when describing the rules on release from this condition of indentured service, though most biblical scholars do use the term. My difference from them lies not in disagreement over the substance of the condition but on the appropriate vocabulary to describe it. Since the phenomenon of enslavement for debt and debt bondage is not limited to the Near East, but also appears in an important set of laws from Gortyn on Crete as well as in several texts from or concerning Athens, the need to find suitable terminology in common requires an adjustment of the vocabulary used in narrower scholarly fields. In other words, when engaging in comparative research, a cross-cultural lingua franca in terms of etic descriptors is indispensable. A third omission is the term ‘unfree’, deriving from the abstract noun Unfreiheit in German. The adjective ‘unfree’ is usually employed to describe a broader set of dependent relationships than slavery sensu stricto, but its usage implies a concrete meaning of freedom that its users seldom, if ever, define. I do not find it ‘admirably vague’, and prefer the term ‘dependent labour’ as the genus of which slavery is a species.¹⁹ (It is perhaps more legitimate in studies of medieval slavery, where the semantic range of the terminology of freedom shifted from its antique usage.) As we will see in Chapter 2, a great deal of (unnecessary) confusion has been created by scholars failing to be specific about what exactly it is that they mean when they employ terms such as ‘free’ and ‘freedom’, and it is all too common that vagueness in understanding or evasiveness of expression is passed off as nuance. ¹⁷ Cf. Dandamaev (1984): 80, rightly noting that although many scholars use the term ‘debt slavery’ when talking about debt bondage, ‘the debtor, who only worked temporarily for the creditor, can hardly be considered a slave’. ¹⁸ That is, at least, the legal distinction; I do not wish to deny that in practice, sometimes debt bondsmen were treated as slaves in regions where the legal institutions that prescribed debt bondage were inadequate to enforce their rules. For the formal distinction between the two conditions, see Harris (2002a). ¹⁹ I borrow this description from Finley (1968): 308: ‘Slavery is a species of dependent labor and not the genus.’

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction

11

Finally, readers will notice that I have avoided using the term ‘serfdom’ to describe helotage and other comparable forms of slavery. Even though some aspects of helot life did approximate to the social condition of serfs, from a legal perspective this terminology is misleading. It engenders the belief that these individuals were ‘bound to the soil’ and possessed legal rights, a view that has little evidentiary basis; furthermore, scholars who use this terminology routinely ignore or attempt to explain away good evidence that these persons could be sold. Often scholars fudge together the economic and the legal, concluding that because this or that enslaved group lived in family groups in the countryside, they must have been serfs, not slaves. This is to ignore the institutional, legal basis of different varieties of forced labour. Slaves are not distinct from other forced labourers because they work in some ‘plantation mode of production’, but because a society’s laws grant their masters the rights of ownership over them. From a legal point of view, a plantation slave is no more or less a slave than one who works in a bank or a mine.²⁰ I will attempt to justify my obstinacy in avoiding the term ‘serfdom’ in more detail in Chapters 6 and 7.

A B R I E F H I S T O R Y OF TH E I S S U E We noted above that Finley’s views have played an important role in shaping the contours of modern inquiry into ancient Greek slavery. But Finley did not develop these views in a vacuum, and it will, I hope, prove useful to take a brief look at scholarship of the nineteenth and twentieth centuries. This will help us to understand why the standard view of the problem that exists today has taken the particular shape that it has. (Here, I focus on the core question of the ‘uniqueness’ of Greek slavery, and not on some of the subjects explored in Part II.²¹) One could begin an overview of this subject earlier, but for present purposes it will suffice to begin with Henri Wallon, whose three-volume Histoire de l’esclavage dans l’antiquité (1847; second edition 1879) stands as an important early landmark in the field.²² This work included a sixty-one-page chapter on ²⁰ That said, I should state that I have no objection to the idea of serfdom as a cross-cultural category, as argued for by de Ste. Croix (1981) and, more recently, Hunt (2016); my difference of viewpoint from these scholars arises from a different interpretation of the evidence for helot status, rather than any schism over theory. ²¹ I can provide only a very cursory review of this issue here, and what follows is very far from being anything like a detailed intellectual history (though such a history would be well worth writing). ²² This was not the first work on ancient slavery; see Finley (1980): 11–40 for a survey of earlier studies. However, it represents a good starting point for tracing the shape of the field as it developed into and during the twentieth century.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

12

Greek Slave Systems

slavery in the ancient Orient that prefaced his subsequent chapters on Greece and Rome. We should recall that cuneiform script was still at this time poorly understood, so Wallon’s focus fell mainly on the Hebrew Scriptures. It is also important to emphasize the moral and political aims of Wallon’s study, formulated at a time when the abolition of slavery in the French colonies remained a key political issue.²³ A deeply religious individual, Wallon aimed to show the incompatibility of slavery with Christian values; and to Wallon’s mind, the Hebrew Bible contained a kernel of abolitionist sentiment: in particular, we may note those laws of the Torah that enjoined good Hebrews to treat their Hebrew slaves well and release them after a brief spell of service.²⁴ This apparently protoabolitionist attitude stood in contrast with the brutal pagan slave systems of Greece and Rome. For Wallon, then, the contrast between Hebrew and GraecoRoman slavery lay chiefly in the moral esprit suffusing each. For the articulation of a proper economic comparison of Greek and Near Eastern slavery we must turn to the work of the German ancient historian Eduard Meyer. Meyer’s famous lecture on slavery (later published as ‘Die Sklaverei im Altertum’) was delivered in Dresden on 15 January 1898.²⁵ One of his chief points in this lecture lay in emphasizing the cyclical nature of history: as the leading proponent of the nascent ‘modernist’ school on the ancient economy, Meyer heavily criticized the ‘primitivist’ view of Karl Bücher and the National Economists that history comprised a series of stages, each leading progressively to modernity, in which antiquity was understood as more economically undeveloped than the medieval period.²⁶ Instead of a simple progressive march from the self-sufficient household economy (Hauswirtschaft) of antiquity to the bourgeois capitalist economy of the nineteenth century, Meyer emphasized rather the ebbs, flows, and cycles of history. In his view, different societies entered different stages of development at certain points in the past, and comparable economic structures might resurface at a later point in time. History thus knew regression as well as progress. According to Meyer, the Homeric period was directly comparable to the European Middle Ages, whereas the heyday of antiquity (‘die Blütezeit des Altertums’), that is, classical

²³ See Wallon (1879): v–clxiv. ²⁴ We will examine the Hebrew material in much greater detail in Chapter 9. Wallon rightly noticed that the laws of the Torah prescribed kindly treatment and release for Israelites alone, and thus not foreigners; but he conceived an ingenious (if wrongheaded) argument to extend this treatment to gentiles as well: since the law at Gen. 17: 12–13 required all male members of an Israelite household (including slaves) to be circumcised, this must have made any slave held by an Israelite, whether a fellow countryman or foreigner, qualify for the social justice legislation described in Exodus, Deuteronomy, and Leviticus. See Wallon (1879): 11–12. It is perhaps a sign of the times that Wallon failed to notice that his argument would not work for foreign female slaves, as female circumcision is unknown in this culture. ²⁵ See Meyer (1910): 171–212. ²⁶ On the Bücher-Meyer debate, see Badian (1981): 52–3; Davies (1998): 233–5; Bresson (2016): 2–5.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction

13

Greece and Late Republican/Early Imperial Rome, corresponded to modernity (i.e. the Early Modern period).²⁷ As was the case for Wallon, the fledgling state of research into cuneiform sources restricted much of Meyer’s view of the east to the Hebrew Scriptures and classical texts. He asserted that ancient Israel was comparable in terms of socio-economic structure to Homeric Greece and archaic Rome: in all three societies, slavery played no real role of importance; a system of feudal serfdom (Hörigkeit), rather than a system of slavery, underpinned the economy.²⁸ Furthermore, it is indicative of the strength of the idea that, despite his conservative politics, Meyer held basically the same view as Marx on the alleged historical stagnancy of the Orient, a view that was instrumental to Marx’s theory of the so-called ‘Asiatic Mode of Production’ (which we will discuss in greater detail in the last section of this Introduction, pp. 15–17).²⁹ Meyer’s essay remained influential in twentieth-century Germany, and it was praised by no less than Joseph Vogt, the doyen of post-WWII German scholarship on classical slavery.³⁰ Its influence also lived on in the AngloAmerican sphere: the American historian W. L. Westermann, who penned the supplement on slavery (published in 1935) for the Pauly-Wissowa Real Encyclopädie, wrote that Meyer’s essay laid the foundations of the contemporary understanding of Greek and Roman slavery.³¹ Westermann’s study was later published in English, with (to the chagrin of his student M. I. Finley) few changes from the earlier German encyclopedia supplement, as The Slave Systems of Greek and Roman Antiquity (1955).³² Some of the main tenets of the current orthodox view of the ‘uniqueness’ of Greece and Rome in the history of slavery were articulated in Westermann’s study. One key point was the idea that a sharp distinction between free men and slaves only existed in Greece and Rome, not in the Near East: In sharp contrast to this clarity of differentiation between those completely free and those who were enslaved, with a special term for ‘freedman’ (ἀπελεύθερος)

²⁷ Meyer (1910): 188: Die erste Epoche des Altertums, die homerische Zeit und ihre Parallelen, steht mit der ersten Epoche der christlich-germanischen Völker auf derselben Linie und verdient wie diese als Mittelalter bezeichnet zu werden; die Blütezeit des Altertums entspricht der Neuzeit, sie is wie diese nach jeder Richtung eine modern Zeit, in der die Anschauungen herrschen, die wir als modern bezeichen müssen. ²⁸ For this comparison, see Meyer (1910): 180. For his view that slavery was economically unimportant in the orient, see Meyer (1910): 190–2. For Meyer’s view of slavery in the Homeric period, see Finley (1980): 46–7; Harris (2012): 346–8. ²⁹ Meyer (1910): 189–90. At p. 190 he made the remarkable claim that ‘Damaskus, Aleppo, Hamat und die Großstädte Ägyptens und Babyloniens haben im Jahre 1000 v. Chr. nicht wesentlich anders ausgesehen, als heutzutage.’ ³⁰ Vogt (1974): 179. See Harris (2012): 348 for further discussion. ³¹ Westermann (1935): 894: ‘Der Grund zu unserem gegenwärtigen Wissen über Sklaverei in der griechischen und römischen Geschichte wurde von Eduard Meyer in Seinem Vortrag . . . gelegt.’ ³² See Finley (1980): 55.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

14

Greek Slave Systems

and regulations which provided the freedman group with legal directives, stands the blurring and overlapping of these social classifications in the Egyptian and the Semitic languages . . . the cuneiform signs for ‘slave’ meant, literally, a ‘man from the mountains’, that is, a captive from an alien land. The same indefiniteness prevailed in the early development of the Egyptian language in the use of the term b’k, to the point that the Egyptologists still disagree as to the exact status of the persons thus designated. The Hebrew word ’ebed suffers from the same vagueness, its application ranging through ‘slave’ or ‘servant of the Lord’ in the phrase ’ebed Jahwe, to the titular epithet describing a high military or civil official as ’ebed el malek, ‘servant of the king’.³³

Westermann thus pointed out an apparent qualitative difference between classical and Near Eastern forms of slavery, one that was later to be adopted by Finley. But he did not insist on a huge quantitative divide in economic terms between the classical and non-classical worlds: for one thing, he tended to play down the economic role of slavery in Greece and Rome where possible; for another, several paragraphs in his book are devoted to the enormous slave population of Carthage during the Hellenistic period, the size of which he made no effort to deny or explain away.³⁴ In a section entitled ‘Basic differences between pre-Greek and Greek slavery’ Westermann had much to say on rational Hellenic improvements that apparently brought conceptual clarity into the distinction between slaves and free men, and he also had things to say about differences in the treatment of slaves; but he had nothing to say at all on whether the Greeks and Romans made greater economic use of slaves than other ancient civilizations.³⁵ A starker divide between the two worlds was more fully set out by M. I. Finley. Although generally critical of his teacher, Finley adopted Westermann’s view of the qualitative differences between the allegedly sharp free–slave divide in the classical civilizations and the allegedly blurred state of affairs that prevailed in the Near East.³⁶ Indeed, he pushed this view further by claiming that the Greeks invented the idea of freedom and situated that development in particular historical circumstances (namely the aftermath of Solon’s reforms of 594/3 BC). Although describing Meyer’s essay on slavery as ‘a succession of ex cathedra assertions’ and ‘as close to nonsense as anything I can remember written by a historian of such eminence’, Finley essentially repeated Meyer’s view that slavery was unimportant in the Homeric period.³⁷ It was the economic transformations of the sixth century that, according to Finley, sparked the ‘rise’ of slavery in the ³³ Westermann (1955): 42–3. ³⁴ Westermann (1955): 58. ³⁵ Westermann (1955): 39–46. ³⁶ Finley (1980): 52–5. For his adoption of Westermann’s stance on the qualitative divide, see Finley (1964): 238 = (1981): 120–1, quoted verbatim in Chapter 3, p. 81 of this volume. ³⁷ Finley (1980): 47–8. For a more balanced assessment of Meyer’s contribution, see Badian (1981): 52–3. On the closeness between Finley and Meyer’s views on early Greek slavery, see the important study of Harris (2012): 346–51.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction

15

Greek world, and this was the direct cause of the development of a sharp distinction between free men and slaves. In essence, before slavery became economically prominent, all ancient societies lived in a world where statuses were blurred together in a continuum of dependency.³⁸ It was the Greeks who pioneered the large-scale exploitation of slave labour, spurring the invention of the concept of freedom, a cultural value that they bequeathed to the Romans (also large-scale exploiters of slave labour). Other ancient societies, according to Finley, failed to make this key leap. Finley’s deterministic causal link between the economic development of slavery and the development of the cultural value ‘freedom’ caused him to deny or ignore the role of slavery in non-classical cultures and forefront the importance of Greece and Rome to the history of slavery.³⁹ At this point, something needs to be said on Marx and his influence on the historiography of ancient slavery. Marx was well acquainted with the ancient world, having written his doctoral dissertation on the Democritean and Epicurean philosophies of nature at the University of Jena (accepted in 1841). In his later political and economic works he sometimes drew on classical texts to illustrate the proto-history of modern bourgeois capitalism, but this was never systematically approached or developed into a rigid theory of human societal evolution. The programmatic division of history into a series of epochs marked by different modes of production (primitive communism; the slave mode of production of antiquity; medieval feudalism; modern bourgeois capitalism; socialism) is more a relic of the twentieth century, more specifically, of Stalinism (in the USSR) and ‘Vulgar Marxism’ (elsewhere). At any rate, Marx did analyse different historical forms of property, and in the course of his inquiries formulated some conclusions about historical conditions in Asia. Like most of his contemporaries, Marx tended to view Asia as essentially uniform and unchanging.⁴⁰ His theory of a so-called ‘Asiatic ³⁸ Finley (1981): 132. ³⁹ One must be fair to Finley by situating his work in its historical context. He can be excused in part, insofar as research into slavery in the Near East at that time (exemplified by Mendelsohn 1949) largely supported his position. But this only explains so much, and it is at times difficult to justify Finley’s view of a stark contrast between the classical and non-classical worlds. For instance, Mendelsohn (1949): 110–11 mentioned that slavery was economically significant in agriculture during the Neo-Assyrian period, but I can find no reference to this inconvenient position in Finley’s works, nor can I find any discussion of de Vaux (1961), who argued that slavery was an important labour source on elite lands in Iron Age Israel. (On Finley’s avoidance of Jewish subject matter in general, see the remarks of Momigliano 1987: 2–3.) More glaring is Finley’s avoidance of Carthaginian slavery: having reviewed Westermann’s Real Encyclopädie supplement (see Finley 1936), he surely knew of at least some of the key evidence for this topic, much deriving from classical sources that he was more than capable of reading in the original. On Finley’s intellectual background and oeuvre, much has recently appeared in print: see inter alia the books of Nafissi (2005); W. V. Harris (2013); Naiden and Talbert (2014); Jew, Osborne, and Scott (2016), and the essays of E. M. Harris (2013d); Tompkins (2014a; 2014b); Lenski (2018b). ⁴⁰ The ubiquity of these views in Europe during the eighteenth–twentieth centuries is extensively documented in Said (1978), though not always in a balanced manner. For criticisms of Said’s assessment, see Irwin (2006).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

16

Greek Slave Systems

Mode of Production’ (AMP) stood outside the historical development of Europe. In its most basic terms, the model worked as follows: in an Asiatic state, the surplus production of the proletariat was appropriated by an administration and delivered to the absolute monarch, who owned all the land in his kingdom. As a result, the monarch became fabulously rich, but the working population lingered on at subsistence level. This system rendered any sense of dynamism or creativity stillborn, and explained the apparent sluggishness of Asiatic societies to European eyes.⁴¹ As with many of Marx’s ideas, this was simply provisional and tentative. But his writings on the topic exercised an important influence on scholarship on slavery during the twentieth century, nowhere more so than in the Soviet Union.⁴² Eastern Bloc scholars working on ancient slavery had to contend with two aspects of Marxist thought that under the political circumstances of the time could be simplified and turned into dogma by hardliners. First, there was the issue of the AMP. Second, there was stage theory, which became enshrined as dogma in Stalin’s Dialectical and Historical Materialism (1938). Daniel Tompkins has shown that scholars in the Eastern Bloc were far from a brainwashed mass dedicated to providing empirical proof of party dogma: there was much variation among scholars regarding the degree to which they toed the party line; and the enforcement of dogma fluctuated considerably during the history of the Soviet Union.⁴³ At any rate, China’s adoption of Communism made little sense if one were to apply mechanically the AMP theory (as an Asiatic country, after all, it should have remained historically stagnant, rather than leapfrogging the intermediate stages leading to socialism), and there was naturally some debate over whether, where, and when it might be applied to historical societies.⁴⁴ Some Soviet scholars, on the other hand, sought to apply the idea of an antique slave mode of production across the whole of the ancient world without positing an East/West divide, as Stalinist stage theory prescribed. Notable in this respect were the efforts of Vasily Struve. Since it is patently clear that slavery sensu stricto was not the sole productive base for any ancient society, Struve had to modify the definition of slavery to make Marx’s concept of an antique ‘slave mode of production’ fit the evidence. He thus defined slavery not in traditional legal terms but in terms of workers deprived of the means of production, a rather broad and vague definition that has not stood the test of time.⁴⁵ ⁴¹ On the AMP, see Thorner (1966); Anderson (1974): 462–549. ⁴² Pečírka (1964). One should note that Marx’s influence on ancient history before 1917 was fairly negligible; and the Grundrisse remained unpublished until 1939. ⁴³ Tompkins (2014a). ⁴⁴ This is well documented in Garlan (1988): 8–11. ⁴⁵ See Chapter 1, p. 27 below. Cf. Heinen (2001): 488, quoting (and translating) the Russian historian Ė. D. Frolov: So paradox das auch klingen mag, so wurde doch die Akademie-Reihe ‘Forschungen zur Geschichte der Sklaverei in der antiken Welt’ [viz. a Russian research programme]

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction

17

Soviet scholarship had a limited impact on research into ancient slavery in the Anglo-American world, apart from producing disdain in several quarters. Some historians went to special pains to keep abreast of developments in Soviet scholarship,⁴⁶ but it would be an exaggeration to say that this work set the agenda for research on slavery in the West. However, Marxism exerted an influence well beyond the USSR, and two fine and influential Marxist works by European scholars are particularly worthy of mention: Yvon Garlan’s Les Esclaves en Grèce ancienne (1982; English translation by J. Lloyd 1988) and G. E. M. de Ste. Croix’s The Class Struggle in the Ancient Greek World (1981). Both scholars considered the place of Greek slavery in a broader geographical context. Garlan, more attached to the theory of the AMP than de Ste. Croix, simply ignores the Near East as a world apart.⁴⁷ De Ste. Croix, on the other hand, rejected the AMP wholesale, and indeed argued that Marx himself had abandoned the idea by the end of his life.⁴⁸ In de Ste. Croix’s view, elites of the ancient Near East relied on forms of serfdom and debt bondage for their surplus production.⁴⁹ His treatment of one episode that we shall analyse in greater detail in Chapter 12 is telling: Xenophon (Anab. 7.8.7–22) describes the slave-worked estate of a wealthy Persian in Asia Minor, but de Ste. Croix claimed (based on his prior assumptions on serfdom and debt bondage in the Near East) that this grandee was simply copying the economic practices of his Greek neighbours.⁵⁰ Beyond the efforts of individual scholars, three international research groups on ancient slavery require our consideration. The first is the Mainz

gewissermaßen zum Grabstein der marxistischen Konzeption von der antiken Sklavenhaltergesellschaft. Sowohl diese Forschungen selber als auch die sich parallel entwickelnde Polemik mit der westlichen Wissenschaft . . . führten letztendlich die russischen Gelehrten zu der Überzeugung, dass die marxistische Vorstellung von der grundlegenden Bedeutung der Sklaverei für das klassische Altertum keine unbedingte Gültigkeit besaß, jedenfalls nicht für alle Epochen und alle Gebiete der antiken Geschichte. ⁴⁶ Germany proved to be the main conduit through which Soviet scholarship came to be known in Europe and America; in particular, the Mainz Academy commissioned translations of key Russian works into German, e.g. Lencman (1966), Blavatskaja, Golubcova, and Pavlovskaja (1972); and H. Heinen published regular summaries of Soviet research on slavery in Historia: Zeitschrift für alte Geschichte throughout the 1970s. Cf. Heinen (2010) for a retrospective view, and Heinen (2001) for an overview of (mainly Russian) work on Black Sea slavery. There were also summaries of Soviet work published in French: see Petit (1970); Pavlovskaja (1979). Y. Garlan, fortuitously married to a Russian speaker (see Garlan 1988: x), is exceptionally familiar with Russian work on slavery among European ancient Greek historians, as a glance at the footnotes of Garlan (1988) conveys. Mrs Garlan also provided translations of Russian works to M. I. Finley: see Finley (1999): 9. ⁴⁷ Garlan (1988): 5–10. Another notable believer in the AMP was Keith Hopkins: see Hopkins (1978): 99 n. 2. ⁴⁸ De Ste. Croix (1981): 155–7. ⁴⁹ De Ste. Croix (1981): 170–2. Like Finley and Westermann, de Ste. Croix relied on Mendelsohn (1949) for his view of slavery in the Near East: see de Ste. Croix (1981): 573 n. 76. ⁵⁰ De Ste. Croix (1981): 569 n. 42a.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

18

Greek Slave Systems

Academy’s research programme on ancient slavery, initiated by Joseph Vogt in 1950. Unfairly dismissed (along with its founder) by Finley as narrowly concerned with justifying ancient slavery as a price worth paying for the cultural legacies of Greece and Rome, the Forschungen have greatly contributed to our knowledge of Greek (and to a slightly greater extent, Roman) slavery.⁵¹ But although the programme was entitled Forschungen zur antiken Sklaverei, its purview has essentially been limited to Graeco-Roman material. Rather than viewing Greece in comparative perspective with non-classical cultures, members of the project working on the Greek side have tended to focus on discrete topics, such as the slave miners of Laurion, slaves in Greek religion and warfare, the papyrological evidence for slavery in the Ptolemaic period, and so on.⁵² One notable exception to this trend, though, is Gundlach’s monograph on forced population movements in Middle Kingdom Egypt;⁵³ one could also point to Heinen’s work on slavery in peripheral regions such as the Black Sea and Egypt as going beyond the traditional core Graeco-Roman focus.⁵⁴ The second research group of note is the Groupe International de Recherche sur l’Esclavage dans l’Antiquité (GIREA), whose conferences have occurred every year or two since 1970, with conference proceedings appearing in print shortly afterwards. The group began in Besançon, though its meetings have been held in a number of other locations; the proceedings in general amount to an eclectic mixture of articles of varying length. For the most part, they focus on Graeco-Roman matters. To its credit, the group has long been open to examinations of slavery in other cultures, both ancient and modern;⁵⁵ in recent years in particular, contributions to the proceedings have covered other ⁵¹ Finley (1980): 55–66. Cf. Wiedemann (2000) on Finley’s ‘grotesque misinterpretation of the aims, methods and achievements of the Mainz research group’ (p. 157) and the judicious comments of McKeown (2007): 30–41, who was a postgraduate student in Cambridge during the 1980s. Finley’s relationship with the Mainz Academy and its founder Joseph Vogt form part of a current research project on Vogt undertaken by the Edinburgh-based historian of ancient slavery Ulrike Roth. See also Tompkins (2014a). ⁵² Greek topics in the Forschungen include slaves in medical literature (Kudlien 1968); slaves in warfare (Welwei 1974; 1977); slave–master relations, as well as economic and political writings on slaves (Klees 1975); Greek terminology for slavery (Gschnitzer 1963; 1976); slave miners at Laurion (Lauffer 1979); slavery and religion (Bömer 1990); Homeric and Mycenaean slavery (Lencman 1966; Wickert-Micknat 1983); mass enslavements in warfare (Volkmann 1990); slave life (Klees 1998); Ptolemaic texts concerning slaves (Scholl 1990). For an overview of the Mainz group, see Wiedemann (2000). ⁵³ Gundlach (1994). ⁵⁴ Heinen (1978); (2001). I should also draw attention to I. Weiler’s (2004) study. A Mitarbeiter of the programme, Weiler draws several parallels between Homeric Greece and the Near East in terms of manumission, flight, ransom, and the idea of the slave as a ‘half-man’. ⁵⁵ Ancient examples include contributions on the Black Sea region (Nadel 1976); Babylonia (Dandamaev 1979); Illyria (Arciniega 1989); Thrace (Tačeva 1996); Spain (Wagner 2000). Modern examples include, inter alia, volumes incorporating essays on the Far East (Yuge and Doi 1988) and the New World (Gonzales 2008).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction

19

ancient societies, especially Egypt.⁵⁶ This diversity is most welcome, but unlike the Mainz group, which has taken a more coordinated approach and led to the publication of numerous monographs, the GIREA has lacked a concerted agenda and programme. As a result, its contributions have remained rather atomized and lacking in overall coherence. The third research group is the much more recently founded Institute for the Study of Slavery (ISOS), established by Thomas Wiedemann in 1998 and based at the University of Nottingham. To date, this programme has produced four edited volumes, which have brought together scholars working on a range of historical societies to study specific issues in relation to slave history.⁵⁷ Like the Mainz Forschungen, the ISOS publications are well organized and based around carefully chosen themes. The Institute’s emphasis on comparative history and its work on engendering dialogue between scholars of GraecoRoman antiquity and scholars of post-antique historical cultures represents an important step towards a better integration of ancient slavery into broader global studies of slavery; but its programme has not yet explored the comparative study of slave systems within the ancient world beyond Greece and Rome. In sum, then, slavery in the non-classical civilizations of the ancient world has not been wholly ignored by classical scholars; but it has never been systematically examined alongside Greek and Roman forms of slavery. Nothing better characterizes the limited horizons of modern research than the syntheses and introductions to the topic published since the millennium. A recent French introduction of note, J. Andreau and R. Descat’s Esclave en Grèce et à Rome (2006), has little to say on how Greek and Roman slavery compares to forms of slavery among their neighbours. Furthermore, the Greek portions of this book entirely sidestep regions such as Sparta, Crete, and Thessaly by characterizing their systems of dependent labour in terms of serfdom.⁵⁸ L. Schumacher’s Sklaverei in der Antike (2001) treats ‘Antike’ as commensurate with Greece and Rome, and suffers from the same drawbacks as Andreau and Descat’s synthesis in relation to helotage and comparable systems of dependent labour. And we have already noted K. Bradley and P. Cartledge’s hefty Cambridge World History of Slavery, Volume I (2011), purporting to cover the ancient Mediterranean world, but overwhelmingly focused on Graeco-Roman subject matter.⁵⁹ Does this state of affairs owe in part to stagnation in research on slavery among scholars of the ancient Near East? In fact, a great deal has occurred in this field since the mid-twentieth century. The main textbook on the topic used by most of the scholars we have noted above is Isaac Mendelsohn’s ⁵⁶ See Campagno, Gallego, and Mac Gaw (2013). ⁵⁷ Wiedemann and Gardner (2002); Kleiwegt (2006); Geary and Vlassopoulos (2009); Hodkinson and Geary (2012). ⁵⁸ Andreau and Descat (2006): 11–12. ⁵⁹ One recent step towards a more inclusive approach is Zurbach (2015), a collection of essays on agrarian labour covering a wide range of archaic Mediterranean societies.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

20

Greek Slave Systems

Slavery in the Ancient Near East (1949), excellent on legal matters, but tending to downplay the economic importance of slavery mainly on the grounds of minimal contribution to overall production. (As we shall see in Chapter 4, classical historians had abandoned this index of the importance of slavery by the 1960s, but it is still employed by some scholars of the ancient Near East.) Of chief importance is the work of the late M. A. Dandamaev, who drew attention to some of the key evidence for slavery in the Persian Empire in articles published during the 1960s and 1970s⁶⁰ and whose remarkable book Slavery in Babylonia, from Nabopolassar to Alexander the Great, 626-331 BC was translated into English and published in 1984. One reviewer, R. J. van der Spek, suggested that Finley’s views on slavery would need to be seriously reworked in light of Dandamaev’s book, but his challenge to produce a revisionist study has not been taken up in the quarter-century that has since elapsed.⁶¹ Other studies of note include Roland de Vaux’s work on slavery in Iron Age Israel (now supplemented by Hezser’s useful monograph, particularly good on the later periods, as well as a useful recent essay by A. Lemaire⁶²), the work of E. Matilla Vicente and J. G. Février on slavery at Carthage,⁶³ the recent work of K. Radner, G. Galil, and H. Baker on Assyria⁶⁴ and C. Wunsch and R. Magdalene on Babylonia during the first millennium BC,⁶⁵ not to mention the work of scholars like P. Briant and G. G. Aperghis on Persia.⁶⁶ (And that is to note merely a few prominent names working on slavery and forced labour; one could add many others whose efforts have revolutionized our broader understanding of these regional economies.) Whilst in most cases these studies do not suggest that we need to alter our picture in terms of the contribution of slavery to overall production in any given Eastern society, they all provide much important discussion on elite exploitation of slave labour that has major implications for our broader view of slavery in the ancient Mediterranean and Near Eastern world. Let us sum up. Several trends have combined to produce the current view of Graeco-Roman exceptionalism: (i) a traditional tendency in the Classics to treat Greece and Rome as fundamentally different from their neighbours, only recently subject to extensive re-examination, and which has not yet affected the field of slavery studies; (ii) limited engagement with recent work on the economic history of the ancient Near East; (iii) the effects of Vulgar Marxism and Orientalism, which have tended to ‘other’ the economic structures of the Near East to an exaggerated degree; and (iv) the profound influence exerted ⁶⁰ ⁶² ⁶³ ⁶⁴ ⁶⁵ ⁶⁶

See Dandamaev (1963; 1975). ⁶¹ Van der Spek (1990): 248. De Vaux (1961): 80–90; Hezser (2005); Lemaire (2015). Matilla Vicente (1977); Février (1951–2; 1961). Radner (1997): 219–48; Galil (2007); Baker (2016a). Wunsch and Magdalene (2012; 2014). e.g. Briant (2002): 422–48; Aperghis (2000).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Introduction

21

by the work of M. I. Finley, whose scholarship is among the first any student of ancient slavery encounters. A well-turned sentence by Finley can have as great an impact on the field as a book by most other scholars; it is no exaggeration to say that his negative remarks about the role of slavery in the East have hitherto rendered the topic a non-starter. Whilst some scholars have recognized the weaknesses of this general outlook,⁶⁷ no thorough revision of its parameters has thus far appeared. On the positive side, this state of affairs leaves much stimulating comparative work to be done, a task to which we will now turn.

⁶⁷ e.g. Foxhall (1997): 62; Vlassopoulos (2007a); Harris (2012): 364–6.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Part I Prolegomena

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

1 Ownership and the Articulation of Slave Status in Greek and Near Eastern Legal Practice In order to compare Greek and Near Eastern forms of slavery profitably, we must settle at the outset on a basic definition of slave status in its legal sense. By this I do not mean that we propose a modern piece of taxonomy into which we crowbar our ancient evidence: rather, our definition and conceptual approach must fit that evidence comfortably. Of course, any global definition of slavery must be by its very nature etic, not emic; but its success or lack thereof will be judged on the degree to which emic concepts of slavery (i.e. the indigenous concepts of slavery of the Greeks, Babylonians, etc.) align with this broader comparative framework.¹ Theory, therefore, must be in full dialogue with the empirical realities reflected in our sources. Traditionally, slavery in its legal sense has been defined in terms of property and ownership: the slave is an article of property (the object of the relationship) that is subject to the ownership (the relationship itself ) of his or her master (the subject of the relationship).² This definition has a distinguished pedigree reaching back to antiquity: Aristotle (Politics 1253b32) defined the slave as an article of property imbued with a soul (ὁ δοῦλος κτῆμά τι ἔμψυχον); and Florentinus in the Digest (1.5.4.1) states that slavery is ‘an institution of the common law of peoples by which a person is put into the ownership of somebody else, contrary to nature’ (servitus est constitutio iuris gentium, qua quis dominio alieno contra naturam subicitur). In his fifteenth Discourse, Dio Chrysostom presents a dialogue in which multiple persons discuss the various shades of meaning associated with the term ‘slavery’. When his disputants

¹ On emic and etic categories, see Chapter 2, section I, pp. 58–9 below for more detailed discussion. ² It is true that the terminology of slavery in ancient Greek, as well as in ancient Near Eastern languages, had a broad semantic range that went beyond legal slavery. This theme will be taken up in Chapter 2; for now, we will restrict our horizons to slavery as a technical legal status.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

26

Greek Slave Systems

come to the basic, legal sense of the word, though, they all agree with one another (15.24): It seemed to them that when someone gains power over a person through acquisition, just as he might for any of his property or cattle so that it is possible for him to use the person as he wishes, then the person might correctly be said to be the slave of the one who acquired him.

As for modern definitions on this theme, perhaps the most well known is that promulgated by the League of Nations in 1926, which defined slavery as ‘the status or condition of a person over whom any or all of the powers attaching to the right of ownership are exercised’.³ Since the latter half of the twentieth century this formulation has come under criticism on several counts. Some have objected to defining the slave as property because it pigeonholes the slave as one of history’s victims: the slave, so the objection goes, was much more than the mere property of his owner, and had a voice and agency of his own.⁴ This criticism misconstrues the purpose of a definition: a definition is not meant as a kind of précis summarizing every noteworthy or important aspect of a given phenomenon. Rather, the function of a definition lies in isolating the distinctive aspect of a given phenomenon.⁵ There is nothing in the traditional ‘property definition’ that denies the humanity or agency of slaves, and leaving out these features from a definition of slavery need not imply that they are unimportant or not worth studying.⁶ Slaves, of course, do have minds of their own, hopes, desires, and ³ See Allain (2008). ⁴ For example, Rankine (2011): 35 writes: what exactly is property? Is Eumaeus simply Odysseus’ property? . . . In Homeric society, a man recognises that he has ‘stuff ’, ktêmata, but there is a clear distinction between the inanimate goods won in war, for example, and a slave. . . . In the Iliad Achilles elevates his slave Briseis to the level of a wife. A slave is more than property and evokes deep emotional and ontological responses. ⁵ It is key to emphasize that a definition is not an analytical tool: it is a classificatory tool. It represents a key preliminary step insofar as it allows us to identify and delimit the subject of our enquiry; but definition tells us nothing about how slavery operates in practice, for which other approaches are required. This point is well made by de Ste. Croix (1981): 92–3 and has been recently reaffirmed by Hodkinson (2008): 286–7. Finley (1980): 72–4 makes several similar key methodological observations. Nor should we think of classification merely in terms of modern scientific attempts at taxonomy and orderliness: all human societies intuitively classify via language. See Goody (1977) passim, esp. ch. 4, and Lakoff (1987) passim. ⁶ This area of research represents an important facet of ancient Greek slavery that is currently being explored by Kostas Vlassopoulos: see, e.g., Vlassopoulos (2011a). My main difference from Vlassopoulos’ approach to slavery concerns the issue of definition, which he views as an ‘essentialist’ and ‘static’ exercise, preferring a ‘processual’ and ‘dynamic’ approach (Vlassopoulos 2016b). I have no objection to a processualist approach to studying slavery; and I hope that something dynamic emerges from Chapter 14 of this book. But such an approach cannot provide the tools to distinguish slaves from non-slaves, which is the task of the legal historian; if this is ‘essentialist’, I can only plead in its defence that it is also essential, lest we ignore the important preliminary task of specifying who is and who is not the subject of our investigation. I do not think

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

27

the capacity for independent action, but so does everyone who is not classed as a slave. Since these qualities are not distinctive to slavery, they are extraneous to the issue of definition. It is the slave’s legal status as the property of his or her owner that places him or her in a specific category, that is, slavery.⁷ Alternatives to the traditional property definition have appeared in various guises over the years. For example, as we noted in the Introduction, the Soviet scholar V. V. Struve rejected the legal definition of slavery and worked with an economic definition that held the slave to be a person who did not possess the means of production. Using this formulation, some Soviet scholars branded antiquity, in both East and West, as an epoch marked by the ‘slave mode of production’. In other words, they claimed that slaves undertook the vast bulk of production. Quite rightly, this view met with vigorous opposition from many European and American scholars as well as several of Struve’s Russian colleagues such as K. K. Zel’in and M. A. Dandamaev.⁸ On the other hand, some scholars have retained the property definition but added various accretions to it. For example, Finley claimed that the slave was not just property, but also an outsider.⁹ Many slaves in Greece and Rome were indeed outsiders to the societies into which they were sold, but it is not immediately obvious why this fact need be incorporated as part of the definition of slavery. In the US South, even before the closure of the slave trade in 1808 nearly all slaves were born on American soil. They were thus not outsiders in the same sense as their progenitors who had been born in Africa; yet, whether born in America or Africa, all of these slaves held the same legal that a processualist approach can dispense with definition; but I recognize that definition itself cannot tell us how slavery worked in practice, or explain historical change. At any rate, as I see it, our approaches are potentially complementary. ⁷ As Finley (1980): 73 n. 18 rightly wrote: 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. 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 property; it merely reveals that he is a peculiar property. Finley was rightly critical of those who objected to cross-cultural definitions of slavery on cultural-relativist grounds: see his remarks on the point raised by S. Lauffer at the Stockholm Historical Congress: Finley (1971): 63; (1980): 69. ⁸ On Soviet definitions of slavery, see Zel’in (1975) and the comments of Dandamaev (1984): 72 and Diakonoff (1980): 15. Cf. Finley (1980): 58–9; Wiedemann (1987): 4–5; Garlan (1988): 9–11. ⁹ See Finley (1968): 307–9. At p. 308 he wrote ‘the slave is an outsider: that alone permits not only his uprooting but also his reduction from a person to a thing which can be owned’. Cf. Finley (1980): 75: ‘the slave was always a deracinated outsider’. He is followed by Garlan (1988): 24, who wrote: ‘The common-sense notion of slavery . . . is dominated by the idea that slaves are par excellence foreigners who are bought and sold as though they were mere objects.’ Vlassopoulos (2016a): 90 rightly notes that Finley contradicted what he wrote elsewhere by insisting on the ‘outsider’ status of slaves.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

28

Greek Slave Systems

status. Those who wish to retain this ‘outsider’ qualification in the definition of slavery thus have to fall back upon notions of the ‘outsider’ that are ever more nebulous and abstract. Moreover, proponents of this view are hard pressed to explain, for instance, those Athenians that Solon tells us were enslaved in Attica before his reforms (Solon fr. 36 [West]). Once more, whether or not a slave is an ‘outsider’ is extraneous to the task of definition (which does not mean that this feature is unworthy of study or is devoid of historical implications, but merely that it need not form a criterion for identifying a given person as a slave in the first place).¹⁰ These criticisms or reformulations of the traditional definition are of relatively minor importance. But in 1982 the sociologist Orlando Patterson delivered a pointed critique of the traditional formulation of slavery in terms of property in his book Slavery and Social Death. Rejecting the property definition, Patterson argued that slavery should be defined as the permanent, violent domination of natally alienated and generally dishonoured persons.¹¹ Patterson’s views have had a profound impact among scholars of ancient slavery, especially in Great Britain and the USA.¹² The traditional approach is untenable, according to Patterson, for two main reasons. First, although Patterson correctly notes that the function of a definition lies in identifying a distinct category of persons, he claims that property fails to achieve this goal because everyone, not just slaves, counts as property of some other person. Everyone, whether beggar or king, is the object of some or other property

¹⁰ Finley’s notion that a slave must be an outsider has seeped into the scholarly consciousness without any real debate over whether or not it belongs in our definition of slavery. For example, Spittler, in his entry on anthropology and ethnology in the Handwörterbuch der antiken Sklaverei writes that whilst slaves have been traditionally defined as property, ‘Diese juristische Perspektive reicht allerdings nicht aus. Finley wies als einer der Ersten darauf hin, dass neben dem Eigentumskriterium die Entwurzelung des Sklaven ebenso wichtig sei. Dieser ist ein Außenseiter.’ See Spittler (2012). Cf. Cartledge in OCD³ s.v. ‘slavery, Greek’. But as Badian (1981): 49, in a review of Finley (1980), points out, ‘we must note that this not an inherent part of the institution’. Cf. Eltis and Engerman (2011): 14: ‘Societies have tended to reserve enslavement for those whom they have defined as not belonging, but this has not always meant that all aliens were enslaved, or that all slaves were aliens.’ ¹¹ Patterson (1982): 13. ¹² See Morris (1987): 174; Fisher (1993): 5–6; Bradley (1994); Callender (1998): 68; Thalmann (1998b): 24; Hunt (1998): 14; Scheidel (2002): 176; Luraghi (2002): 233; Zelnick-Abramovitz (2005): 25–7; Hezser (2005): 26; Dal Lago and Katsari (2008b): 200; Dubois (2010): 4–5; Kleber (2011); Rankine (2011): 35; Kamen (2012); Ismard (2015): 21–2. Patterson (2008) has attempted to analyse slavery in Homeric Greece using his own definition, whilst Cartledge (2011a): 79 endorses Patterson’s approach and elsewhere (2003: 17) suggests that ownership is an anachronistic concept to apply to the Greek world. One should note that not all scholars have been convinced by Patterson’s claims: Harper (2011a): 35 n. 11 notes, in relation to Patterson’s view, that ‘definitions that identify the commodification of the human person as the essence of slavery are more persuasive’. Cruz-Uribe (1982): 65–7 criticizes Patterson in some detail. Harris (2002a) criticized Patterson’s misunderstanding of ownership and expressed the desire to treat the topic at greater length in the future (cf. Harris 2012).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

29

relationship.¹³ Second, Patterson claimed that property is cross-culturally inconsistent in content; many societies even lacked the very concept of ownership.¹⁴ Since the nature of property varies from place to place, it cannot stand as a common denominator in cross-cultural perspective, and thus fails as a sound basis for any general definition of slavery. There are good reasons to contest both of these claims. I have written a more thorough critique of Patterson’s method elsewhere, which I do not propose to repeat in detail here. Essentially, Patterson’s view of property is sui generis, and not at all comparable to what any practising lawyer would call property: for example, he cites contractual relationships between persons as examples of property relations in an attempt to confound the traditional approach; but his arguments fall flat once one realizes that Patterson does not recognize the basic distinction between property and contract.¹⁵ Nevertheless, his views do not stand alone, but rather represent the most famous and influential of a broader set of misconceptions regarding the nature of property and ownership that have considerable currency in modern scholarly literature and have hindered the development of a sensible consensus on the question of definition. The legal approach to ancient Greek slavery is rather unfashionable at present, but, as I hope to show in this chapter and in Chapter 2, the Greeks had their own indigenous, technical (i.e. legal) understanding of slavery that allowed them to distinguish easily between legal slavery and the various metaphorical usages of slave language that, over time, developed from it. Unfashionable as the legal approach may be, it is nonetheless fundamental to an understanding of the issue; and if we avoid it, we will fail to grasp fully the sophistication of the Greeks’ own, emic conceptualization of slavery.

¹³ Patterson (1982): 21. ¹⁴ Patterson (1982): 20; (2012): 323. ¹⁵ See Lewis (2016), the conclusions of which are endorsed in Harper and Scheidel (2018): 92. I have since found that the legal scholar Paul Finkelman spotted the same error in Patterson’s knowledge of property law over thirty years ago: Finkleman (1985). See also Allain (2015): 109–10, who shows that the idea that the ‘sale’ of sportspersons (Patterson 1982: 24) amounts to a property relationship is fallacious. For further discussion on Patterson’s approach, see Penner (2012): 244–8. I am not at all convinced by Patterson’s response to my criticisms (Patterson 2016), which simply adds further unforced errors to his account, and largely relies on an epideictic display of reading in legal literature to convince readers ill-versed in the topic that he is siding with the experts. If one wishes to gain a clearer idea of what actual lawyers think, see Allain (2015): 107–8, who writes regarding the recent Bellagio-Harvard Guidelines on the Legal Parameters of Slavery that this approach: brings to bear the foundations of property law so as to provide a coherent reading of the definition of slavery. Not surprising, for those versed in property law, is that the starting point for engaging with the topic is Antony Honoré’s work in developing the standard incidents of ownership. As Robin Hickey has shown, those powers attaching to the right of ownership should be understood as being synonymous with Honoré’s standard incidents of ownership. On Honoré, see section II, pp. 34–6 below.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

30

Greek Slave Systems

In what follows, I will set out five of the most common misconceptions about property. Following this, I illustrate the nature of ownership according to modern comparative law and legal anthropology, which undermines the set of five common myths about property and provides a more robust picture of the content and nature of ownership in cross-cultural perspective. The bulk of this chapter consists of two detailed case studies, the first analysing slave law in Athens during the classical period, the second the corresponding laws in Babylonia during the Neo-Babylonian and Persian periods. As we will see, this conceptual framework maps accurately onto Greek and Babylonian legal practices.

I . F IVE P O PU LAR MI SCON CEP TI ONS ABOUT PRO PE RTY

1. Pre-Roman Societies Lacked a Concept of Ownership Many scholars believe that the concept of property is a Roman invention, and that it is meaningless to speak of property and ownership in earlier cultures. Patterson wrote of the irrelevance of the Roman concept of absolute property outside civil law systems,¹⁶ and has recently re-emphasized this point: most traditional societies had no such formal systems of law and it is anachronistic, as well as legally ethnocentric, to claim that legal ownership was generally asserted in the old systems of slavery. Exclusive legal ownership is a distinctive principle originating in ancient Roman law and is not attested in many traditional systems of law.¹⁷

Similarly, E. M. Wood has written: The often ambiguous shadings of meaning in the spectrum between dependent labourer, tenant, and freeholder reflects the complicated spectrum of conditions and tenures typical of smallholders in many parts of the world. This is especially true where there is no concept of ‘absolute’ private property but rather various shades of conditional property; where there is no clear dividing line between outright ownership and mere possession or occupancy; and where there is no rigid dichotomy between property and propertylessness of the kind that exists in modern capitalist society. The indeterminacy of peasant conditions and tenures would have been especially characteristic of classical Greece, which, before the development of Roman property law, had no clear legal conception of ownership at all.¹⁸ ¹⁶ Patterson (1982): 20. ¹⁷ Patterson (2012): 323. ¹⁸ Wood (1988): 75, her emphasis. Cf. Morris (2002): 43 n. 8 for a similar view. Wood cites Harrison (1968): 200–5, but Harrison (at p. 201 n. 2) writes that in relation to Rome and Athens ‘the importance of the distinction [i.e. between ownership in the two societies] so far as it concerns the basic conception of ownership has perhaps been exaggerated’.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

31

2. Legal Concepts only Apply in Advanced, Literate Cultures A corollary of this view holds that to speak of law and legal status in certain places and at certain times in ancient Greece is anachronistic. For example, Raaflaub has argued that the language of law and legal status is misleading if applied to ‘Homeric society’: it ‘seems inappropriate since in this early period legal forms were undeveloped and relationships based on power and influence predominated’.¹⁹ Similarly, Cartledge has claimed that individual ownership as a concept is probably not applicable to Sparta’s domination of the helots, or indeed any Greek society: he implies is that it is an anachronism and did not exist as an organizational principle in the Greek world.²⁰ These views can be related to an evolutionist approach to property law of the sort ventured by thinkers such as Marx and Durkheim, who believed that primitive societies were marked by communal ownership of property which only much later evolved into a system of fully fledged private property. (Such assumptions, it should be noted, lie behind a number of attempts to deny the existence of private, alienable land in archaic Greece.²¹)

3. Ownership Is a Set of Absolute Rights As we saw in the quotation from Wood on p. 30 in section I.1 above, ‘ownership’ is often seen as a set of absolute rights. This ‘absolute’ ownership is often asserted as a principle of Roman law; any systems that compromise on this principle, admitting restrictions and limits, thus fall short of this ideal, and must have a ‘conditional’ and ‘indeterminate’ approach to ownership, that is, without any clear boundaries. This view can be seen clearly in the work of Sir Paul Vinogradoff, who wrote that ‘Rome developed the conception of absolute property for the citizen—the dominium ex jure Quiritium—while Athens worked out a conception of relative property rights (κυρίως κτῆσθαι or ἔχειν καὶ κρατεῖν).’²²

¹⁹ Raaflaub (2004): 291 n. 54. Cf. Harris (2012): 353, who points out the flaws in Raaflaub’s position. ²⁰ Cartledge (2003): 17. Cf. Cartledge (1979): 143: ‘it may be anachronistic to think of ownership in juridical terms at so early a date’. ²¹ e.g. Rihll (1991): 103. Fisher (1993): 16 claims that ‘the concept of “ownership” was not well defined in relation to land in archaic Attica’. Foxhall (1997): 63 delivers decisive objections to this view. Cf. Figueira (2004): 54, who believes that helotage originated in a social stage ‘before control and usage had differentiated into categories like public or private ownership’. ²² Vinogradoff (1922): 198.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

32

Greek Slave Systems

4. Ownership Can only Exist where it Is Formulated in an Abstract Definition Another misconception lies in the belief that a phenomenon did not exist or indeed was meaningless to the Greeks or other ancient peoples unless they formulated an explicit, abstract definition of it. For example, R. ZelnickAbramovitz has recently underscored the complex uses of the vocabulary of freedom in Greek literature, and from this conglomerate drawn out one strand: the notion of a free man as one who does not live for another, a formulation ventured by Aristotle in the Rhetoric (Rhet. 1367a27–8) and repeated with minor variations by other ancient authors such as Dio Chrysostom (14.3–4) and Epictetus (4.1.34). Since this is a notion without hard edges, she opts for a fluid distinction between slaves and free persons based on the degree of compulsion under which they find themselves. Her approach entails rejecting a juridical notion of slavery, a notion that she views as an anachronism foisted upon the evidence by modern scholars, and indeed one which was foreign to the Greek mind: ‘The words doulos and eleutheros thus carry cultural, moral, political, and social connotations . . . but these connotations say very little about the legal definition of slavery. Indeed, we look in vain in the Greek sources for a legal and coherent definition of slavery.’²³ Deriding the legal approach as ‘futile’, she instead utilizes a sociological approach inspired by Patterson.²⁴ (In the case of the Near East this argument is particularly problematic, since we have little in the way of abstract definitions of anything, let alone property or slavery.²⁵)

5. Concepts of Property Vary greatly from Culture to Culture One of the reasons why Patterson rejected ownership and property as a suitable basis of a cross-cultural definition of slavery was because he believed that the content of ownership varied greatly from one place to the next. As we saw in his claim quoted on p. 30 in section I.1 above, Patterson believes that whereas slavery is a phenomenon that exists worldwide, legal ownership is not: the latter cannot, therefore, serve as the definitional principle upon which slavery is formulated in cross-cultural perspective. This view has been endorsed for the ancient world by Ian Morris.²⁶ Likewise, in his monograph The Shape of Athenian Law (1993), Stephen Todd writes that ‘it may be asked whether we can legitimately speak of questions like “ownership” at all, because such terms necessarily convey implications about packaging. Ownership and ²³ Zelnick-Abramovitz (2005): 35. ²⁴ Zelnick-Abramovitz (2005): 38; 9. ²⁵ Beaulieu (2007): 474. ²⁶ Morris (1987): 174.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

33

possession are bundles of rights which we package together in particular ways; Athenians may have packed them differently.’²⁷ Todd here is talking about the problem in the abstract, and in this sense he is right to point out that there is little in the way of theory or jurisprudence of ownership and possession in classical Greek writings. However, it is important to make a distinction between abstract formulations and observable practices: Todd claims that the Athenians may have packaged ownership differently, but a glance at the sources suggests that it is unlikely that they (or other contemporary Greeks) bothered to package it at all, the closest thing to a theory of ownership being a couple of throwaway remarks by Plato and Aristotle (Pl. Euthyd. 301e–302a; Arist. Rhet. 1.5.1361a21). Such theoretical abstractions emerged to classify existing practices: they did not call those practices into being any more than Isaac Newton’s discovery of gravity affected how firmly his contemporaries were attached to the ground. But the lack of a jurisprudence of ownership need in no wise hamper the pragmatic utilization of private property and its regulation by law. Since we are in possession of around a hundred law court speeches from classical Athens, we can in fact test empirically whether Greek practices regarding what they called ktesis resembled those that we call ‘ownership’. In other words, differences in the intellectual packaging of legal practices can distract scholars from substantive similarities between the practices themselves across a range of societies.

II. PROPERTY AND S LAVERY IN ANTIQUITY: A P RAGMATIC A PPROA CH All five of the aforementioned views are untenable: this will become clear once we examine them in light of the understanding of ownership current in both legal theory and legal anthropology. But a theoretical refutation of these views is by itself inadequate: vital to the case argued in this book is a pragmatic demonstration of how property and status were understood and articulated in practice in ancient Greek and Near Eastern societies, one that is based empirically on extensive evidence from these cultures. In what follows, I will set out the concept of ownership in modern comparative legal theory. From there, our analysis will proceed to examine the legal relationship between masters and their slaves in classical Athens and in Babylonia during the Neo-Babylonian and Persian periods. These two case studies illustrate in great detail the fact that this was indeed, in legal terms, a relationship based

²⁷ Todd (1993): 243.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

34

Greek Slave Systems

on the fact that the slave was the property of his or her owner, displaying all the basic features required by modern comparative legal theory to qualify for the label ‘ownership’. The most influential modern study of the theory of property law is A. M. Honoré’s essay ‘Ownership’ (1961).²⁸ In this seminal study, Honoré studied the definitions and conceptions of ownership in a variety of legal cultures, including Continental civil law systems, the common law systems of Britain and America, the Marxist-inspired Soviet legal system, and preliterate cultures such as the Trobriand Islanders of Melanesia documented and analysed by Brolinsław Malinowski. Despite the vast differences between these societies, they all conceived of an owner’s relationship to his or her property in a fundamentally similar fashion. Honoré broke this relationship down into ten aspects or ‘incidents’, which remained present from one society to the next: 1. Right to possess. Honoré found that in all of the societies he investigated, there was a clear recognition of a right of possession. This amounted to more than a de facto recognition that a specific item might be in a certain person’s keeping at a certain point in time. Rather, it recognized that an owner had an exclusive right to possession of that property, one that was not shared by other members of society. This aspect is closely linked to the protection of this right (see Right to security below). Intellectual packaging aside, any legal system with a concept of theft recognizes a practical distinction between possession and ownership. 2. Right to use. Ownership in all societies is the ‘greatest possible interest in a thing which a mature system of law recognizes’. This means that the owner’s liberty to make use of his property is more extensive than any other form of legal interest; an owner’s rights over his property constitute an ‘open list’, the limits of which are set by the stipulation that an owner cannot use his property in an illegal manner (see Prohibition of harmful use below). 3. Right to manage. Ownership of property entails the right to decide how and by whom one’s property may be put to use. For example, ‘an owner may not merely sit in his own deck chair but may validly license others to sit in it, lend it, impose conditions on the borrower, direct how it is to be painted or cleaned, contract for it to be mended in a particular way’.²⁹

²⁸ Honoré (1961); for its importance, see Hann (1998): 6; Benson (2002): 773–7; Allain and Hickey (2012): 925 n. 35. ²⁹ Honoré (1961): 116.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

35

4. Right to income. If an owner chooses to loan his property to a third party, he or she is the rightful recipient of any income generated by that loan. Likewise, the owner of a field and its crops reaps the benefits of a harvest, just as the owner of a fruit tree gains the fruits that grow upon it. 5. Right to capital refers to the right of owners to alienate their property, which normally amounts to sale or gift. This transfers the rights of ownership to the new owner, who may in turn alienate the property if he or she wishes. 6. Right to security. Owners of property in any society are guaranteed exclusive rights to their property and protection of their possession. This means that third parties cannot interfere with the property, damage it, or take it from the owner without due cause. All legal systems, then, provide rules on damage and theft that protect the rights of owners and set out the means by which owners can recover stolen property or receive compensation if this is not possible. 7. & 8. Transmissibility and Absence of term. Ownership in all societies involves an aspect called ‘duration’. This comprises two features: transmissibility and absence of term. An owner’s rights over their property are essentially permanent: unless he or she chooses to alienate their property, or unless the property is seized by the state or a third party for legally legitimate reasons (see Liability to execution below), his or her interest in the item is permanent. If a man buys a watch from a shop, he does not have to give it back after a certain period of time (this aspect marks an important difference from legal arrangements such as hire and usufruct). Transmissibility means that the permanence of this interest is so strong that it outlasts the lifespan of the owner: the property thus passes to his heir after the owner’s death. The watch, then, passes to a new owner rather than reverting to, for instance, the state. Different societies will construct different rules on wills, intestate succession, and so on; but the principle that the property passes to heirs is common to all human societies. 9. Prohibition of harmful use. The principle of prohibition of harmful use means that an owner cannot use his or her property to illegal ends. For example, a person might own a car, but they cannot drive that car onto private land without permission, nor drive without insurance, never mind run over their neighbour with it. 10. Liability to execution. Another aspect of ownership in all societies is the fact that although owners generally enjoy exclusive rights of possession, some circumstances allow third parties to deprive the owner of his or her property. One common instance lies in loan agreements, where owners can pledge their property as security for a loan, meaning that in the event of default, creditors can seize the property and dispose of it. Individuals with taxes in arrears in modern Britain or the USA might also have items of property seized by the government. In certain circumstances, too, the state may seize property. For example, when Baron Haussmann remodelled Paris in the 1850s, the French

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

36

Greek Slave Systems

state’s interests trumped those of private property owners, allowing him to demolish large parts of the city and rebuild it to his own design.³⁰ Likewise, the UK government is currently planning a high-speed rail link between London and Birmingham, and is forcibly purchasing land and houses lying on the proposed route. There is a limit, then, to owners’ rights over their property, and when the interests of private owners run up against the interest of the state, it is usually the latter that triumphs. Honoré’s analysis of ownership was not, however, a rigid one, nor did he claim that the precise content of property law was identical from one place to the next. He noted three areas in which significant variation might occur from one legal system to the next: 1. Who can own? Ownership of property is often limited to different groups of persons in different societies. For example, in classical Athens, only citizens were allowed to own land; metics (resident foreigners) were only able to rent land and houses. As a modern example, one could note that only individuals who hold the correct licence in the UK can own firearms, and so on. 2. What can be owned? Another area of variation is the notion of what can be owned. For example, the UK government classifies some substances as illegal drugs, banning their ownership altogether. Other countries take a more liberal line, permitting private ownership of these substances. In the former Soviet Union, various things such as factories could not be owned by private individuals.³¹ 3. Restrictions on ownership. Even if two societies permit the same sort of person to own the same kind of property, the specific rules that govern the use of that property will usually differ. For example, a person may own the same model of car in the UK, France, or Germany; but in the UK, he may only drive it up to 112 kph. His French counterpart can drive a little faster, up to 130 kph. But in Germany, where large stretches of the Autobahn lack speed restrictions, one might drive the car at the highest speed that can be coaxed out of the engine. The relationship between an owner and his or her property may then vary from place to place in terms of details, but according to Honoré it maintains a consistent set of features across all human societies. The areas of variation, though, have allowed a vast number of specific configurations of property relations to develop: an item which can be privately owned in one society might be public property, or the property of an institution, in another; and where collective interests impinge on private interests only slightly in one place, they may make

³⁰ On Haussmann’s remodelling of Paris, see Jordan (1995).

³¹ Birks (1985): 8.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

37

much deeper inroads in another, and so forth.³² In some societies, certain things simply cannot be owned. This approach is in accord with the findings of legal anthropology, which has given the lie to several of the popular misconceptions of ownership mentioned in section I above. For example, when Patterson claims that ‘exclusive legal ownership is a distinctive principle originating in ancient Roman law and is not attested in many traditional systems of law’, one assumes that this opinion is based on modern anthropological research. But one leading anthropologist, the late Jack Goody, wrote as recently as 2006 that ‘the widespread idea that individual property rights are an invention of Roman law—or of the west—completely overlooks the sophisticated analysis by anthropologists of the jural order in oral cultures’.³³ In fact, the idea that private rights of ownership do not exist in primitive cultures was challenged on empirical grounds nearly a century ago by Malinowski, who pointed out this view simply bore no resemblance to the realities that he had observed in his close studies of the Trobriand Islanders of Melanesia, where private property rights were extensive.³⁴ Similarly, Pospisil’s field studies of the Kapauku Papuans, living to all intents and purposes in a Stone Age society, illustrate the complex system of private property that they had devised: A house, a boat, bow and arrows, fields, crops, patches of second-growth forest, or even a meal shared by a family or household is always owned by one person. Individual ownership . . . is so extensive in the Kamu valley that we find virgin forests divided into tracts which belong to single individuals. Relatives, husbands and wives do not own anything in common. Even an eleven-year-old boy can own his field and his money and play the role of debtor and creditor as well.³⁵

As Benson has rightly written, ‘the emphasis on private property may seem surprising to those who think of tribal society as some sort of socialist or communal system. On the contrary, however, private property rights are a common characteristic of primitive societies; they constitute the most important primary rules of conduct.’³⁶ (Honoré’s approach has, it should be noted, been endorsed by social anthropologists such as Chris Hann, leader of the Halle Focus Group on property relations.³⁷) From where, then, does the popular misconception spring? One answer may lie in Honoré’s list of cross-culturally variant features of ownership, in particular the issue of what can be owned. Many nomadic societies lack a concept of private land ownership because the itinerant nature of their ³² On these various regimes, see Hann (1998) and the work of the Halle Focus Group on property relations, on which, see Hann (2005). For a splendid ethnographical overview of property, inheritance, and marriage practices in global perspective, see Goody (1969). ³³ Goody (2006): 58. ³⁴ Malinowski (1926): 17–21. ³⁵ Pospisil 1971: 66. ³⁶ Benson (1989): 8. ³⁷ Hann (1998): 6.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

38

Greek Slave Systems

activities precludes any permanent settlement, with the implication that individuals are unlikely to lay claims to specific plots of land (although they may have a broader notion of tribal territory). At the same time, the need for mobility means that personal possessions are kept to a minimum: the result is an undeveloped sense of materialism which appears strikingly different from the property regimes of more developed societies.³⁸ But the contrast does not lie (as the popular view has it) in an inability to conceive of private property rights, but rather in a pragmatic need to keep cumbersome possessions to a bare minimum and a lack of any real need to vest individual rights in specific parcels of land. Can we then agree with Wood in making a distinction between an ‘indeterminate’ form of private property without clear boundaries and a Roman concept of absolute property in which owners exercise unfettered rights over their property, or with Vinogradoff in making a distinction between ‘absolute’ ownership in Rome and ‘relative’ ownership in Athens? This view also stems from a popular misconception that has been rejected by jurists working on Roman law. Peter Birks has shown that despite dominium being ‘absolute’ in abstract terms, legal practice in Rome admitted a wide variety of state-imposed limits and restrictions on owners’ rights. Even in Rome, then, ‘absolute’ ownership did not exist as a practical reality.³⁹ As Honoré points out (in his third element of cross-culturally variant aspects of private ownership), liberties and restrictions that frame the rights of owners may vary in their specifics from one legal system to the next, but we cannot insist on grouping some legal systems in a category of ‘indeterminate’ private property and others in a category of ‘absolute’ ‘Roman’-style ownership. Furthermore, and pace Wood, the ‘indeterminacy’ she asserts for some Greek systems is really just a chimera: legal systems in the Greek world, at the very least, were extremely specific about what liberties and restrictions applied to an owner’s rights over his property: they did not exist within an ‘indeterminate’ spectrum where possession or occupancy are difficult to distinguish from ownership sensu stricto or where the one position shades into the other.⁴⁰ For example, in the inheritance cases of the orator Isaeus, no litigant ever makes an argument based on the indeterminacy or ‘fuzziness’ of property rights; all the parties share an implicit understanding of what property rights amount to in practice. Their disagreements lie in who is entitled to exercise these rights, and the ³⁸ This trend is described well by Sahlins (1972): 11–13. ³⁹ Birks (1985). In a review of Kränzlein (1963), C. Bradford Welles (1965: 66) makes the same basic point: ‘dominium, however absolute in theory, was never absolute in practice, but hemmed in by precisely the same kind of restrictions and controls imposed by the state which we find in Greek law’. Writing in the anthropological tradition, Hann (1998): 2 notes that ‘in all societies the property rights of individuals are subject to political as well as legal regulation’. ⁴⁰ On the pragmatic distinction between ownership and possession in Greek law, see Kränzlein (1963).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

39

evidence for why this or that person holds such a right.⁴¹ In this instance, Wood’s claims (quoted on p. 30 in section I.1 above) amount to little more than vagueness masquerading as nuance.

III. THE P RAGMATIC APPROACH: TWO TEST CASES So far we have dealt only with theoretical arguments: the concept of property that Patterson has attacked and rejected as a basis for defining slavery simply fails to match up to the view of property as understood in modern comparative legal theory. Furthermore, the popular view that esteems private property rights to be an advanced development of Roman legal theory ignores the findings of almost a century of legal anthropology, which has observed private property systems in a variety of tribal social systems that were far less advanced in terms of technological and social complexity than even the society imagined in Homer’s epics. But does the alternative approach advocated by Honoré bear any real resemblance to what we find in ancient Greek and Near Eastern legal practices? Or does choosing this approach merely forefront a different set of anachronisms that obstruct our evaluation of ancient evidence? In order to demonstrate the utility of this approach, I will take soundings in two ancient societies: Athens during the classical period, and Babylonia during the Neo-Babylonian and Persian periods. In both cases, I will illustrate the legal position of owners vis-à-vis their slaves in relation to legal practice, focusing not on abstract definitions found in the works of this or that philosopher, or even on ‘Code’ law, but in the quotidian, empirical evidence provided by real legal cases that were argued before judges in Athenian and Babylonian courts, as well as business documents and records. These case studies show in detail that the description of ownership adumbrated by Honoré maps accurately onto the real legal position of slaves in these ancient societies.

Classical Athens The interest of the Athenian state in recognizing and protecting the title of slaveholders (right to possess, right to security) is evident in the hefty fines ⁴¹ Cf. Dover (1974): 170, who rightly points out that most Greeks were: born in a society which at that time had stable assumptions about the difference between ‘mine’ and ‘yours’; these assumptions underlie the provisions made by the law for the restraint of any ways of in which I may increase what is mine by depriving you, without your consent, of what is yours, and they also underlie the application of words such as ‘theft’, ‘fraud’ and ‘honesty’.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

40

Greek Slave Systems

handed out to individuals who sought to deprive owners of their slaves. In one case, the father of a man named Theocrines was fined 500 drachmas—around a year’s wages at the average rate for a craftsman—for falsely asserting that a private citizen’s slave was a free woman ([Dem.] 58.19). Conversely, harsh penalties applied to those who sought to enslave free persons ([Arist.] Ath.Pol. 52.1; Din. 1.23). It is no surprise, then, that when one individual tried to carry off his opponent’s son, believing him to be a slave, all it took was a word from a neighbour regarding the boy’s status to effect his release ([Dem.] 47.61). A procedure called aphairesis eis eleutherian (‘removal to freedom’) allowed individuals claimed as slaves to be rescued from detention by their friends, with comparable effects to habeas corpus in British law; but these friends would then have to face the man claiming to be the rescued individual’s owner in court.⁴² If the court decided that the slave had been falsely asserted to be free, half the penalty was paid into the public treasury ([Dem.] 58.21). This shows a clear public interest in enforcing payment of fines for this crime, for if the guilty party did not pay, he became a public debtor. It is perhaps telling that suits concerning claims over slaves were treated in the same category as those concerning animals ([Arist.] Ath.Pol. 52.2). Slaves could be tattooed to make their status clear and (perhaps) to indicate the identity of their owner.⁴³ The rights of owners to use and abuse their slaves in Athens were extensive. Violent abuse was routine, but not universal: Athenian writers noted that some masters were brutal whilst others were comparatively lenient (Pl. Leg. 6.776d–e; 777a; cf. Xen. Oec. 3.4); Xenophon took an intermediate position, accepting the need to punish slaves but warning against doing so in a state of anger (Xen. Hell. 5.3.7; cf. Hdt. 1.137). It is more important for the present purposes that we understand what the law permitted. What is striking is the near total lack of restrictions on mistreatment of slaves, to a greater degree even than in Roman law.⁴⁴ Masters could corporally punish their slaves in whatever manner they deemed fit, and Athenian comedy is replete with references to whippings and abuse.⁴⁵ Aristophanes (fr. 197 K-A with Suda ⁴² On the procedure of aphairesis eis eleutherian, see Harrison (1968): 178–9. ⁴³ On tattooing, see Jones (1987), who argues that στίζω and its cognates mean ‘tattoo’ in our period, and do not pertain to branding. Kamen (2010) cogently argues that we are better off translating the term as ‘mark’, as it is not usually clear whether branding or tattooing is implied. One must make a distinction between slaves who belonged to peoples with a cultural tradition of tattooing, such as Thracians, and slaves tattooed to mark their status as property, or for a punishment. For the former, see Tsiafakis (2015); for the latter, see Bremmer (2015): 139–40, and especially Kamen (2010). ⁴⁴ As Nicholas (1962): 69 points out in respect to Roman law, these restrictions took ‘the same form as our legislation for the protection of animals’. Hunter (2000): 9 rightly notes that at Athens ‘the punishment of slaves was a matter of private practice and left to the discretion of the master’. ⁴⁵ Hunter (1994): 165–73. On the beating of slaves in Aristophanes, see Lévy (1974): 39–41. On violence towards slaves in Menander, see Konstan (2013). On slaves in comedy, see the essays in Akrigg and Tordoff (2013). Aristophanes (Vesp. 441–3) refers to tying a slave up to a tree and

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

41

δ 1267) records a proverb about just punishments, comparing the beaten victim to an octopus pulverized to make it tender. Pollux (3.78–9) notes various names of implements for punishing slaves; of chief importance was the whip. The Corcyraeans manufactured the most famous variety (Ar. Av. 1461–4; Phrynichus fr. 47 K-A; Arist. fr. 513 [Rose]).⁴⁶ Masters could also chain or fetter their slaves if they proved recalcitrant (Xen. Mem. 2.1.16; Ar. Ran. 1508–14; Vesp. 435; Plu. 276; Ar. fr. 871 K-A; Dem. 18.129).⁴⁷ Oppression was not only corporeal: many owners changed the names of their slaves in an attempt to strip them of their identity (Pl. Cra. 384d).⁴⁸ Of course, not all masters chose to exercise their full rights of punishment over their slaves, but the degree to which violence towards slaves was socially acceptable is vividly conveyed in Lysias 1. 18–22, where a litigant actually relates to the court how he had threatened to torture information out of one of his slaves and throw her into a mill.⁴⁹ Such an admission would have been a disastrous rhetorical strategy had there existed any real popular sentiment against the mistreatment of slaves (cf. Xen. Mem. 2.1.16; Hyp. 3.5; Arist. Rhet. 2.3.5. 1380a14–24).⁵⁰ (This is one of the most telling pieces of evidence for the treatment of slaves in classical Athens, for it allows us to generalize in a way that no collection of individual instances of abuse can match.) Owners could also abuse their slaves sexually: the same litigant in Lysias I, Euphiletus, describes how he had assaulted one of his slave girls on an occasion when he was drunk (Lys. 1.12). Molesting slave girls was unlikely to disgust the judges of the case, or Euphiletus would not have mentioned it. Athenians could also laugh about such things in the theatre (e.g. Ar. Pax 573, 1138; Ach. 273). Perhaps the most blatant statement of this principle can be found in Xenophon’s Oeconomicus (10.12), where Ischomachus nonchalantly whipping him; a graphic representation of such a procedure can be seen in the name vase of the Beldam Painter, c.480 BC. For an image, see Haspels (1936) plate 49a–c. ⁴⁶ Masters may well have been able to emasculate their slaves: we certainly know of no rule against it. The practice is attested elsewhere in the Greek world: see Hdt. 3.48, 8.105; on the latter passage, see now the remarks of Braund (2008): 15–16. We do hear of a eunuch doorkeeper at Kallias’ house in Plato’s Protagoras 314c, but it is unclear where his castration occurred. Cf. the eunuch (porter?) who fell foul of Diogenes the Cynic’s wit (Diogenes Laertius 6.39). ⁴⁷ On slave punishment, see further Klees (1998): 176–217. ⁴⁸ Slaves lack patronymics in Athenian inscriptions: see, e.g., IG I³ 1032, where the slaves are identified with their owner’s name in the genitive case. On Attic slave names, see Vlassopoulos (2010; 2015); Lewis (forthcoming d). ⁴⁹ For the mill as a place of slave punishment, see Aristophanes fr. 95 K-A, Men. Asp. 238–45, Heros 1–3, Dem. 45.33, with Hunter (1994): 171–2. For a clay model from Boeotia where slaves(?) grind meal to the tune of an aulos, see Ehrenberg (1961) plate XI b. For a collar device that mill slaves wore that stopped them from eating that which they kneaded, see Ar. fr. 314 K-A. Slave women in mills were given the nickname ‘mill roach’ (μυλακρίς): Ar. fr. 600 K-A. Cf. also Din. 1.23. The image of the mill slave as a stereotype of the most degraded of persons was found also in the ancient Near East: see Baker (2016): 19–20. ⁵⁰ It is not greatly surprising that we hear stories about slave suicide: Theopr. Char. 12.12; Hist. Pl. 9.16.6.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

42

Greek Slave Systems

tells Socrates that a wife is more attractive than a slave because wives give their consent to sexual intercourse, whereas slaves have no choice but to submit (cf. Xen. Hier. 1.28).⁵¹ The plaintiff Epicrates in Hyperides’ speech Against Athenogenes attributes a similar sentiment to Athenogenes, who was baffled by Epicrates’ efforts to woo a slave boy whom he had bought when the boy was his to do with as he pleased (Hyp. 3.24).⁵² Owners could also place their slaves in a brothel to bring in an income.⁵³ Athenian slaveholders might even get away with killing their slaves. The speaker in Ant. 5.47 makes it clear that the execution of a slave by his or her owner was illegal and required official sanction (and was probably carried out by the public executioner⁵⁴); but another passage (Ant. 6.4) suggests that men who killed their own slaves might only perform purificatory rituals to assuage miasma and faced no legal penalties (cf. Isoc. 12.181, Pl. Leg. 9.868a).⁵⁵ On the other hand, some scholars have claimed that slaves in Athens had various rights. One of the key texts marshalled in support of this view is the law on hubris mentioned in Demosthenes’ speech Against Meidias and Aeschines’ speech Against Timarchus. This law has been interpreted as a humanitarian measure designed to prevent anyone—even masters—from inflicting violence on slaves.⁵⁶ It is difficult to accept this interpretation. The inserted texts of the law itself are forgeries,⁵⁷ though the surrounding text does make it clear that the law forbade hubristic acts against slaves (Aeschin.1.15; Dem. 21.46; 48; cf. Hyp. fr. 120). However, Aeschines (1.17) emphasizes that the law was not designed to protect slaves but to punish a specific antisocial kind of behaviour. To suppose that the law applied to a master’s treatment of his own slave requires us to ignore the plentiful evidence for routine and socially acceptable violence towards slaves by their owners, and is difficult to reconcile with the statement of Plato that a slave had no way of helping himself when he or his loved ones were trampled upon (Gorg. 483a–b). As Canevaro has shown, the law included slaves because the lawgiver wished to curtail the spread of hubristic behaviour among the citizens tout court. The hubris law was designed to engender respect and orderly conduct among citizens, not to ⁵¹ Though one should not underestimate the amount of marital rape: see Llewellyn-Jones (2011). ⁵² These examples create significant problems for the view of Cohen (2000): 160–7 that the hubris law protected slaves from sexual abuse at the hands of their owners. ⁵³ On brothels, see Davidson (1997): 83–91; Glazebrook and Tsakirgis (2016). ⁵⁴ It is the public executioner, at least, who kills the slave poisoner in Ant. 1.20. ⁵⁵ Pace Cohen (1998): 116 n. 62 ‘the slave . . . did have a right to live: he could not be killed with impunity by his masters’ (Cohen cites Ant. 5.47 but ignores Ant. 6.4). ⁵⁶ Morrow (1937): 37–41, and for a more extreme view, Cohen (2000): 160–7, who believes that the hubris law protected slaves from all kinds of abuse at the hands of their owners. Dmitriev (2016) takes a more measured view, and interprets the law in terms of protecting the household and household slaves from hubris at the hands of other citizens. ⁵⁷ Drerup (1898): 305–8; Harris in Canevaro (2013): 224–31.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

43

protect slaves.⁵⁸ There was one feature of Athenian law, however, that told in the slave’s favour: he or she might run away and take sanctuary in the Theseion in the agora⁵⁹ or the shrine of the Eumenides on the Areopagus (Ar. Eq. 1312; Thesm. 224), and demand to be sold to another owner. However, we do not know whether or not this was a common occurrence, or what would happen if a potential buyer did not come forward. In terms of their rights of management, owners could establish their slaves in whatever occupations they deemed fit, and slaves could be found in many areas of the Athenian economy.⁶⁰ They might also hire out their slaves to third parties (Xen. Vect. 4.14; [Xen.] Ath.Pol. 1.17; Dem. 27.20; 28.12; 53.20; [Dem.] 59.20; Is. 8.35; Theophr. Char. 30.17; Cratinus fr. 171 K-A ll. 73–4; Hyp. 1A fr. 4.1). If an owner were to apprentice one of his slaves to a craftsman, he would draw up a contract stipulating what the slave should have learned in a given period (Xen. Eq. 2.2). Owners enjoyed full rights to the income of their slaves. This principle was generally unrestricted, though owners of girls who played the aulos, harp, or lyre could not hire them out for more than two drachmas per night ([Arist.] Ath.Pol. 50.2, cf. Hyp. 4.3; Suda δ 528). In some circumstances, a slave might be allowed to live separately and earn his own wages, handing over a fixed sum (apophora) to his owner and keeping the remainder (Aeschin. 1.97; Theophr. Char. 30.15; Men. Epitr. 378–80; Andoc. 1.38; [Xen.] Ath.Pol. 1.11); but it should be emphasized that this was a de facto arrangement at the discretion of the owner, and did not grant the slave any right to the remainder.⁶¹ For example, in the confiscation records relating to the affair of the herms and the profanation of the Eleusinian Mysteries, there is a record of the goods of a slave leather worker named Aristarchus, listed along with Aristarchus himself among the property confiscated from his owner, Adeimantus (IG I³ 426.14; 24–39). Aristarchus had presumably been working on his own and paying an apophora to Adeimantus, but when the latter was convicted and his property seized, Aristarchus and all of his possessions were seized as well. Likewise, in [Dem.] 53.20 Apollodorus uses the fact that the slave Cerdon (who was let out on hire) had his wages paid to Arethusius by those who hired him to prove that Arethusius was his master. Clearly, the owner enjoyed full legal title over anything the slave earned, even if he might de facto allow the slave keep a portion of his or her wages or accumulate some savings (cf. Hyp. 3.22; [Dem.] 59.21; Theophr. Char. 30.9). This principle applied to all fruits of slave labour: children born to slaves became ⁵⁸ Canevaro (2018). ⁵⁹ MacDowell (1978): 81. For slaves seeking asylum in temples, see Naiden (2006): 373–4. ⁶⁰ For the tasks and jobs of slaves, see Ehrenberg (1961): 181–4; Rihll (2011). ⁶¹ These slaves are often identified with the khoris oikountes, but the latter were more likely a specific category of manumitted slaves: see Canevaro and Lewis (2014). For apophora in Athens, see Kamen (2016).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

44

Greek Slave Systems

the slaves of their parents’ owners (Men. Epitr. 469).⁶² This principle applied to the phenomenon of slaves belonging to other slaves: in Chapter 5 we will discuss the slave Mesaulius who belongs to the swineherd Eumaeus in the Odyssey; by the classical period, this phenomenon had become more complex: Theopompus of Chios (FGrHist 115 F253) mentions a Thracian madam named Sinope who moved her brothel from Aegina to Athens; among her slaves was a flute player named Bakchis, who in turn sub-owned a girl named Pythionike. This Pythionike became a famous hetaera and the lover of Alexander’s treasurer Harpalus. Theopompus derisively refers to Pythionike as a ‘triple slave’ (τρίδουλος), emphasizing the layers of sub-ownership under which she lay.⁶³ Slaves were fully alienable under Athenian law (right to capital), and could be given away (e.g. [Dem.] 53.6; Hyp. 3.23–8; Men. Sam. 380–2) or sold (Xen. Mem. 2.5.2), though if a slave were sold many times, it was seen as indicative of the slave’s lack of worth (Poll. 3.125). Slave sales were warranted to protect the buyer against any hidden ‘latent defects’ that the vendor had failed to declare, allowing the dissatisfied customer to return the slave and get his money back (Hyp. 3.15; cf. Pl. Leg. 11.916a). Private manumission does not seem to have been restricted, apart from a law that prevented the declaration of manumission in the theatre due to its disruptive effect on performances (Aeschin. 3.41–4).⁶⁴ Freedom could be granted by the owner when alive, or posthumously through his will,⁶⁵ but it appears to have been an informal procedure.⁶⁶ Let us now turn to the master’s security of ownership. Generally speaking, the principle that only a slave’s owner might beat a slave was widely observed, a fact that the Old Oligarch found objectionable, preferring the more ‘liberal’ Spartan rule that allowed any citizen to beat capriciously the helot of one of his ⁶² Cf. the mention of home-born slaves in Xen. Oec. 9.5; Pl., Meno 82a–b; Lys. 13.18; [Dem.] 53.19; IG I³ 422.100–2; 422.103–5; 426.16; Σηκίς in Pherecretes fr. 10 K-A and Ar. Vesp. 768 with Poll. 3.76; Ar. fr. 596 K-A; Arist. Pol. 1255a; Ar. fr. 596 K-A; Hesychius s.v. ἀμφίδουλος; Men. fr. 411 K-A = Suda s.v. ἅβρα. See also the remarks of Harrison (1968): 164 on slave parentage. ⁶³ See J. P. Lewis (2012): 13 and appendix B on the term τρίδουλος (which, as he points out, is not a technical term). Good evidence that slaves had no title over their sub-owned slaves can be found in the will of Aristotle (Diogenes Laertius 5.14): here, we find a slave named Ambrakis who is to be granted her freedom and, on the occasion of Aristotle’s daughter’s marriage, is to have 500 drachmas and title to the slave girl she already has (εἶναι δὲ καὶ Ἀμβρακίδα ἐλευθέραν καὶ δοῦναι αὐτῇ, ὅταν ἡ παῖς ἐκδοθῇ, πεντακοσίας δραχμὰς καὶ τὴν παιδίσκην ἣν ἔχει.). This shows quite clearly that sub-ownership of slaves by slaves was a privilege, not a right. On the authenticity of Aristotle’s will, see Gottschalk (1972) and Canevaro and Lewis (2014). ⁶⁴ See also Plu. Nic. 3. For further discussion, see Mactoux (2008). ⁶⁵ The best evidence for the latter lies in the wills of the philosophers preserved by Diogenes Laertius. See Canevaro and Lewis (2014). ⁶⁶ I leave aside here the notorious phialai inscriptions. Meyer (2010) has convincingly demonstrated that they do not relate to a manumission procedure, but her own reconstruction is quite speculative, and a robust explanation of these documents remains a desideratum (I have changed my position from that expressed in Lewis 2011: 102 n. 46). On these documents, see Harris (forthcoming).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

45

peers ([Xen.] Ath.Pol. 1.11). However, there were several exceptions. If a slave trespassed on a citizen’s land in order to steal or damage his property, the wronged party was entitled to punish the slave ([Dem.] 53.16; Ar. Ach. 271–6).⁶⁷ (A contemporary parallel can be found in a fourth-century lease from Thasos that empowers the lessee to whip any slaves who are caught dumping kopros on the property: LSCG 115 = IG XII 8 no. 265.) A lead letter from the Athenian agora describes how a foundry owner badly beat a slave given over to work for him, and this may show that slaves hired out in such a fashion could be corporally punished by the hirer.⁶⁸ Moreover, certain offences by slaves could be punished by state officials regardless of the consent of the slave’s owner.⁶⁹ These instances show that although the master’s rights of security were strong, they were not absolute. One matter in which the owner exercised discretion over any violence done to his slave was in requests to have the slave subjected to torture (basanos), for he chose whether or not to accede to the request.⁷⁰ Since slaves lived in close proximity to their owners, they were potentially privy to various secrets and had extensive knowledge about their owner’s family and dealings:⁷¹ an exception to testimony extracted under torture was in cases of menusis in which slaves might denounce their owners for certain offences.⁷² This was a high-risk

⁶⁷ Intent is crucial: see the rationale stated in Men. Dysc. 141–3, with Harris (2004: 162 = 2006: 275–6). Apart from in the special circumstances noted above, it was an offence to rape another person’s slave; in Athens, the fine exacted was half the amount imposed for raping a free person: see Lys. 1.32. ⁶⁸ See Harris (2004; 2006: 271–9); see also Harvey (2007). For slaves hired out in the workshops of other persons, see Hyp. 1A fr. 4.1. ⁶⁹ A Solonian law set fifty stripes as the punishment for slaves who took free boys as lovers (Aeschin. 1.139), and the same penalty appears in classical inscriptions for slaves offering false coinage (Stroud 1974 lines 30–2) and stealing wood from a sanctuary (IG II² 1362.9–10). See also the case of fly-tipping (IG II² 380.40–2). Whipping was also applied to state-owned slaves in cases of infraction: Stroud (1974) lines 13–16; IG II² 333.6–7; IG II² 1013.5 (fifty strokes, restored, but securely so); cf. ll. 45–9. See Hunter (1994): 155–62 for further discussion. ⁷⁰ This point is not in dispute; but the use of torture is subject to some debate. For Thür (1977; 1996), and Gagarin (1996), evidentiary torture was a legal fiction and not actually practised in classical Athens. Against this view, see the more compelling analyses of Hunter (1994): 91–5 and Mirhady (1996; 2000). A key point is made by Mirhady (2000): 72: in many cases ‘the evidentiary basanos is fictive, a rhetorical trick, but for its rhetorical effectiveness it had to rely on a reality in which violence and torture against slaves were common practice. Otherwise, the fiction would have been utterly transparent.’ One might add that Athenian litigants, constrained by the clepsydra, could not afford the time to devote to rhetorically dud strategies. ⁷¹ Sources collected and discussed in Hunter (1994): 70–91. ⁷² Osborne (2000) argues that this rule was limited to cases of impiety. He argues ex silentio since the tyranny law SEG 12:87 and the Decree of Demophantus (Andoc. 1.96–8) contain no provision for slave informers. But as Osborne notes (at 83 n. 14), the tyranny law is incomplete; moreover, the Decree of Demophantus is a forgery: see Canevaro and Harris (2012): 119–25. I am, therefore, not confident that we have enough evidence to contrast Athenian practice with that of Thasos, Keos, and Ilion along the lines suggested by Osborne.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

46

Greek Slave Systems

strategy for the slave: were his testimony believed, he would be freed; if not, he might be executed.⁷³ We have little information on the theft of slaves. The speaker at [Dem.] 47.52–3 implies that slaves could only be taken from their owners if a third party had won a judgment against them and had a right to seize items of their property. We hear from Lys. 13.65 of a man condemned to death at Corinth for the theft of a slave; the same penalty must have applied to the theft of slaves at Athens, since thieves were counted as kakourgoi and subject to the death penalty ([Arist.] Ath.Pol. 52.1).⁷⁴ Trials concerning the murder of slaves by third parties took place at the Palladium ([Arist.] Ath.Pol. 57.3). In [Dem.] 47.70 we learn of a case where a freedwoman was murdered; the son of her ex-owner consulted the exegetai on how to proceed against them, since she was neither his kin nor his slave, implying that murderers of slaves would normally be prosecuted by the slave’s owner. Isocrates (18.52–4) describes such a case: two men, Cratinus and Callimachus, were engaged in a dispute over a farm. Callimachus hid one of his slave women, claimed that Cratinus had killed her, and brought an indictment against him at the Palladium. However, Cratinus succeeded in capturing the slave and presented her to the court. Callimachus failed to get a single vote (cf. [Dem.] 59.9). It is likely that the penalty for those convicted of killing slaves was a fine (Lyc. Leocr. 65). Owners held an indefinite interest in their slaves, one that transferred to their heirs after they died (absence of term; transmissibility): several wills attest the bequeathal of slaves in this manner (Aeschin. 1.97; Dem. 27.9; 45.28; Dem. 48.12; Diogenes Laertius 5.13; 54; 62–3; 73). They could, of course, sell, give away, or manumit the slave at any time, and manumission was sometimes cited as a useful incentive to motivate slaves ([Arist.] Oec. 1344b; Arist. Pol. 1330a32–3); but the choice of whether to keep a promise of manumission or to renege on it was at the discretion of the owner.⁷⁵ Athenian slaveholders were liable for the actions of their slaves (prohibition of harmful use). One of the best examples of this principle is the case Against Athenogenes (Hyp. 3). Athenogenes, according to the litigant Epicrates, sold him a slave named Midas and his two sons. Epicrates later discovered that he had unwittingly acquired responsibility for Midas’ debts of five talents, a crippling sum (Hyp. 3.9). The litigant mentions a law that made the expenses (or fines) and misdeeds (tas zemias . . . kai ta a[dikem]ata) of a slave the responsibility of the man who owned the slave when the misdeed ⁷³ Harrison (1968): 171 with n. 1. ⁷⁴ One should make a distinction between the thief (kleptes) and the enslaver (andrapodistes) mentioned at [Arist.] Ath.Pol. 52.1. Andrapodistai were those who enslaved free persons; those who stole slaves counted rather as kleptai. But for an atypical usage, see Lycurg. fr. 11.1. ⁷⁵ The institution of debt bondage provides a useful contrast to slavery in terms of the duration of the master’s rights, temporary in the former case, permanent in the latter case. See Harris (2002a = 2006): 249–69.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

47

occurred (Hyp. 3.22).⁷⁶ This principle seems to have applied to debts incurred by slaves: since the debts had been accumulated when Midas was owned by Athenogenes, the latter inserted a specific clause in the small print of the sale contract that made the buyer assume any debts Midas had incurred (Hyp. 3.11).⁷⁷ One should note that Midas seems to have exercised some degree of economic independence in the running of his perfume business. Apollodorus uses a similar argument to prove that Arethusius owned another slave, Cerdon: whenever Cerdon was the perpetrator or victim of an offence, Arethusius paid the fines or received the compensation respectively ([Dem.] 53.20; cf. Right to income in section II above).⁷⁸ Owners could not use their slaves to commit murder: both were liable to the death penalty in such cases.⁷⁹ The Aristotelian Athenaion Politeia (59.5) also mentions actions in cases where slaves slandered free persons. In several situations the confiscation of slaves from an owner was permissible (liability to execution). One of the most common was distraint on slaves pledged as security for loans in the event of the borrower’s default. There are several examples of slaves pledged as security in the horos inscriptions (IG II² 2747; 2748; 2749; 2751; SEG 51.162; 54.256), but these inscriptions probably underrepresent the frequency of pledging slaves in everyday life (the examples all involve the pledge of slaves in combination with land or buildings; obviously a horos could not be placed on a slave alone!). Forensic oratory provides ⁷⁶ Cf. Lys. 10.19 with the remarks of Todd (2007): 684–5, citing the archaic language of Solon’s statutes and a law concerning slaves and fines. This could, however, be read as involving slave victims rather than perpetrators; if so, then this law would apply to right to security (see section II above). On this Solonian law, see Leão and Rhodes (2015): 53–4. ⁷⁷ Harrison (1968): 175 believes that the zemias . . . kai ta a[dikem]ata relate to delicts but not contracts. But if the law covered only delicts, or if contractual liabilities incurred by the slave at any time automatically transferred to the new owner via sale, why did Athenogenes need to add the small print to the sale contract (3.10 ‘καὶ εἴ τῳ ἄλλῳ ὀφείλει τι Μίδας’)? The fact that the contract had a clause explicitly acknowledging the new owner’s assumption of Midas’ debts ipso facto shows that the law cited at Hyp. 3.22 was relevant to contracts. ⁷⁸ There seem to have been two ways to prosecute in cases where a slave had perpetrated an offence: if the owner had (or was accused of having) given the slave the order to commit the act, the owner was sued directly. If the initiative was the slave’s, then the slave was sued (though the owner bore legal liability for any fines and probably represented him in court). See Dem. 37.51 and Dem. 55.31–3, with Harrison (1968): 174. Cf. [Arist.] Ath.Pol. 59.5. ⁷⁹ See Ant. 1.20. Here, a pallake (surely a slave: see 1.14) administers poison to two men, quite possibly without knowing the nature of the concoction, and causes their deaths. She is tortured on the wheel and killed by the public executioner, and the stepmother who allegedly plotted the murder and provided the potion is on trial for her life. Cf. Is. 8.41. The basic principle in both cases is that a person who plotted to kill was guilty of the killing as well as the actual perpetrator. See Harris (2006): 391–404. In any event, a slave could not ‘get off ’ because he or she was acting on instructions from his or her owner. In one fascinating passage, we hear of an 11-year-old slave boy who had stabbed his owner, and the clear implication is that if the owner had died and there had been doubt about which slave was the killer, all of the slaves in the household would have been killed (Ant. 5.69, much like the circumstances in Tac. Ann. 14.42). For a tongue-in-cheek enumeration of the misfortunes that can befall a rich man, including being murdered by his slaves, see Antiphanes fr. 202 K-A.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

48

Greek Slave Systems

a few further examples: a fragment of Lysias (fr. 1 [Thalheim]) refers to a slave pledged as security for a loan. Demosthenes’ father held a kline (couch) workshop and its twenty slave artisans as security for a loan of 40 mnai; it would appear that this was a kind of antichretic loan since Demosthenes’ father received an annual income of twelve minae from the security (Dem. 27.9). If an Athenian had won a suit and his opponent failed to pay the damages in time, he was permitted to seize items of property from his opponent, including slaves (Dem. 30.27). In Dem. 33.8–9 we hear about a ship and some slaves pledged as security for a loan, which the creditors aimed to seize. Alternatively, a debtor, if unable to repay his loan in cash, might offer a slave in lieu of cash to his creditor ([Dem.] 53.20). State debtors, according to Apollodorus, could have their slaves confiscated ([Dem.] 53.27). In some exceptional circumstances, a slave could be liberated by the state regardless of the consent of his owner, such as in cases where he informed on his owner for certain crimes, or in military emergencies such as the Battle of Arginusae (Xen. Hell. 1.6.24; Ar. Ran. 693–4; Hellanicus FGrHist 323a F 25; cf. Diod. 13.97.1). The basic picture of Athenian slave law that our sources portray is of a system in which the owner’s rights were extensive, though by no means absolute. And as we can see, the basic cross-cultural ‘incidents’ of ownership map accurately onto Athenian legal practices of the fifth–fourth centuries BC.

Neo- and Achaemenid-Era Babylonia Very little in the way of legislation per se exists for this period of Mesopotamian history; our fuller legal codes, such as those of Hammurabi or Lipit-Ištar, come for the most part from an earlier epoch.⁸⁰ To a great extent these deficiencies are negated by an abundance of business documents such as sale and loan contracts, wills, dowry agreements, and so on. We do possess, however, a fragmentary set of laws from Sippar, c.700 BC that mention slavery, and allow us to see that the principle of the right to possess was clearly observed in this culture in relation to slave ownership. LNB 6 sets out rules regarding slave sales: if someone sold a slave woman who did not belong to him, and the rightful owner made a claim to regain possession of his slave, the fraudulent seller had to compensate the buyer for the capital amount paid as set out in the sale document. This clearly shows that Babylonian law recognized the owner’s legitimate title to his slave; it also shows that the rightful owner ⁸⁰ See Roth (1995): 1–9; Westbrook (1995) passim. What follows draws heavily on Dandamaev (1984); the documents to which I refer (on pp. 49–52 below) are translated and discussed by Dandamaev; I provide specific details on pp. 49–52 below.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

49

could regain his slave, which demonstrates his right to security.⁸¹ As in Athens, Babylonian courts were able to determine the status of individuals in disputed cases. Nbn. 1113 describes how status distinctions could be enforced, a case that we will examine in greater detail in Chapter 3: a slave named Bariki-ili lodged a complaint claiming that he was a free man; the case came to court, and Bariki-ili’s owner presented documents that showed that he had been bought, pawned, and then later given away as part of a dowry. The court asked Bariki-ili to produce the document that would prove he had been manumitted; he was unable to do so, and he soon confessed to having twice run away and having made a false claim to being a free man. The court then returned him to his owner.⁸² In this concrete instance we can see a similar picture to that given in LNB 6: Babylonian law protected the master’s rights of ownership over his slave and put procedures in place to keep the slave under the master’s control. Runaway slaves emerge as a well-attested problem in the Babylonian documents, and it was illegal to harbour a fugitive slave, since this transgressed the master’s security of ownership.⁸³ Slave owners in Babylonia during this era had an extensive right to use in respect to their slaves. It was common to brand slaves, which clearly marked them out as the property of a particular person. This seems to have been a standard practice throughout the Persian Empire, and we find it as far from the Babylonian heartland as Egypt.⁸⁴ In Camb. 143 we find a slave who had her master’s name branded on her hand in both Akkadian and Aramaic, which made her easily identifiable among Aramaic readers as well as Babylonians.⁸⁵ Masters could sleep with their slaves and they could chain their slaves up if they proved recalcitrant. One particular solution to slaves repeatedly attempting to escape known from this period was a type of workhouse, where slaves could be chained to their tasks. The master could be thus freed from the concern of constant supervision—for a fee, of course.⁸⁶ Because of the nature of our documentation, we possess little direct evidence for the beating of slaves by their masters (there is no equivalent, for example, of Attic comedy or the generalizing literature on treatment of slaves of the kind which we find in Xenophon), but a good deal of evidence shows that many slaves ran away from their masters, which indicates that mistreatment occurred.⁸⁷ ⁸¹ See Roth (1995): 143–9. One document (Dar. 53) concerns complex pretrial negotiations relating to a case that shares some elements of LNB 6. For a comprehensive study of the case, see Magdalene, Wells, and Wunsch (2008). ⁸² Translation in Dandamaev (1984): 441–2. ⁸³ Snell (2001); Dandamaev (1984): 220–8. ⁸⁴ Driver (1957): 29–30; Porten (1996): 199–201, B33. See also Dandamaev (1984): 229–34; Mendelsohn (1949): 42–50. ⁸⁵ Dandamaev (1984): 133; Beaulieu (2006): 199–200. (My thanks to Heather Baker, who kindly pointed me to Beaulieu’s study.) ⁸⁶ Dandamaev (1984): 235–8. ⁸⁷ See Dandamaev (1984): 220–8; Westbrook (1995): 1670–3; Snell (2001).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

50

Greek Slave Systems

As we might expect, the Babylonian owner had full rights of management over his slave. He could decide how to employ his slave, and whom he might contract the slave to work for if he were to hire out the slave. Slaves worked in many sectors of the Babylonian economy: in agriculture, leatherworking, shoemaking, seal engraving, weaving, sack making, dyeing, bakery, carpentry, building, brewing, metalworking, and as barbers. They were used for domestic work, and sometimes acted as business agents for their masters.⁸⁸ The master could also hire his slave out as a prostitute. In Nbk. 409 we hear of an arrangement by which a Babylonian named Nabu-aḫḫe-iddin hired out his slave girl to a brothel-keeper named Kalba. According to the agreement, the brothel keeper received a quarter of the slave girl’s income for his work as a pimp bringing in clients; the owner received a three-quarter share.⁸⁹ The principle of Right to income is also clear in Babylonian slave law, and can be seen quite explicitly in the documents Nbn. 679 and 682, which concern a slave woman named Amtija employed as a prostitute by her master, Itti-Marduk-balat ụ . In Nbn. 679, the master makes an agreement by which Nur-Sin, evidently a popular client, was to pay him for the privilege of sleeping with Amtija; in a later document (Nbn. 682), the client has changed (this document concerns a certain Guzanu), but the principle remains the same: Guzanu must pay Itti-Marduk-balat ụ three sūt of barley per day to have sex with Amtija. There is no indication whatsoever that Amtija is to be paid for her services.⁹⁰ In Babylonia there also existed an arrangement along the same lines as the apophora arrangement at Athens. This payment was known in Akkadian as mandattu, and, as at Athens, it was the master who ultimately owned any of the possessions accrued through the slave’s work. In Dar. 509 we find the master of one such slave stepping in to safeguard his financial interests and overriding any de facto control the slave may have had over his ‘peculium’. The slave, Madanu-bel-uṣur, had a contractual relationship with a man named Bel-upaḫḫir, and the latter owed him money. When Bel-upaḫḫir refused to pay the debt, Madanu-bel-uṣur’s owner demanded that either Bel-upaḫḫir prove that the contract was false or pay him (i.e. the owner) the full amount owed.⁹¹ As was the case in Attica, successful slaves in Babylonia might purchase slaves of their own from the money they had in their charge, as is attested in a number of documents.⁹² Babylonian masters had a right to capital in respect to their slaves.⁹³ Many slave sale documents contain a warranty clause, which protected the buyer from any undeclared defects, problems, or liabilities a newly purchased slave ⁸⁸ Dandamaev (1984): 246–320. ⁸⁹ Dandamaev (1984): 134. ⁹⁰ Dandamaev (1984): 134. ⁹¹ Dandamaev (1984): 390–1. ⁹² For the Babylonian equivalent of servi vicarii, see Dandamaev (1984): 372–8. ⁹³ For slave prices in Babylonia during our period, see Dandamaev (1984): 246–8; Jursa (2010): 741–5; Oelsner (2014). For taxes on slave sales, see Stolper (1989). For the layout of sale contracts, see Jursa (2005): 367.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

51

might bring with him which were not obvious to the buyer. For example, in the document Nbn. 693, two slave sellers, Aḫija-likin and Ḫ ašdija, provided guarantees when they sold their slave girl, Nana-silim, to cover suits of false claims, vindication suits, or suits claiming the status of a royal slave or a free person.⁹⁴ A similar principle prevailed in Persian Egypt.⁹⁵ Owners could also give or manumit the slave as they wished; and slaves often appear as dowry gifts among well-heeled families.⁹⁶ One document in particular demonstrates that a single slave might be alienated numerous times and pass to a number of different owners in the course of his life. Cyr. 332 describes the history of a slave named Mušezib-Šamaš, who was originally owned by a man (whose name is not preserved) and his wife Ajartu, through a number of legal transactions. He was then sold to a new owner, Nur-Šamaš; later, NurŠamaš gave Mušezib-Šamaš to his wife, Burašu, in place of half a mina of silver which was her dowry. After Nur-Šamaš died, his wife remarried, and her second husband, Tabbanea, pledged Mušezib-Šamaš as security for a loan of half a mina of silver in what appears to be an antichretic arrangement. The loan was evidently paid off, because the document tells us that the couple in turn sold Mušezib-Šamaš to a new owner for one mina and fifty shekels of silver.⁹⁷ Masters could also manumit their slaves. In Nbn. 697 we see a striking parallel to the Greek practice of paramone, which was a conditional manumission by which the slave received his freedom but was legally obligated to provide determined services to his ex-owner.⁹⁸ In our Babylonian example a man named Iqiša drew up a document bestowing freedom upon his slave Rimanni-Bel, on condition that the slave provided him with food and clothing; on receipt of the document, Rimanni-Bel fled, and did not fulfil his obligations. Iqiša cancelled the deal and re-enslaved Rimanni-Bel, giving him as a gift to a woman named Esagil-ramat. A very similar situation can be observed in Cyr. 339, where a woman named Ḫ ibta manumitted her slave Bazuzu on condition that he supply her with a certain amount of food, beer, salt, and wool. Unfortunately, the documentation surviving on manumission is not extensive enough to determine whether or not this Babylonian equivalent of paramone was commonplace.⁹⁹ The duration of slavery in Babylonia was permanent; it encompassed both the incidents of transmissibility and the absence of term. Slaves were bequeathed in wills: in YOS 6 143 we hear of a division of slaves between two beneficiaries: Ištar-mukin-apli and his uncle Nabu-mušetiq-uddi.¹⁰⁰ In Camb. 365, we can see the division of six slaves as well as a sum of money among

⁹⁴ ⁹⁶ ⁹⁷ ⁹⁹ ¹⁰⁰

Dandamaev (1984): 192. ⁹⁵ Seidl (1968): 62–3. See Dandamaev (1984): 207–10; Roth (1991); Baker (2004) nos. 4, 12. Dandamaev (1984): 192. ⁹⁸ See Wunsch and Magdalene (2014). See Dandamaev (1984): 438–55; Westbrook (1995): 1648–51. Dandamaev (1984): 213.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

52

Greek Slave Systems

members of a wealthy Babylonian banking family; and BE 8 123 documents certain items of property left by will (three silver minae and a slave woman) which were to be divided among the beneficiaries.¹⁰¹ This leads us to the incident of absence of term. As property, the Babylonian slave was held indefinitely; although the owner was free to manumit his slave, he was under no compulsion to do so. Sale documents often indicate that a slave shall become the property of the buyer ‘in perpetuity’. In NRVU 27 we hear of an arrangement by which a woman named Ḫ anna gave seven slaves ‘in perpetuity’ to her adopted son and his mother (i.e. they became the owners), but she retained usufruct of the slaves for the remainder of her life.¹⁰² In TMH 2/3 121 we hear of a slave woman named Nana-reṣua pledged as security for a loan, with the proviso that in the event of default, the slave would belong to the creditor ‘for all time’.¹⁰³ The fourth-century Aramaic slave-sale papyri from Samaria found at Wadi Daliyeh also demonstrate this principle, since they stipulate that upon sale a slave becomes property of the buyer and his sons after him in perpetuity.¹⁰⁴ As in Athens, the use of slaves for purposes deemed harmful by the law was prohibited. In YOS 7 189 we can see this principle in action. Two slaves, named Pudija and Ša-Nabu-taqum, physically assaulted a herdsman and stole the flock he was in charge of (which belonged to a temple of Ištar). The owner of the slaves, a man named Kina, was required to produce his slaves before a court, and if he failed to do so, liability lay with him to compensate the temple for the stolen property.¹⁰⁵ Finally, slave ownership in Babylonia entailed a liability to execution. Slaves could be pledged as security for loans, and could be taken by the creditor in the event of default, and became the creditor’s property. In the document TMH 2/3 121 (which we have just considered in relation to absence of term) we can see such a situation: a woman named Aḫušunu pledged her slave Nana-reṣua as security for a loan of one mina and fifty-two shekels of silver, and in the event of default the contract stipulated that Nana-reṣua would belong to the creditor, Nabu-šum-ukin, for all time.¹⁰⁶ Slaves could be confiscated for other reasons. In YOS 3 165: 32–4, for example, we hear of a man who owed taxes to the state and had fled the country along with his son; the governor of Uruk took possession of his slave and confined him to a workhouse.¹⁰⁷ *

* *

¹⁰¹ Dandamaev (1984): 213. See also Porten (1996): 199–201 (B 33). ¹⁰² Dandamaev (1984): 212. ¹⁰³ Dandamaev (1984): 142. ¹⁰⁴ Gropp et al. (2001) no. 1 line 4 (‘‫’לעלמא‬, ‘in perpetuity’). See also no. 2 line 4, no. 3 line 4, no. 4 line 5, no. 5 line 5, no. 6 line 4, no. 7 line 6. For the formula, see Gropp et al. (2001): 39. On these documents, see further Chapter 12, p. 253 n. 36 below. ¹⁰⁵ Dandamaev (1984): 420–8. ¹⁰⁶ Dandamaev (1984): 142; see also Westbrook and Jasnow (2001) passim. ¹⁰⁷ Dandamaev (1984): 237–8.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

53

As these two detailed case studies show, the legal position of owners vis-à-vis their slaves in Athens and Babylonia fulfils all of the criteria for ownership as understood in modern comparative legal theory. Leaving aside metaphorical formulations for now (a subject that we will return to in Chapter 2), it is clear that in pragmatic terms individuals who are termed doulos and oiketes in Greek, and ardu and qallu in Akkadian, do indeed hold a fundamentally similar relationship with their masters: this is a legal relationship of ownership.¹⁰⁸ Our case studies, furthermore, support what Hann has called ‘the embeddedness of property’.¹⁰⁹ In other words, law cannot simply be passed off as an epiphenomenon which bears no relation to social practices: laws are enmeshed with, and emerge from, other aspects of society, be they political, religious, or otherwise; and these factors give rise to the specific institutional forms that slave laws assume. Viewing slave law as an embedded aspect of society, as we shall see, will enable us to make sense of the slave laws of regions such as Sparta and Crete (Chapters 6 and 7). Another insight of social anthropology— that property is not an immutable fact, but a powerful, institutionalized social construct—is worth bearing in mind: The essential nature of property is to be found in social relations rather than in any inherent attributes of the thing or object that we call property. Property, in other words, is not a thing, but a network of social relations that governs the conduct of people with respect to the use and disposition of things.¹¹⁰

If this all sounds too vague and nebulous, let us note Plato’s imaginary scenario in the Republic (9.578e). Here, a rich man with some fifty slaves is all of a sudden whisked off by some god to a desert place along with his slaves and family. Abstracted from the society whose legal institutions have forged for him the enforceable right to treat other individuals as his property, to treat them like dirt, this man—vastly outnumbered—begs and grovels lest his slaves murder him in revenge for his former treatment of them. Once the master and his slaves are taken out of their social and institutional context, they are just human beings, without any inherent differentiating feature. It is the fact that societies construct and enforce legal institutions that enables one man’s tyrannical governance of the lives of fifty of his fellows to be a practical proposition. In other words, ‘slave’ and ‘master’ do not exist as atomized units, or immutable ‘statuses’, but only have meaning once they are viewed as an embedded feature of social relations, made concrete through legal institutions.¹¹¹

¹⁰⁸ Cf. Wunsch and Magdalene (2014): 338. ¹⁰⁹ Hann (1998). ¹¹⁰ Hoebel (1966): 424, cited in Hann (1998): 4. ¹¹¹ By ‘embedded’, I imply the sense adumbrated by Mark Granovetter, e.g. Granovetter (1985), rather than the Polanyian sense that has gained significant currency in the debate over the ancient economy.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

54

Greek Slave Systems

CONCLUSIONS Several conclusions follow from the analysis set out in this chapter. First, it is clear that the two societies we have examined had a working concept of property, even if they did not do us the courtesy of formulating this concept into an abstract jurisprudential theory. Much of what has been written about the Greeks lacking a concept of property or a legal definition of slavery can be dispensed with when we abandon the search for abstract definitions in philosophical or legal texts and focus instead on pragmatic concepts as revealed by law in practice. As Wittgenstein forcefully stated in his Philosophische Untersuchungen (§ 70), ‘When I give the description: “The ground was quite covered with plants”—do you want to say I don’t know what I am talking about until I can give a definition of a plant?’¹¹² We noted above on p. 30 in section I.1 Wood’s claim that the Greeks had no concept of ownership. But the falsity of this view is clear once we shift our attention from trawling the works of Plato and Aristotle for abstract formulations and pay closer attention to the position of slaves as revealed by the laws of Athens and the arguments of Athenian litigants in real court cases. Similarly, the weakness of ZelnickAbramovitz’s claim that the Greeks lacked a juridical concept of slavery becomes clear when we realize that it hinges on a search for the presence or absence of an abstract, jurisprudential definition of slavery in Greek texts. In fact, there was never any need to formulate such a definition, because the rights of slave owners were perfectly clear in Attic law.¹¹³ This is why Aristotle can offhandedly refer to the slave as an article of property imbued with a soul (Pol. 1253b32): he does not need to expend any effort on justifying this view of the slave as an article of property, because this was the common-sense view of slave status. It is crucial to see this problem in the light of practices rather than in terms of theoretical abstractions. It is highly unlikely that the average man on the street in a modern European nation state can produce an abstract theory of ownership of the sort ventured by Honoré. It does not follow, however, that this somehow prevents him from inter alia buying and selling, reporting theft to the authorities, taking out an insurance policy on his property, or inheriting the property of his deceased relatives. The focus on abstract theories of slavery and property has distracted many scholars from the fact that Athenians had a pragmatic, empirically informed concept of slavery that they understood from

¹¹² Cf. Wittgenstein (1963): 33. Cf. Hann (1998): 2: ‘If I ask my ten-year-old son to explain the meaning of the word “property”, he answers promptly that it has to do with owning things. If a thing is your property, it belongs to you. Others cannot take it away from you.’ ¹¹³ The Athenians only defined key legal terms when ambiguity existed about their meanings; otherwise, a workaday conception of their meaning normally sufficed. See Harris (2013a): 166–73.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Ownership and Articulation of Slave Status

55

their everyday dealings and encounters with the law.¹¹⁴ To claim that the presence of an abstract legal formulation of slavery is a sine qua non for a juridical conception of slavery in terms of ownership is no more sensible than claiming that a child cannot learn to speak a language without being drilled in the formal rules of grammar and the intricacies of linguistic theory. We will look at this problem in more detail in Chapter 2, when we turn to the meaning of eleutheria (freedom) in Greek texts. In terms of the broader theme of this book, though, there is a more central point that our above analysis underscores: M. A. Dandamaev was quite correct to write that ‘Babylonian slaves of the seventh to the fourth century, with regard to their legal status, were not notably different from those of the classical world.’¹¹⁵ The alleged qualitative difference between Graeco-Roman and Near Eastern slavery appears much less stark than is often asserted. Two further issues require treatment, however, before we can properly comprehend the real difference in conceptions of slavery between the Greek world and the Near East: first, what difference did the Greek concept of freedom make? And second, could Near Eastern societies make crisp status distinctions of the sort made in the Greek and Roman worlds? These crucial questions are the subject of Chapters 2 and 3.

¹¹⁴ Cf. Harris (2006): 188–9. ¹¹⁵ Dandamaev (1984): 79. Nor was this status a recent invention: Cf. Thompson (2011): 194, 195 and 202, writing of the spread of ‘Greek style chattel slavery’ in the wake of Alexander’s conquest, and at 194 of ‘the introduction of chattel slavery to areas where previously it had not formed part of the culture’. One struggles to pinpoint what is distinctively Greek about this form of slavery, for both the legal regulation of the slave as property of his or her owner and the traffic in slaves from foreign regions was common in the East long before the Greek classical or even archaic periods: see Bayram and Çeçen (1996); van Koppen (2004). As Part III of this book will show, slavery sensu stricto was also extremely widespread.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

2 The Riddle of Freedom As we noted in the Introduction, one of the key concepts that scholars have emphasized when asserting the differences between the Greek world and the Near East is the element of freedom. Finley, for instance, argued that in Greece in the days of Pericles and Rome in the days of Cicero, everyone could be sorted into two categories: free and slave. This sharp dichotomy did not develop in the Near East, so the argument goes, where status formed a continuum, with one dependent status shading into the next.¹ This idea of an indeterminate continuum need not be confused with Finley’s tendency to describe dependent groups in the Greek world in terms of a ‘spectrum’ of statuses running the gamut between full citizen at one end and chattel slave at the other.² Although these views have received some criticism, praise of Finley’s approach has been more vocal.³ Although Finley’s views have been highly influential, there have been other approaches to the issue of freedom that further complicate the issue. As we noted in Chapter 1, R. Zelnick-Abramovitz has recently argued against defining slavery and freedom in juridical terms, an idea that she believes ¹ Finley (1981): 120–1; cf. p. 132. ² Analysing classical Greek dependency using the notion of a ‘spectrum’ of statuses might seem at odds with Finley’s claim that classical Greeks and Romans divided everyone into the categories of free and slave; however, Finley’s ‘spectrum’ approach was not intended as a description of the Graeco-Roman method of articulating status distinctions (i.e. an emic view), but as an observer’s (i.e. etic) approach to representing concrete realities accurately (e.g. Finley 1981: 118: ‘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 two claims are thus not necessarily logically incompatible. Finley’s idea that before the Greek classical period, statuses were blurred into one another (e.g. 1981: 132) can be seen as a substantively different proposition. I may be mistaken, however: Zurbach (2013): 960 has noted that Finley did alter his views over time and is not always clear in saying what he means; whilst Vlassopoulos (2016a): 85–8 calls the existence of both ideas in Finley’s thought ‘schizophrenic contradiction’. For emic and etic categories, see section I below. ³ Criticism: de Ste. Croix (1981): 91–6; 137–8; Garlan (1988): 87–8; van Wees (2009): 451–2; Praise: Hopkins and Roscoe (1978): 137; Vidal-Naquet (1986): 160; Carlsen (2002): 120; Braund (2008): 3; Vlassopoulos (2009): 347; Hunt (2011): 46; Ismard (2011) passim; Kamen (2013): 1–7. Vlassopoulos (2016a): 86 calls Finley’s spectrum idea ‘a step in the right direction’ but levels several criticisms.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

58

Greek Slave Systems

was foreign to Greek thought: ‘It is principally independence of anyone or anything that distinguishes the free from the non-free. The quality of the eleutheros, according to Aristotle, is not to live for another.’⁴ And again, she writes: ‘freedom was some degree or other of the absence of constraint and compulsion; slavery was some degree or other of the absence of freedom. This formulation is not just academic wordplay, but a reflection of reality.’⁵ As with the concept of property treated in Chapter 1, a variety of interpretations of both the meaning and the significance of freedom populate much recent scholarly literature. Amid this cacophony of dissenting voices, it is easy to become perplexed. In order to understand how this issue bears on our comparison of Greece and the Near East, we first need to examine carefully how these terms are used in Greek texts. Once we have established some basic conclusions about Greek usage of freedom terminology, we will be in a position to compare Greece with the Near East, a task taken up in Chapter 3.

I. EMIC AND ETIC CATEGORIES One key distinction utilized in the social sciences, but rarely if ever in discussions on the categorization of dependent statuses in ancient Greece, is the distinction between emic and etic categories. Emic (actors’) categories are those categories used by the subjects under consideration. When studying the ancient Greeks, we must learn about various concepts such as ate, nemesis, and koros, which—though we may translate them into English—have a very different cultural currency in the Greek world. This allows us to understand how the Greeks conceptually organized their world. Etic (observers’) categories are those of the examiner and not of the subject. We might take as an example the idea of ‘the economy’. The Greeks had no analytical category corresponding to our word ‘economy’, but they certainly had an economy in the sense of organized forms of production and exchange, which we can fruitfully investigate without imputing false categories to our subjects. It is no use arguing that we need only use one or the other category: both are important heuristic tools, but they are useful for different tasks, and it will be useful to maintain a clear distinction between them as we proceed.⁶ The issue of freedom is complicated because, broadly speaking, it is a value shared both by the Greeks and by us. It is vitally important, then, that we look ⁴ Zelnick-Abramovitz (2005): 34, quoting Arist. Rhet. 1367a27–8. ⁵ Zelnick-Abramovitz (2005): 38. ⁶ On emic and etic categories, see Scott and Marshall (2005) s.v. ‘Emic and etic analysis’. Cf. Lloyd (1990): 7–20; much useful reflection in Ginzberg (2013).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

59

contextually at what the term eleutheria and related words meant to the Greeks without importing any of our own modern views of freedom into our analysis, or deciding a priori that there is some transcendent and immutable ‘Western’ value of freedom that runs as a common thread throughout European history.⁷ In analysing the Greek terminology for freedom and its usage, close attention is required on two issues in particular. The first is the nature of freedom: Zelnick-Abramovitz has rightly observed that the Greek terminology for freedom is flexible and can be used in a variety of different ways. Based on this observation, she proposes that eleutheria was a vague and fluid concept related to the absence of compulsion. Is this correct? Did eleutheria have a vague, indistinct meaning, or did it have a number of precise meanings grouped around a common theme? Second is the development of the concepts of freedom and slavery from the archaic period down to the fourth century BC. Various opinions have been ventured on this topic: Raaflaub, for example, argues that a political concept of freedom (that is, the absence of an internal or external oppressor in a polity) developed after the Persian Wars as a kind of exercise in ‘othering’ whereby Greek political arrangements were contrasted with those of the Persians in order to construct their own distinct identity.⁸ Vlassopoulos, on the other hand, holds a different view. He rightly observes that the terminology of freedom is inseparable from the terminology of douleia.⁹ But the latter, he believes, meant ‘domination’ to the Greeks, and only came to be focused on the issue of human property after Aristotle’s famous formulation of the slave as ‘animate property’ (Pol. 1253b32). Modern scholars, in his view, have mistakenly seized on the Aristotelian view of douleia as a property relationship, and dismissed other usages as mere metaphors. Are these rather divergent views simply reflections of the muddled nature of the Greek sources? A fresh study of these terms and their usage is required in order for us to solve this riddle and evaluate the arguments of these scholars.

I I . P O L Y S E M Y A N D CO N TE X T Greek literature displays a complex variety of uses for the term eleutheria and related words. In other words, these terms are polysemous, that is, they can bear multiple meanings. It does not follow, however, that these different usages were somehow combined into a nebulous, conglomerate ‘Greek concept of freedom’. Two key points need to be emphasized. First, despite the ⁷ Cf. Raaflaub (2004): 10. ⁸ Raaflaub (2004). ⁹ Vlassopoulos (2011c): 117; cf. Vlassopoulos (2012b), rightly praising Tamiolaki (2010) for viewing the two concepts as inseparable and studying them together.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

60

Greek Slave Systems

terminology for slavery and freedom having multiple meanings, each meaning belongs to a specific context, and a study of their usage shows that these generic boundaries were easily distinguished in practice. In other words, we should not be looking for a single definition of freedom in ancient Greece, but we should rather explore the multiple contexts in which this language could operate, how these were conceptualized, and how they were applied in practice. In essence I am proposing that we take Wittgenstein’s approach to the understanding of language and observe the usage of different terms, rather than trying to root about for one single definition and ignore the contexts in which the terms are applied.¹⁰ Much recent work on the language of slavery in ancient Greece is frustrated by a simple error: scholars rightly recognize that these terms can be applied fairly broadly, but this leads them to a false conclusion, i.e. that the terms in question are ‘vague’, ‘fuzzy’, ‘blurred’, or such like. If we disembed these words from the texts in which they are found and treat them in isolation, then such a conclusion is justified. But observations on the ‘fuzziness’ of these terms effectively ignore the issue of context, which more often than not disambiguates the precise shade of meaning that the word is meant to express. As Umberto Eco has rightly written, ‘A text is a place where the irreducible polysemy of symbols is in fact reduced because in a text symbols are anchored to their context.’¹¹ Second, it is important to recognize that the complex shades of meanings that these terms can bear evolved over time, beginning with a legal relationship and only later being applied in an increasingly creative way to other asymmetrical relationships. It is perhaps best to discuss these points by looking at the development of the concepts of slavery and freedom first and then the context-specific nature of later usages second. In its original sense, the doulos–eleutheros dichotomy related only to personal status, as is evident from several well-known passages of Homer. For example, in the Iliad (6.455) Hector justifies going into battle to his wife Andromache, but laments the day when Troy will fall, a day he knows is not far off. The aspect of this impending disaster that pains him most, he says, is that some bronzearmoured Achaean will take away Andromache’s day of liberty (ἐλεύθερον ἦμαρ) and lead her off to Argos as a slave. She would then be widowed of the man who could fight off her day of slavery (δούλιον ἦμαρ; Il. 6.462–3). Hector later remarks, when standing over the body of Patroclus, ‘Patroclus, you thought perhaps of devastating our city, of stripping from the Trojan women the day of their liberty (ἐλεύθερον ἦμαρ) and dragging them off in ¹⁰ For example, in §§ 66–75 of the Philosophical Investigations, Wittgenstein draws attention to the problem of defining the term ‘game’, putting forward in turn several common features of games, but showing that there are always examples of games that fail to fit the proposed definitions. He points to the usage of the terminology as key: the meaning of a term cannot be understood once the term is abstracted from the context in which it is employed. ¹¹ Eco (1990): 21 (my emphasis).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

61

ships to the beloved land of your fathers’ (Il. 16.830–2, tr. Lattimore). These passages display no ambiguity: they deal only with literal enslavement, and treat it as an abrupt and violent act that transforms a free person into a slave. It is the act of a single day. The same usage is evident in the Odyssey. In spinning his Cretan yarn, Odysseus talks of his voyage with some Thesprotian sailors: when they had travelled out to sea, they devised the ‘day of slavery’ (δούλιον ἦμαρ) for him, dressing him in rags and binding him (Od. 14.340), obviously as a preliminary to selling him for a profit. And at Od. 17.320–3 Eumaeus remarks that Zeus takes away half a man’s arete when the day of slavery (δούλιον ἦμαρ) befalls him. In all of these cases the language of slavery and freedom relates to literal enslavement in war or through kidnapping and reduction to the status of property of the captor.¹² They do not relate to a vague domination of one individual by another, or to personal service for another, or even to an internal, dispositional domination by this or that appetite. Nor is this language related to political entities: despite plentiful opportunities, there is no talk of the enslavement of Troy in the sense of the city as a political entity, nothing to parallel the language of later Greek writers such as Thucydides who spoke of the Athenians enslaving (i.e. politically dominating) their allies.¹³ Holding human beings as property did not, as we saw in Chapter 1, require conceptual leaps associated with written law, state formation, or advanced jurisprudence.¹⁴ It is clearly present in Homer, and the terminology of slavery in Homeric epic is solely concerned with this individual legal status. As we shall see in greater detail in Chapter 5, this condition was deeply asymmetrical, regulated only by customary law. It was also an extremely familiar one: slaves formed the backbone of the elite economy in the archaic period and were a highly visible feature of early archaic society. It is no great surprise, then, that the language of slavery (since slavery was the most conspicuous example of a relationship of asymmetrical power) was gradually extrapolated to other contexts in which one entity exercised extensive power over another. Solon’s poetry provides our first example of the extension of the language of slavery to the sphere of politics. Solon claimed that he removed the boundary stones from the earth, which, previously enslaved, was now free (Solon fr. 36 [West]). ¹² The only possible exception to this trend is the mysterious crater called ἐλεύθερον mentioned at Il. 6.528. Rocca (forthcoming) makes the interesting suggestion that this vessel was used as part of a manumission ritual, and draws attention to water in manumission rituals in later Greek texts. On manumission in Homer more generally, see Zanovello (2016): 22–40. Raaflaub (2004): 26 suggests that this crater was used to mark the departure of the enemy. ¹³ Given that these are the sum total of extant examples found in Homeric epic, I do not understand how Raaflaub can write (2004: 31) that ‘the boundary between free and unfree, marked sharply later on, existed but tended to be blurred’. Such a claim owes more to faith in the writings of M. I. Finley than to what can actually be read in the sources. ¹⁴ Indeed, slavery had a long history before the archaic period. For a recent study of slavery in neolithic Europe, see Testart et al. (2012).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

62

Greek Slave Systems

Historians may argue endlessly about the meaning of the former phrase, but the idea that the earth was ‘once enslaved’ and is now ‘free’ cannot be interpreted in the juridical, property sense.¹⁵ Likewise, Solon warned his fellow Athenians of the slavery (doulosyne) under which they would fall if a tyrant rose to power in Athens (fr. 9 [West]). Since it is hardly probable that a tyrant would or could literally enslave all of the Athenians in a legal sense, we are dealing here once more with a political usage of the terminology of slavery. Similarly, Anacreon (fr. 419 [West]) praises his friend Aristoclides, who died warding off slavery from his homeland—once again, to be sure, a political metaphor. These notions were extremely common in fifth- and fourth-century Greek political discourse.¹⁶ In Thucydides the Athenians are represented as having enslaved their allies, and the rallying cry of the Spartans in 431 BC was to free Greece from Athenian subjugation (Thuc. 4.85). These terms clearly do not denote legal slavery, but isolate the aspects of asymmetry of power and the element of domination inherent in the relationship of legal slavery and apply it to the political sphere. In one passage, Thucydides, in describing the growth of Athenian power, uses two different verbs to make clear the distinction between the juridical meaning of slavery and its political usage: the Athenians captured Eion and Scyrus, selling their inhabitants into slavery (ἠνδραπόδισαν), but politically subjected (ἐδουλώθη) Naxos (Thuc. 1.98).¹⁷ The plasticity of this political usage of the ideas of freedom and slavery was such that when the Athenians endeavoured to revive their maritime empire in the fourth century and attract member states to their league with a guarantee of ‘freedom’, they

¹⁵ The most methodologically streamlined and sensible explanation of the removal of the horoi is, to my mind, Harris (1997). ¹⁶ For an in-depth study, see Brock (2013): 107–96. ¹⁷ The same distinction can be found in Herodotus: see Brock (2013): 109. Vlassopoulos (2011c): 119 argues for a functional distinction between δοῦλος and ἀνδράποδον and the related verbs δουλόω and ἀνδραποδίζω: ‘The different semantic fields [sic] of doulos and andrapodon can best be seen in the use of the related verbs douloô and andrapodizô . . . the verb douloô describes a relationship of power; andrapodizô a relationship of property.’ Whilst I agree that the latter verb is generally used in a specific sense in relation to property (often in connection with enslavement in war), it does not follow that the former verb never has a legal meaning; besides, the noun δοῦλος in Thuc. 8.28.3–4, cited by Vlassopoulos on p. 119, is used in a technical legal sense to distinguish slaves from free people among a group that has been subjected to andrapodismos. Since all of the captives are subject to domination and are all equally under the power of their captors, the word δοῦλος must be used here in terms of a legal status. This distinction is not unique to Thucydides but was common across the Greek world. Thus, we find the same distinction made in, for example, Amorgos (SIG³ 521) and Crete (IC I Knossos no. 6) during the third century BC with reference to the victims of kidnapping, who, though all are kidnap victims and thus potential slaves, are discussed in terms of the legal statuses they held prior to capture. Cf. Bielman (1999): 180–1, ‘en Grèce, le statut juridique n’était pas affecté par la capture. Le citoyen privé momentanément de liberté demeurait théoriquement un citoyen: on rencontre ainsi des captifs appelés des politika sômata (« personnes appartenant au corps de la cité ») ou des éleuthéra sômata (« personnes de condition libre »).’ Here, I should emphasize that I have rethought the position I expressed in Lewis (2016): 50–1 n. 15.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

63

were forced to be precise about the exact content that this advertised ‘freedom’ entailed, given their dubious track record in the previous century: allied states were to have their own constitutions and self-government, would not have Athenian garrisons foisted on them, and did not have to pay tribute (IG II² 43.19–44). This language was extended to social contexts as well. At Athens, working for another person was seen as slavish: the reality of depending for any length of time on another citizen for wages, taking orders, and being at another’s beck and call ran contrary to the civic ideology of egalitarianism, where at least theoretically all citizens existed on a level playing field. For example, Socrates, in a well-known anecdote related in Xenophon’s Memorabilia (2.8), bumps into an old acquaintance named Eutherus who has hit rock bottom; his property, located abroad, has been lost in the war, leaving him destitute. Socrates advises Eutherus to apply to a fellow citizen for work as a bailiff on a farm, but Eutherus counters that he would not consider doing so as it would mean becoming subjected to slavery (χαλεπῶς ἄν, ἔφη, ἐγώ, ὦ Σώκρατες, δουλείαν ὑπομείναιμι). Obviously, Eutherus is not speaking in juridical terms: Athenians had been banned from enslaving fellow Athenians since the time of Solon. Rather, we find here a further extrapolation of the originally narrow, legal meaning of ‘slavery’ into the realm of asymmetrical social relations. The language of slavery can also be applied to those who practise ignoble occupations of the sort often practised by slaves. Thus, in another story from the Memorabilia, Socrates asks Euthydemus whether he knows that some people are referred to as ‘slavish’ (ἀνδραποδώδης). Euthydemus replies that he does, and Socrates proceeds to list several ‘slavish’ occupations: smiths, cobblers, and carpenters (Xen. Mem. 4.2.22). Xenophon expands on this idea in the Oeconomicus, writing that certain occupations are called ‘banausic’, render the body unfit for military service, and provide their practitioners with too little leisure to cultivate friendships (Xen. Oec. 4.2–4). Aristotle too discusses this prejudice in his discussion on citizenship in the Politics, and argues that such people are slave-like and should not be admitted into the ranks of the citizenry (Arist. Pol. 1277b33–1278a13). It is in relation to such ‘servile’ occupations that he claims that the free man is one who does not live for another (Rhet. 1367a27–8). Other fourth-century texts take this sense of the ‘free man’ in terms of the model citizen: Theophrastus, in characterizing the man of aneleutheria, or illiberal temperament, depicts a skinflint who refuses to spend money on either himself or the state (cf. [Arist]. Virt. 1251b); the implication is that the man of a liberal temperament will give generously to others and maintain himself in style. Once again, this usage is only loosely related to the strict legal meaning of eleutheria. There was yet another context in which the language of slavery could be invoked: in philosophical discourse it is often related to dispositional vices. In

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

64

Greek Slave Systems

a philosophical context a man can be said to be enslaved to one of his passions or appetites that exercises despotic power over him. In Plato’s Symposium, Pausanias’ speech on love describes the degrading ways in which men behave towards their boyfriends: they beg and beseech them, swear oaths to their promises, sleep in their doorways, and submit to a slavery that even slaves would not endure (καὶ ἐθέλοντες δουλείας δουλεύειν οἵας οὐδ᾽ ἂν δοῦλος οὐδείς: Pl. Sym. 183a; cf. 184c, 185a). In Xenophon’s Memorabilia (1.3.9–11), Socrates notes how Critobulus, once a sober-minded man, has by kissing Alcibiades’ attractive son risked turning from a free man into a slave.¹⁸ In another story in the Memorabilia, Xenophon notes that Socrates possessed the greatest selfcontrol, and remained master over his desires for carnal pleasure as well as the lust for money, which, if allowed free rein, subjects the sufferer to the worst kind of slavery (Xen. Mem. 1.5.6).¹⁹ And in an important passage in Xenophon’s Oeconomicus, Socrates describes at length the various passions that could enthral a person and subject them to a kind of inner slavery (Xen. Oec. 13–23). Similar usages of slave imagery can be found in the writings of Democritus of Abdera (68 F 50, F 214, F 251 D-K). Two conclusions follow from this review of the evidence. First, the multifarious usages of douleia and related terms were not present in Greek culture from the outset. They stemmed from a property relationship concerned with personal status that is visible in the Homeric poems (and eleutheria as an antonym to this, i.e. for all those who were not slaves), and over time its conspicuousness as the asymmetrical power relationship par excellence meant that its terminology could be enlisted and mapped onto a wider variety of asymmetrical relationships in a number of extralegal contexts.²⁰ In other words, there is a ‘family resemblance’ regarding the usage of these terms built around the idea of asymmetrical power relations.²¹ But the property element was present from the outset and was the kernel from which the other usages sprang.²² In other words, we are looking at a process that was essentially ¹⁸ On slavery and love, see Tamiolaki (2010): 386–90. ¹⁹ On Socrates’ self-restraint, see Tamiolaki (2010): 371–94. ²⁰ Cf. Samuel (1965): 243–4 on the evolution of the usage of the verb paramenein. On the development of metaphor as a process of domain mapping, see Lakoff (2008); Gentner and Bowdle (2008). ²¹ Cf. Wittgenstein (1963): 32 § 67. As a parallel, cf. the discussion of Cairns and Fulkerson (2015): 12–14 on the term aidos. This term can mean ‘shame’ but it can also mean ‘deference’, ‘respect for others’. These seem like rather separate issues to the modern Briton; but the conceptual link (or family resemblance) between the two lies in the issue of honour, time. Shame concerns one’s own honour, deference the honour of others; so the two meanings of the word cluster around a single central concept. ²² Pace Vlassopoulos (2011c): 121–2, who argues that Aristotle was innovating when he called the slave a ktema. Cf. Thalmann (1998a): 53–4, who points out that passages such as Od. 7.225, Od. 19.525–9, and Il. 19.33 show that the conception of slaves as articles of property is present in Homer: ‘in all three places, the line reflects a view of dmōes (or dmōiai) as forming part of the property, just as a house does. That is, in a way that anticipates Aristotle’s definition of a slave,

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

65

the opposite of what Finley thought occurred: instead of a fluid, complex archaic situation that later resolved into a black-and-white legal distinction, we start with a clear legal distinction that, over time, became more complex (though not necessarily more confused). The gradual process presented above in which the terms doulos and eleutheros (and related words) were extended in meaning beyond their earlier, Homeric semantic range is no accident of survival: had philosophical or political concepts of freedom and slavery existed c.700 BC, there would have been plenty of opportunities for them to be deployed in the Iliad— the fall of Troy, for example, or the anger of Achilles, the lust of Paris, or the greed of Agamemnon. But in no case do we find the language of slavery and freedom applied to these themes. The argument from silence is, in this case, a strong one: in the Homeric period these other usages of slavery and freedom terminology had not yet developed. At any rate, the process of development sketched above is in many ways unremarkable and follows a common path of semantic change well known to modern linguistic theory. As McMahon writes: Metaphor . . . involves similarity of meanings. An imagined link is established between two concepts, allowing the transfer of a label from one to the other: the human foot is the lowest part of the body, just as the foot of a hill is the lowest part of a hill. People may be owlish, mulish or catty, if their behaviour matches some (real or imagined) characteristic of the animal concerned. Metaphors often shift meanings from concrete to abstract; thus grasp means ‘to take hold of something’ mentally as well as physically, while to cast light on something may mean ‘to make something understandable’ as well as ‘to make something visible’. Eliminate, from Latin elīmīnare ‘to put out of the house’, exhibits a similar metaphorical extension. The transfer of a word or phrase from specialised into general use may also entail metaphor; outside discussions of space travel, French être sur orbite ‘to be in orbit’ now means ‘to be very successful’.²³

When we encounter extralegal usages of the vocabulary of slavery and freedom in classical Greek, one option is to refer to them as metaphors. Vlassopoulos has objected to this categorization, writing that it is ‘totally misconceived’. He points to the concrete social consequences of conditions in which slavery language is invoked: ‘there is nothing metaphorical about the use of doulos to describe a member of the Athenian Empire’.²⁴ In a certain sense, Vlassopoulos is right: if we take ‘metaphor’ in its vulgar meaning, that is, as simply a literary device, it is obvious that the circumstances that he describes are trivialized by the label when they in fact they carry considerable social and historical import.²⁵

they are living property.’ Cf. Ndoye (2010): 207. On Homeric slaves as property, see further Chapter 5. ²³ McMahon (1994): 182–3. ²⁴ Vlassopoulos (2011c): 118. ²⁵ On the vulgar sense of the word, see Lakoff and Johnson (1980): 3.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

66

Greek Slave Systems

However, there are good reasons to retain the term ‘metaphor’ for the extralegal usages of freedom and slavery; for in a technical sense that is precisely what they are. In their groundbreaking work Metaphors We Live By (1980), George Lakoff (a specialist in linguistics) and Mark Johnson (a philosopher) proved that metaphor is anything but trivial and inconsequential, and far more than just a literary device.²⁶ Not only is metaphor surprisingly pervasive in spoken language, but it is also an embedded feature of human cognition: humans think relationally and by analogy, and metaphorical language shows the logical connections forged between different areas of reality as experienced by the brain. Metaphor shows analogous thinking in action.²⁷ It is precisely this process that can be seen in the evolution of the language of slavery and freedom in Greek texts: the relationship between master and slave in a concrete personal sense is, over time, extrapolated through metaphorical extension to analogous contexts where asymmetries of power are also in play.²⁸ At any rate, we have seen that the extension of meaning from the concrete, status connotations of doulos and eleutheros to other areas of life over time reflects a natural linguistic progression. It cannot be overemphasized that these terms, though malleable, function at a number of registers; there is not one nebulous meaning of douleia but several distinct meanings clustered around a common theme; the same goes for eleutheria and related terms. There is nothing particularly surprising or unusual in this: when we speak of terms such as ‘liquidity’ in modern Britain, we mean something very different in a conversation about finances from a conversation on chemistry. Context allows us to be precise on which meaning is intended, and there is no reason to suppose that the same principle does not apply to the Greek usage of terminology for freedom and slavery. As Borowsky and Masson point out regarding polysemous words: When reading text, the context provided by the preceding words and sentences provides a means of disambiguating such words. As a result, we may not even notice the ambiguity in words that we are reading in context. If, on the other

²⁶ Lakoff and Johnson (1980). ²⁷ This ‘thinking by analogy’ appears to be powerfully wired into the human mind: the neuroscientist David Eagleman, for instance, points out how contact with hot or cold substances can condition ‘warm’ or ‘frosty’ feelings towards other persons (Eagleman 2015: 95), whilst scientists have recently shown the strong associations between physical and moral uncleanliness (Chen-Bo Zhong and Liljenquist 2006; Lee and Schwartz 2010). For an introduction to how metaphor is based in the physical circuitry of the brain, see Lakoff (2008). ²⁸ It seems to me that Garnsey (1996): 16–19 and Lavan (2013): 75–6 correctly capture the essence of the phenomenon: these metaphorical usages are conceptually distinct from legal slavery (insofar as the two usages are not muddled together), but the force and resonance of the metaphor depends on quotidian acquaintance with legal slavery.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

67

hand, semantically ambiguous words are presented in isolation, their alternative meanings are readily accessible, and thus their ambiguous nature is noticed.²⁹

III. CONCEPTS I N CONTEXT: FREEDOM AND SLAVERY I N THE ATHENIAN COURTROOM Whilst the claims made in section II may be theoretically coherent, they need to be tested against the evidence in order for their validity to be ascertained. To do so we may turn to three well-known Attic trials, represented by the speeches Against Pancleon by Lysias (Lys. 23), Against Neaera by Apollodorus ([Dem.] 59), and Against Euboulides by Demosthenes (Dem. 57). In Against Pancleon the litigant who delivered the speech tells us that he had been the victim of some manner of connivances by a man named Pancleon and had decided to proceed against him with a lawsuit (23.2). Thinking him to be a metic, the litigant summoned him before the polemarch; but Pancleon claimed to be a Plataean (a group that had been inducted into Athenian citizenship) and to belong to the deme of Decelea. The litigant investigated further and found that Pancleon was unknown to either the Deceleans or the Plataeans in town, and had been involved in other lawsuits under the jurisdiction of the polemarch, hinting at a non-citizen status (23.3–6). One of the Plataeans, a man named Nicomedes, claimed that he had a slave named Pancleon, who had absconded (23.7), and later arrested Pancleon as that very man (23.9). Pancleon’s friends used the aphairesis eis eleutherian procedure to halt the arrest, for which they had to provide sureties. But remarkably, instead of the sureties coming forward, a woman appeared, asserting that Pancleon was in fact her slave (23.10). In his own prosecution of Pancleon, the litigant who delivered the speech pointed out that in an earlier case Pancleon had been proven not to be a Plataean, had lost his case, and had not pursued the matter—suspicious behaviour for a person who maintained that he was a citizen (23.14). Now, for our purposes it is irrelevant whether the litigant’s version of events is strictly true:³⁰ more important is the fact that he uses specific arguments to convince the judges of Pancleon’s dubious claims to citizen status, and avoids other arguments. Arguing on a point of juridical status, he sticks to points of fact and law. Pancleon’s status is considered in pragmatic legal terms: had Nicomedes or the anonymous woman ever owned Pancleon? Could his sureties prove that the claims of these two individuals were false? ²⁹ Borowsky and Masson (1996): 63. On the importance of context, see also Lakoff (2008): 23–4. ³⁰ Cf. Vlassopoulos (2007b): 34 n. 10.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

68

Greek Slave Systems

Did Pancleon’s previous lawsuits undermine his claim to be a Plataean, and thus a citizen? Did the other Plataeans in town know him? Conspicuous by their absence are any arguments for slavery or freedom based on ‘service for another’, ‘domination’, or belonging to a ‘slavish’ occupation, because the dicasts evidently knew that these usages of the language of ‘slavery’ or ‘freedom’ were irrelevant to the task of determining Pancleon’s legal status. The same point can be made in relation to the Against Neaera. In this speech, the litigant attempts to prove that Neaera was an ex-slave and former prostitute who had been living in marriage with an Athenian, Stephanus, and was passing off her illegitimate children as his own legitimate Athenian progeny. Again, when it comes to proving that Neaera was a slave, the litigant sticks to pragmatic facts and the laws, providing witnesses to Neaera’s life as a slave prostitute ([Dem.] 59.23–5, 28) and her manumission ([Dem] 59.32). He again brings witnesses to show that Neaera’s daughter Phano had been passed off as an Athenian virgin and married to the archon basileus, before the latter found out the truth and the Areopagus conducted an enquiry that settled the point ([Dem.] 59.84). Once more, when issues of personal status are being tried in Athenian courtrooms, litigants rarely drag in the various shades of meaning of slavery and freedom that were appropriate to other contexts such as philosophical, political, or social discussions: rather, the litigant in this case deals only with pragmatic facts relevant to legal status, mentioning details such as specific individuals who had hired her from her owner ([Dem.] 59.26), how much money had been paid for Neaera by two Corinthians ([Dem.] 59.29), and how much they charged for her manumission ([Dem.] 59.30).³¹ The strongest example of the pragmatic Athenian approach to determining legal status can be found in Demosthenes’ speech Against Euboulides, in which the litigant, Euxitheus, seeks to defend his citizen status from aspersions cast by another citizen, Euboulides, who inter alia drew attention to his mother’s ‘slavish’ occupation as a proof that she was not a citizen, and therefore that he should be ejected from the ranks of the citizenry. As Euxitheus protests (Dem. 57.34): He claims her to be a ribbon seller and that everyone has seen her. I suppose then that there are many who can witness as to who she is from knowledge, and not just from rumour. But if she was a foreigner, they should have scrutinized the tax records in the marketplace to see if she paid the foreigner tax, and point out the country from which she came; and if she was a slave, then her buyer should have come forward to bear witness, or, in lieu of him, her seller; or in lieu of either of them, someone else who knew that she had been a slave or had been set free. Now, he [i.e. Euboulides] has produced none of these, as I see it, but has instead only

³¹ For the Athenians’ sensitivity to context and genre regarding the courts, see also Harris (2013b).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

69

engaged in abuse. For this is what a sykophantes is, alleging many things, but proving nothing.³²

Here we have an exact demonstration of the difference between a mere ‘slavish’ occupation and slavery sensu stricto, the explicit contrast between the two conditions from the mouth of an Athenian litigant, and precisely the sort of pragmatic approach to status distinctions as that outlined above. As Euxitheus points out, poverty compelled many free men to perform slavish deeds (πολλὰ δουλικὰ καὶ ταπεινὰ πράγματα τοὺς ἐλευθέρους ἡ πενία βιάζεται ποιεῖν: Dem. 57.45), but there is a clear distinction between slavish deeds and legal slave status. He rightly points out that though his family had been engaged in the ‘slavish’ business of selling ribbons, this has nothing whatsoever to do with the technical matter of his descent from free Athenian citizens (Dem. 57.35); and that to prove that his parents had been either metics or slaves, one would have to consult metic tax records (to prove metic status) or records of sale or manumission (to prove slave status). One could not find a clearer example of the competence of Athenians in disentangling the legal meaning of slavery from its extralegal usages. In arguing this case, I do not claim that legal status was immediately distinguishable at sight for individuals either in classical Athens or elsewhere. Even individuals in Athens with some of the most visibly ‘slavish’ features— such as certain kinds of tattoo or a foreign accent—might turn out on inquiry not to have been slaves, but freedmen or visiting foreigners. Indeed many slaves and metics simply blended in with the majority of the Athenian working population. The Old Oligarch famously wrote that one could not distinguish between slaves and Athenians at sight ([Xen.] Ath.Pol. 1.11), which is certainly true in a selective sense, though we should not make too much of it as a generalization for all slaves in Attica. But it is vital that we understand precisely why slaves might have been mistaken for free persons: this was not because of some putative failure in the conceptual apparatus of the Athenian mind, unable somehow to grasp the distinction between legal slavery and other uses of the terminology, but because of the concrete and practical problems of the sort highlighted in the Pancleon case: to identify an individual’s legal status, one required information that was not always, or perhaps even usually, available at sight.³³ Thus, in the Against Euboulides the reason why we do find extralegal facets of slavery (i.e. the ‘slavish’ features of Euxitheus’ parents) dragged into the case is not because Euboulides thinks he can hoodwink the judges by exploiting their (putative) inability to disentangle the legal and extralegal meanings of slavery. Rather, the difficulty in

³² On the term sykophantes and its connotations, see Harvey (1990). ³³ On these issues, see the fine analysis of Vlassopoulos (2007b) and, for information-related difficulties in determining Pancleon’s status, Sobak (2015): 682–9.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

70

Greek Slave Systems

obtaining secure information about one’s legal status allows Euboulides to exploit his opponent’s unfortunate background and play on stereotype and popular prejudice to engender a preponderance of doubt among the judges as to Euxitheus’ citizen credentials (much as Aeschines relied on rumour and stereotype rather than evidence to secure his conviction of Timarchus). Ultimately, it is necessary make a strong distinction between conceptual blurring, on the one hand, and visual indistinctiveness and problems of information gathering, on the other. It is the latter, not the former, that created the problems we find in our sources. It may seem that I am labouring an obvious point, but it will become clear why this is crucial when we turn to specific instances where scholars have invoked the vague social register of freedom to analyse juridical status, in other words, applying a meaning of freedom that is specialized to one context to analyse a phenomenon pertinent to a different context. I will limit myself to two examples: first, Zelnick-Abramovitz’s claim that the Greeks lacked a juridical distinction between slavery and freedom; second, Finley’s claim that some groups lay ‘between the free men and slaves’, and that we should conceptualize status in the Greek world in terms of a spectrum running between full freedom and complete slavery.

IV. F REEDOM AND MANUMISSION As we noted earlier in this chapter, on pp. 57–8, Zelnick-Abramovitz, in her recent monograph on manumission in the Greek world, rejects a legal distinction between free men and slaves as ‘futile’ and adopts a sociological approach inspired by Patterson.³⁴ But does this really reflect a genuine vagueness and indeterminacy in the Greek understanding of slavery? One should note a key point that is missing from her quotation of Aristotle’s Rhetoric: the formulation of ‘not living for another’ follows a discussion in which Aristotle considers the occupations suitable for a free man, noting that in Sparta men wear their hair long, since long hair is impractical for the performance of the work of a hireling. It is only then that he says that one should not carry out the work of a craftsman, since a free man does not live for another. In this passage Aristotle is not concocting a general definition of freedom suitable to all contexts, but merely commenting on a specific context in which the terminology of slavery and freedom might be invoked: the situation of working in an ignoble occupation. This, as we noted on p. 63 in section II above, is present in the writings of Xenophon and does not represent an original insight. ³⁴ Zelnick-Abramovitz (2005): 38; 9.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

71

Yet, assuming that this explanation of freedom is an axiomatic definition suitable for both social and legal analysis, Zelnick-Abramovitz proceeds to analyse Greek manumission documents, with misleading results. For example, she refers to all manumissions with paramone obligations³⁵ as ‘delayed manumissions’ because the freedman is not ‘wholly free’ until the paramone obligations have been cancelled: her point is that since the manumitted slave remains subject to his ex-master through the obligations attached to his manumission, he is not wholly a free man.³⁶ This fudges the legal reality of the situation by describing it in social terms only, and denying the existence of the legal register of slave terminology. Manumission—even with obligations— profoundly altered the legal relationship between the two parties. After manumission the master could no longer sell, pledge, or bequeath the freedman, even if he retained the right to his services and enforced his domicile.³⁷ In fact, if we contrast the legal position of slaves and freedmen with paramone obligations at Delphi, the contrast is very clear. Manumission was often paid for out of the savings of slaves, but formally speaking the slaves could not have title to this money as ownership rested with their masters. In pragmatic terms, then, a slave could not legally pay for his own freedom. This necessitated an elaborate procedure where the slave entrusted the money for the ‘sale’ to the god Apollo, who could be a valid party to the manumission transaction. The god (via the agency of his priests) then paid the master for the slave’s freedom.³⁸ The contrast with freedmen (apeleutheroi) bound by paramone obligations is a stark one. First, some paramone clauses stipulate apolysis, or the possibility of early release from paramone, if the freedman pays a certain sum of money to his ex-owner. In these clauses, it is perfectly clear that the freedman can enjoy legal title to the money paid for apolysis. Second, in some clauses legal remedies are mentioned in the event that the freedman is accused of not fulfilling his paramone duties. Both parties (freedman and exowner) state their cases before arbitrators whose decision is equally binding on both of them. These facts show very clearly that the freedman was understood to be a legally capable free person.³⁹ Zelnick-Abramovitz’s use of the loose sense of ‘freedom’ that she has culled from Aristotle leads her to elide these two conditions and to play down the significant legal differences between slaves and freedmen with paramone obligations. She is correct to point out that in a (certain non-technical) sense freedmen with paramone obligations

³⁵ These were legal obligations attached to manumission that required the freedman or freedwoman to remain in the household of his or her ex-owner, often for specified periods of time, and provide determinate services. See Samuel (1965); Zanovello (2016): 62–71. ³⁶ Zelnick-Abramovitz (2005): 222–3. ³⁷ Cf. Canevaro and Lewis (2014); Lewis (forthcoming a). ³⁸ Zanovello (2016): 44–61. ³⁹ See further Zanovello (2016): 62–82.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

72

Greek Slave Systems

were ‘less free’ than those without, but this is only one sense of eleutheria, and not the one that Greeks would apply in a juridical context.⁴⁰

V. THE ‘SPECTRUM’ APPROACH TO DEPENDENT LABOUR IN ANCIENT GREECE A very similar approach to freedom can be found in several essays of M. I. Finley. Although on one occasion he derided any attempt to define freedom as doomed to failure,⁴¹ his two articles on dependent statuses in the Greek world implicitly use a similar conception of freedom to that employed by Zelnick-Abramovitz: the idea that a free man is one who does not live under the compulsion of another. It should be stressed that whilst this assumption underpinned Finley’s analysis, he shied away from explicitly stating what he meant by the term ‘freedom’. And rather than seeing this formulation as a conception of freedom (indeed, one tailored to a specific context) or admitting that in Greek texts there are different meanings of freedom that are linked to specific contexts, Finley treats this view as the Greek, emic conception of freedom to the exclusion of any alternatives. Having implicitly accepted this notion of freedom, he then applies it in terms of a graded scale: at one end is the chattel slave and at the other the full citizen. In between, he locates various ‘halfway’ statuses, such as debt bondsmen, helots, and manumitted slaves with paramone obligations.⁴² Those whose status binds them more closely and oppressively to a master cluster towards the ‘slavery’ end of the spectrum, and those who have fewer constraints on their actions cluster towards the ‘freedom’ end of the spectrum. A close look at Finley’s arguments shows, however, that the ‘spectrum’ is indeed (to quote G. E. M. de Ste. Croix) an ‘unfortunate metaphor’, and one that obscures more than clarifies dependent status in ancient Greece.⁴³ Let us begin with one of Finley’s ‘halfway statuses’ that we have already encountered: the manumitted slave whose emancipation entailed the acceptance of legal obligations known as paramone. Our earliest Greek evidence for this practice comes from late classical and early Hellenistic Athens, and since we have already encountered the Delphic evidence, it may be useful to view the ⁴⁰ In an important recent article Joshua Sosin (2015: 328) rightly points out that ‘there were no juridical monsters, half-free, half-slave or otherwise’. On his view that freedmen held under paramone were still slaves, though, see the criticisms of Zanovello (2016): 64 n. 192. ⁴¹ Finley (1981): 77. ⁴² For the formula found in Pollux (3.83) describing certain groups as ‘between free men and slaves’ that serves as a starting point for Finley (1960), see Chapter 6, section III, pp. 143–6 below. ⁴³ De Ste. Croix (1981): 92.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

73

problem from an Athenian perspective as well.⁴⁴ Harpocration (s.v. dike apostasiou) mentions that a manumitter can bring a charge of apostasiou (‘desertion’) against his manumitted slave if that slave absconded, registered a different prostates, or failed to do ‘what the laws prescribe’ (καὶ ἃ κελεύουσιν οἱ νόμοι μὴ ποιῶσιν). If the freedman won the suit, he became ‘fully free’ (τοὺς δὲ νικήσαντας τελέως ἤδη ἐλευθέρους), even though he was already a free man. Here, we find the inelegant but unavoidable combination of two specific meanings of freedom: the juridical sense of freedom and the sense of freedom mentioned in section II on p. 63 above (as illustrated in Xenophon’s story of Eutherus in the Memorabilia) which relates to living at the beck and call of another. It is clear, legally speaking, that the slave ceased to be his owner’s property upon manumission: he was free, from a juridical point of view, and could not (for example) be pledged or sold by his ex-owner. But since he could be lumbered with legal obligations that kept him under the domination of his ex-owner, he could be seen from the point of view of Aristotle’s formulation in the Rhetoric to be not ‘fully free’. By acquittal in a case of apostasiou, his obligations were cancelled and he was no longer under the control of his exowner: he was free in both senses of the word. Finley views these cases as proof of a confused view of freedom among the Greeks.⁴⁵ But from a juridical point of view, the former was a slave, the latter a free person: this is why Harpocration can call the freedman with obligations an apeleutheros and the freedman without obligations ‘fully free’—he is overlapping two meanings of the notion of freedom and linking two contexts together.⁴⁶ A further example, Gortyn on Crete, illustrates both an ancient Greek (i.e. emic) approach to categorizing dependent status and the confusion rather than clarification that results if we follow Finley’s ‘spectrum’ method. Finley draws attention to the status called katakeimenos, or debt bondsman, in

⁴⁴ See Canevaro and Lewis (2014) for a full study of paramone in fourth- and third-century Athens. ⁴⁵ Finley (1981): 141. ⁴⁶ Finley (1981): 141, after citing a variety of these status problems, wrote that ‘the evidence is unmistakable that in all sorts of terminology the Greeks recognised the existence of halfway statuses’. This is simply not true. For example, one of Finley’s examples concerns the girl Plangon from Menander’s Heros, who is held in debt bondage, her late father having been a manumitted slave (see Harris 2006: 256–8 for the context). Upon one of the characters, Getas, asking another whether Plangon was a slave, his interlocutor, Daos, replies ‘yes – well, sort of – in a certain manner’ (ΓΕΤΑΣ: δούλη’στιν; ΔΑΟΣ: οὕτως, ἡσυχῇ, τρόπον τινά; Men. Heros 20). Plangon is indeed a slave in the sense that Zelnick-Abramovitz employs, as noted earlier in this chapter, on p. 58 above, but she is not a slave in the legal sense. All this shows is that the Athenians could recognize that an individual could be a slave if viewed from one point of view (or in one specific register) but not from another, i.e. the legal register. The very fact that Daos says τρόπον τινά is good evidence that simply ‘living under the compulsion of another’ was not in itself an adequate criterion for someone to qualify as a slave sensu stricto in Athenian eyes and that the emic view that Zelnick-Abramovitz and Finley claim to be elucidating is in fact no such thing (or to give them their fair due, only part of the picture).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

74

Greek Slave Systems

Gortynian law. This group, he asserts, ‘were somehow neither free nor slave’.⁴⁷ That was not how the Gortynians viewed this condition. In Gortyn, different rules applied to the debt bondsman depending on whether he was juridically speaking a free man or a slave.⁴⁸ This juridical division of all members of society into two categories, free and slave, based on the criterion of property, is evident in several passages in Gortyn’s laws (e.g. IC IV 72 I 2, 4–5, 9, 15–16). Among Gortyn’s legal inscriptions we find certain rules that applied to a free man who was unable to pay his debts: he became a katakeimenos (debt bondsman) of his katathemenos (temporary master, i.e. his creditor). We possess a fragmentary rule governing cases where the bondsman commits an offence, and a complete rule governing cases where he is the victim of an offence. In the latter case, the katathemenos sued the perpetrator on the bondsman’s behalf, and if successful, they split the winnings. If the katathemenos neglected to sue, the bondsman could only bring a suit once he had amortized his debt (IC IV 41 VI). But rather remarkably, Finley fails to mention another inscription from Gortyn that treats the case of an indebted man giving his slave into debt bondage to repay his creditor.⁴⁹ In this law (IC IV 47), quite different rules apply to slaves from those governing the free man: if the slave katakeimenos commits an offence against a third party, his original owner is held liable provided the slave acted on his own initiative; if the creditor had commanded the slave to commit the offence, then the creditor is held liable (IC IV 47 1–10). If the slave were the victim of an offence, the original owner of the slave and his creditor could join together to prosecute the offender and, if successful, split the damages equally (the slave, though, gets nothing); but if one of them refused to go to trial, the other could prosecute the offender unilaterally and, if successful, kept all of the damages (again, the slave got nothing) (IC IV 47 10–16). These remarkable rules provide our best evidence for debt bondage in classical Greek law,⁵⁰ but they are almost never mentioned in the most oftcited studies in Anglo-American scholarship. G. E. M. de Ste. Croix, for example, who wrote influentially on debt bondage, mentions these laws only in a footnote, and appears to have been unaware that a slave could also be a debt bondsman at the same time.⁵¹ Finley asserted that the debt bondsman in Gortyn held a ‘halfway status’ between slavery and freedom.⁵² That conclusion only follows if we define a free man as someone who does not live under the ⁴⁷ Finley (1981): 138. ⁴⁸ Gortynian slave status is treated in much greater detail in Chapter 7 below. ⁴⁹ This document was available in the same volume of inscriptions from which Finley cites the aforementioned debt bondage laws relating to free persons, Guarducci (1950). Lotze (1959: 15) knew of it, but I can find no trace of discussion by Finley. ⁵⁰ They are discussed in an excellent recent study, Kristensen (2004a). ⁵¹ De Ste. Croix (1981): 162–70. Gortyn is mentioned on p. 571 n. 58, with no discussion. ⁵² Finley (1981): 138.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

75

compulsion of another. Since the debt bondsman is under more compulsion than a citizen unencumbered by debt, yet under less compulsion than a slave, this ‘halfway’ label makes a certain kind of sense. But it is not the way the Gortynians looked at the problem: they made a juridical distinction based on whether the individual in question was or was not the property of another, because different sets of rules governed debt bondsmen depending on their juridical status. From this perspective, the bondsman was either (juridically) free or (juridically) a slave: the issue of ‘living under the compulsion of another’ was irrelevant. From this (we should note, emic) point of view, therefore, there was no such thing as a ‘halfway’ status. That said, the juridical understanding of freedom was not the only kind that the Gortynians possessed: they also used the word eleutheros as a technical term for citizens: we know that eleutheroi (citizens) were distinguished from apetairoi (a kind of non-citizen). Both were free, juridically speaking, but only the former held citizenship rights (see, e.g., IC IV 72 II 1–5). Again, the distinction is made with regard to a specific context: when Gortyn’s laws consider juridical status, they make a juridical distinction between free and slave based on the criterion of property. When the subject matter switches, different criteria are employed. We have, therefore, not one ‘Gortynian concept of freedom’, but at least two meanings, each of which is context-specific in its application. Finley’s idea of a spectrum involves two further problems that render it untenable as an approach to dependent status in ancient Greece. First, its poles are delimited by social criteria, more specifically, the idea of freedom in terms of not living under the compulsion of another person. Groups that have a more favourable ratio of rights over obligations cluster towards the ‘free’ end of the spectrum, whereas groups with few rights and many obligations cluster towards the opposite pole. Yet, along this spectrum, Finley seeks to locate entire juridical groups that are defined not on social but on formal juridical grounds.⁵³ Finley was aware of this problem: ‘It may be objected that I am confusing political and social categories with proper juristic ones.’ But he swatted aside this possible objection by writing: ‘to that I would reply that such ‘confusion’ is inherent in Greek thinking and Greek institutions. To separate them might be more elegant, more Roman, but it would no longer be Greek.’⁵⁴ Here, Finley explicitly claims that these categories were muddled together in the Greek mind; only the Romans made a clear-cut distinction between juridical and social categories. He lays claim to analysing the problem from an emic—that is, a Greek—perspective. But we saw earlier in this section, on pp. 67–70, that the Greeks used different meanings of freedom in different contexts, and that they did not muddle them together whatsoever. Gortynians made two sets ⁵³ Cf. Rotman (2004): 43: different meanings of freedom belong to different registers and cannot be mixed together as Finley does. ⁵⁴ Finley (1981): 147.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

76

Greek Slave Systems

of rules for debt bondage depending on whether the bondsman was juridically a slave or not, yet could use the term eleutheros as a technical term for ‘citizen’ without causing confusion. Athenian litigants made juridical arguments about status when it was contended in court and did not fudge these ideas together with philosophical or political meanings of slavery and freedom that belonged in other contexts. Finley’s claim that these notions were jumbled together in the Greek mind is demonstrably false; it is tantamount (if I might resurrect our earlier example) to claiming that a modern Briton cannot not tell the difference between a scientific concept of liquidity and an economic one because such a distinction does not exist in ‘the British mind’.⁵⁵ The second problem of the spectrum idea is that in a spectrum one colour shades gradually into the next.⁵⁶ But with the ragtag collection of dependent juridical groups to which Finley refers, there is no question of one status gradually shading into the next. An individual did not drift from being a helot one day to becoming an Athenian-style ‘chattel slave’ the next; at Gortyn, a man could be a slave and a debt bondsman at one and the same time (a possibility that cannot be accommodated in Finley’s spectrum); and if a slave at Athens or anywhere else in the Greek world were to be freed, that was the act of a single moment, not a gradual metamorphosis of one status into another. Finley was right to point to the complex variety of status differences in the Greek world, a theme that has recently been taken up again in a useful study by Kamen.⁵⁷ But this complexity should be considered as a miscellany, not as a spectrum; and the complexities of status more broadly defined (e.g. one’s education, social connections, charisma, and the various forms of ‘capital’ later discussed by Bourdieu) need to be explained alongside and in relation to legal status, but not compressed into a single unilateral ‘spectrum’ called ‘status’.⁵⁸

VI. A REVISED APPROACH Finley’s work on servile statuses represented a valuable step forward from earlier work insofar as he drew attention to the considerable diversity among ⁵⁵ This issue stems from a larger problem with the concept of ‘mentality’ that marks some of Finley’s most influential work, most notably his ideas of an economically unproductive mentality that was allegedly hardwired into the minds of Greek and Roman elites throughout antiquity, a view fully articulated in Finley (1999). Against the concept of ‘mentality’, see Lloyd (1990), and in terms of the argument of this chapter, see especially Lloyd’s point about multiple mentalities on p. 5. Goody (1977) provides much useful discussion on this problem from an anthropological point of view. ⁵⁶ As noted by de Ste. Croix (1981): 137. ⁵⁷ Kamen (2013). ⁵⁸ In a major contribution to the methodology of social history, P. A. Davies (2017) outlines a simple but effective method for avoiding these pitfalls. Davies’ method should serve as a baseline for subsequent studies of status in the ancient Greek world.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

77

Greek poleis in terms of local manifestations of slavery and dependent status. Comparing ‘servile statuses’ across the Greek world to the variations between poleis in terms of weights, measures, and coinage standards, Finley championed an approach that avoided reductive generalization or extrapolation of the Athenian model across the Greek world.⁵⁹ As we have seen, however, his method of articulating this genuine diversity is riddled with methodological problems. Can we do better? Five points are worth bearing in mind: (i) If we draw together the conclusions of this chapter and Chapter 1, a sensible way forward can be set out. First, the Greeks did have a working concept of ownership (and thus a legal concept of slavery), even if they did not formulate this into an abstract jurisprudential definition. Greeks bought and sold other human beings, and had a pragmatic understanding of the legal rights and responsibilities that being a slaveholder entailed. When they freed their slaves, they knew that they were giving up these rights, though they could in many cases append legal obligations to the act of manumission that resembled slavery in certain respects, yet fell short of true ownership: they indicated this change in legal status by coining specific terms for manumitted slaves (apeleutheroi, exeleutheroi), and created institutions and procedures that recognized this change in status. This emerges clearly from a study of the evidence that concentrates on practices rather than trying to reverseengineer practices from philosophical abstractions. (ii) Second, we can speak of ‘slave status’ in general terms without implying that the whole body of slave law in, let us say, Athens was replicated in carbon copy at, for example, Gortyn. So long as the basic elements of ownership are present in both places, we may speak of slavery, even if the specific manifestation of these elements varied between different regions at the level of detail. Such an approach aligns with the terminology used in the sources, where Gortynians used terms like dolos and woikeus that were cognate with doulos and oiketes in Attic Greek. The concept of ownership used here is flexible enough to admit all manner of local variations and restrictions, as we saw in our comparison of slave law in Athens and Babylonia in Chapter I. In practical terms, then, this enables us to accommodate the local differences in slave status that existed across the Greek world without having to ring-fence the term ‘slave’ as applicable only to Athens, Chios, Corinth, and so on, but to root about for different terms to apply in the case of Sparta, Thessaly, or Crete (cf. Chapters 6 and 7 below). (iii) Third, when we find the terminology of slavery and freedom in our sources, context is crucial. In a legal context, a slave is a person who legally

⁵⁹ Finley (1981): 140. Cf. Garlan (1988): 85. Finley’s comparison of regional differences between Greek slave systems to those in coinage and weight standards is an excellent simile. For coinage standards across the Greek world, see the overview provided in Psoma (2015).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

78

Greek Slave Systems

belongs to another person, and a free person is anyone who does not fall into this category; in other contexts the terminology could have different, distinct meanings. This rich variety of meanings was not fused into an amorphous ‘Greek concept of freedom’, but was rather a set of related ideas that stemmed from an asymmetrical legal relationship. In abstract terms, these multiple nuances cluster around the issue of the asymmetrical power relationship: in that sense, they share a ‘family resemblance’. In practice, though, the meaning of eleutheros or doulos and related words was calibrated through the context in which this vocabulary was employed. (iv) Scholars sometimes write that the concepts of slavery and freedom were ‘blurred’ or ‘overlapping’ in ancient Greece.⁶⁰ To some extent the latter term is justified: as we noted with the example of Athenian freedmen (who could be legally free, yet remain subject to the domination of their former owners to whom they owed formal obligations) or with Plangon in Menander’s Heros (who was not legally a slave, but shared many other attributes with slaves), the different shades of meaning that the language of slavery and freedom entailed could be mapped in multiple ways onto specific individuals. In this respect, individuals could be slaves in one sense, but free in another. However, the word ‘blurred’ is very different from the term ‘overlapping’, for it conveys a sense of conceptual opacity and an inability to pin down whether this or that person really was a slave. As we noted with the examples from the Against Neaera, the Against Pancleon, and the Against Euboulides, Athenians of the classical period were perfectly aware of the distinction between slavery as a legal status and extralegal usages of the terminology of slavery and freedom. Ultimately, we must settle on descriptive language that can capture the nuance in our sources without allowing nuance to drift into vagueness and imprecision. Status was not blurred in classical Athens, at least not in the sense that it led to the Athenians being tied up in conceptual knots; but it was complex and multilayered.⁶¹ (v) Finally, when looking at status it is important to distinguish between institutionalized aspects of status (e.g. legal statuses such as ‘slave’, ‘citizen’, ‘freedman’, with formal rights, and obligations) and non-institutionalized aspects of status (physique, charisma, education, networks of friends and associates, reputation for competence at a given task, etc.). The former aspects allow us to look at legal status groups in aggregate; but when it comes to

⁶⁰ Blurred: Zelnick-Abramovitz (2005): 295; 310; Vlassopoulos (2007b): 34–6; 46–7; 51; overlapping: Zelnick-Abramovitz (2005): 27; 179. A perfect example of this idea of ‘blurred boundaries’ that overlooks context and fails to distinguish between the technical legal condition and the various other shades of meaning slave terminology could have: Hezser (2016). ⁶¹ Once more, with the caveat that this complexity was navigable, that legal status was not conceptually fudged into other aspects of status, and that problems in identifying slaves arose from pragmatic issues of information gathering and identification.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Riddle of Freedom

79

individuals, we should consider the entirety of this constellation of variables without compressing them into a unidimensional schema. A robust social analysis can neither neglect the legal structures that define the legal status group to which one belongs nor neglect the vast array of other spectra by which social rank is negotiated.⁶²

⁶² Again, this point is discussed at greater length in Davies (2017).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

3 Status Distinctions in Greece and the Ancient Near East Having examined the Greek conception(s) of freedom and slavery in law and metaphor, we are now in a position to turn our attention to the ancient Near East, where, according to some scholars, the concept of freedom did not exist. In arguing for a fundamental qualitative difference between Greek and Near Eastern forms of slavery, M. I. Finley wrote the following: If one examines the various law codes of the ancient Near East, stretching back into the third millennium B.C., 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 required 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 has been different. Hence, for example, careful Hittologists 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 emphasise the fact that slaves were never very significant and never indispensable in the ancient Near East, unlike Greece and Rome.¹

Here and elsewhere in his work Finley constructs a black-and-white contrast: in Greece and Rome everyone could be thought of as either free or slave; in the Near East, there existed a hierarchy of statuses, a continuum in which one

¹ Finley 1964: 238 (= 1981: 120–1).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

82

Greek Slave Systems

servile status shaded into the next.² The reason why the Greeks and Romans conceived of a simple bifurcation between slaves and free men, according to Finley, lay in the growing economic significance of slavery in Greece from the sixth century onwards. There, chattel slaves came to be the dominant labour source on which the elite depended for their income: the rising economic significance of slavery thus spurred a conceptual development whereby all individuals were divided into the categories of free and slave. Hence Finley’s famous formulation: ‘one aspect of Greek history, in short, is the advance, hand in hand, of freedom and slavery’.³ There are five basic problems with these claims. First, Finley sets up the contrast between Greece and the Near East in too simplistic a fashion: Greek society can, of course, be viewed from the vantage point of property law, in which case all individuals were indeed either slave or free. But it can also be seen in terms of a much wider conglomeration of status groups with juridically prescribed rights and obligations. The two approaches are not mutually incompatible: they simply view the situation from different perspectives. Second, the idea that Near Eastern status terms were blurred—an idea ventured by Westermann and adopted by Finley—rests on their allegedly indeterminate meaning; but, as we shall see, it is rather the case that Near Eastern slave terminology is, like the Greek terminology of freedom and slavery that we explored in Chapter 2, polysemous: it can be used in more than one context, and its meaning shifts accordingly. In this respect, there is no fundamental difference between the Greek and Near Eastern worlds. Third, the ‘blurring’ of status boundaries asserted by Finley does not tally with the evidence of Near Eastern legal documents: individuals in Near Eastern societies could make razor-sharp status distinctions between different juridical groups, and there was no conceptual blurring of one group into another whatsoever. Fourth, terminology for freedom did exist in the ancient Near East, although its usages were not identical to those of its Greek equivalent. Fifth, Finley represented the development of the conceptual division of society into free men and slaves in a deterministic sense: not only is the economic importance of slave labour both a necessary and sufficient condition for this conceptual development, but there is a mechanical causal connection established between economic conditions and their alleged conceptual consequences. This deterministic link has led Raaflaub to claim that the absence of a conceptual distinction between free men and slaves must indicate the economic unimportance of slavery in a given society. To use a Marxist metaphor, the ² Finley (1981): 132 in relation to his continuum of statuses: ‘it is not a bad metaphor when applied to the ancient Near East or the earliest periods of Greek and Roman history. There one status did shade into another’ (my emphasis). Cf. Raaflaub (2004): 14–15. As we noted in the Introduction (pp. 13–14 above), Finley probably adopted this view from his teacher W. L. Westermann. ³ Finley (1981): 115.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Status Distinctions

83

presence or absence of this distinction in the conceptual ‘superstructure’ can serve as a proxy that allows us to discount its importance at the ‘base’ level of economic organization.⁴ Each of these misconceptions needs to be explored in more detail.

I. HIERARCHIES OF S TATUS We have already seen that the Greeks maintained a distinction between slaves and free men based on the criterion of property. That is the way that the Gortynians formulated their laws on debt bondage, and that is the distinction that Thucydides (8.28.3–4) makes when he describes, for example, the sack of Iasus by the Peloponnesians and the forces of Tissaphernes: the army, he writes, handed over the town and the captives (andrapoda) to Tissaphernes, both free and slave (doula kai eleuthera). This simple division views the captives from the juridical register, in terms of the statuses they held prior to capture, and divides them into property (slaves) and non-property (free persons). Was this the only type of status distinction that Greeks of the classical period could make? Such a view is evidently mistaken. We might take as illustrative column II of the Great Code of Gortyn (IC IV 72 II 2–45), which provides an example of a law code dealing with a hierarchy of statuses with differentiated penalties according to the status of the perpetrators and victims, exactly what Finley finds in the Near East, yet implies is missing from Greece. One cannot simply dismiss this point by arguing that Gortyn is an ‘early Greek society’ of the sort in which Finley claimed that a continuum of statuses, each one shading into another, existed.⁵ Two reasons rule out such a conclusion: first, as we noted in Chapter 2, the Gortynians on several occasions make the very same division of society into free and slave (using the property criterion) that Thucydides does in the above example. Second, even in classical Athens, which Finley takes as a paradigmatic case of the simple division of society into free men and slaves, we can discern a ‘hierarchy of statuses’. Deborah Kamen has illustrated this fact in her recent book on status in classical Athens, where she has shown that the traditional division of Athenian society into three orders (citizens, metics, and slaves) obscures a more complex legal reality. Manumitted slaves, for example, although in some senses subsumed into the metic population (they paid the metic tax and were required to have a citizen prostates), were formally distinct ⁴ Raaflaub (2004): 7 ‘this interactive relationship permits us to use conceptual changes as indicators of how contemporaries experienced and reacted to ongoing changes’. See more broadly Raaflaub’s discussion on pp. 1–9. ⁵ Finley (1981): 125, 128, 132.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

84

Greek Slave Systems

from the metics in several respects, since they were forced to register their former owner as their prostates (whereas normal metics had a free choice of prostates) and their property passed to their former owner after they died if they did not have children (whereas for normal metics it passed to their next of kin).⁶ In other words, the division of society into a hierarchy (or better, a miscellany) of status groups is not limited to the Near Eastern world, but can apply to the Greek world as well. One does not need to take a zero-sum approach and set the juridical division of society into free and slaves (based on the property criterion) at loggerheads with the more elaborate description ventured in Kamen’s book: both are legitimate ways of describing Greek society, but simply view the problem from different perspectives.

II. L EGAL AND E XTRALEGAL USAGES O F S LAV E TERM INO LO GY If we cast our minds back to the statement of Westermann (quoted in the Introduction, on pp. 13–14 above), we will recall that his idea that Near Eastern societies lacked a sharp legal distinction between free persons and slaves is based on an alleged terminological indeterminacy. He pointed to Egyptian terminology as paradigmatic of this trend, and then proceeded to discuss the Hebrew word for slave, ‫’( עבד‬bd, pronounced eved): ‘The Hebrew word ’ebed suffers from the same vagueness, its application ranging through “slave” or “servant of the Lord” in the phrase ’ebed Jahwe, to the titular epithet describing a high military or civil official as ’ebed el malek, “servant of the king”’.⁷ Westermann contrasted this state of affairs with the ‘clarity of differentiation’ found in Greek terminology. But if we pause for a moment and consider the problem more carefully, we can see that the contrast that Westermann asserts does not in fact exist. We noted in Chapter 2 that the Greek terminology for slavery and freedom can be employed in many extralegal circumstances, and this elasticity of language has engendered great confusion among scholars, some of whom have asserted that these Greek concepts were blurred and indeterminate. We saw too that the Greeks did in fact have a clear idea of the legal difference between free persons and slaves, notwithstanding the various other usages of slavery terminology, and that context allows one to make sense of which shade of meaning is implied. As we shall see in the coming chapters, the same problem prevails in studies of Near Eastern slavery. In Hebrew, ‫ עבד‬does have a basic legal meaning, but it ⁶ Kamen (2013); Dimopoulou-Piliouni (2008); Canevaro and Lewis (2014). ⁷ Westermann (1955): 42–3.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Status Distinctions

85

can be used in polite discourse to express deference to a social superior, or indeed man’s relation with God. Akkadian slave terminology is used in a similar manner. For instance, the Assyrian word urdu denotes someone who is a slave in the property sense; but it can be used in situations of social deference, where individuals can be described as the urdu of their superior. This state of affairs has engendered some debate among Assyriologists about the precise meaning of the word, but as Ahmad, Postgate, and Baker have shown, the simple fact is that the word is polysemous, and context generally makes clear which is implied.⁸ Whilst it is true that slave terminology in Greek and in Near Eastern languages can have a strict legal meaning as well as auxiliary meanings that are applied in extralegal contexts, it is important to point out that these auxiliary usages reveal a deep cultural rift between the Greek world and that of the Near East. In Hebrew, for instance, one can find individuals addressing their social superior as ‘my master’ or ‘my lord’ (‫אדני‬, e.g. Gen. 42:10). In the Greek world this would have been viewed as intolerable grovelling. In a study of the application of the word δεσπότης (‘master’) in classical Attic, Harris has found no instances when it is applied by a free man to another free man who is his social superior: the term is always applied by slaves to their masters or by free men to the gods.⁹ It is no wonder that easterners appeared servile and submissive to the Greeks, a point stressed in a careful comparison of Greek and Persian terminology by Anna Missiou.¹⁰ (These distinct traditions of slave metaphor were later fused in Hellenistic and Roman-era Jewish discourse, in which we find an admixture of traditional Hebrew usages with distinctively Greek usages, for example the extension of slave terminology to emotions and states of mind.¹¹) We may admit provisionally, then, that it is possible that the alleged ‘blurriness’ of Eastern slave terminology may simply reflect polysemy and a plurality of usages, just as slave terminology in Greek has a plurality of usages (though applied in a rather different way), and that context will allow us to pinpoint the implied meaning. Whether this alleged conceptual blurriness really did exist in the Near East can be put to the test if we analyse Babylonian legal documents deriving from courtroom contexts, a task we shall turn to very soon. First, though, is a concept of freedom a necessary prerequisite for a sharp legal concept of slavery, as Finley argued? Jack Goody has recently criticized Finley’s claims on pragmatic grounds: Since an institution approximating to slavery existed in the other societies he mentions [i.e. the ancient Near East], whether or not it can be considered ‘basic’

⁸ Ahmad and Postgate (2007): ix; Baker (2016): 24–5. Baker (2016): 20 shows that slave imagery and terminology can also be used in a negative sense in Assyrian texts. ⁹ Harris (2006): 278–9. ¹⁰ Missiou (1993). ¹¹ Hezser (2005): 333–6.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

86

Greek Slave Systems

or ‘dominant’, it seems inconceivable that there was no recognition of the difference between slavery and its absence, even if there was not a single noun to designate it. While slavery has been present among the groups with whom I worked in northern Ghana, there was no specific word used to describe being free; nevertheless people had no difficulty whatsoever in making the distinction between a ‘slave’ (or ‘pawn’) and other people.¹²

Likewise, Vlassopoulos has recently emphasized that the opposite of slavery in most cultures is not freedom, but mastery.¹³ We may provisionally conclude, then, that at least on a theoretical level, individuals do not require a special term for freedom to possess a clear concept of slavery. Rather, all that is needed is a clear idea of the slave qua property: everything outside this category can thus be distinguished as something else. The idea that one needs to have a clear antithesis to slavery—freedom—in order to have a clear concept of slavery itself is based on a flawed view of human cognition: concepts may or may not have an antonym, but existence of the latter is not a precondition for the clarity of the former.

I II . S T A T U S D I STI N C T I O N S I N A C T I O N : TWO BABYLONIAN COURT CASES Records of court cases from the Neo-Babylonian period provide empirical confirmation that legal status distinctions could be made in a sophisticated and pragmatic manner.¹⁴ Our first example, related in the document Nbn. 1113, concerns a case dating to 548 BC.¹⁵ The dispute centres around a slave named Bariki-ili, who may have been a Judahite captured during the Babylonian conquest of Judah several decades earlier. Bariki-ili, it appears, made the claim that he was a free man: A court examination was conducted before the judicial officer, important men, and the judges of Nabonidus, king of Babylon. They presented their case, and (the judges) heard their testimony. (The judges) had the document read aloud (that demonstrated) the slave status of Bariki-ili, who had been purchased, pawned, (and) given as part of a dowry to Nupta daughter of Gaga, between the 35th year of Nebuchadnezzar, king of Babylon, and the 7th year of Nabonidus, king of Babylon, (and which demonstrated that) Nupta had later legally transferred him together with a share of the house and slaves to her son Zababa-iddin and to her ¹² Goody 2006: 58. On pawns and status distinctions between pawns and slaves in Western Africa, see Lovejoy (2014). ¹³ Vlassopoulos (2016b): 10. ¹⁴ Cf. Vlassopoulos (2007a): 105. For Neo-Babylonian court procedure see Holz (2009). ¹⁵ See Dandamaev (1984): 221–2; 440–3.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Status Distinctions

87

husband Iddina. Whereupon they said the following to Bariki-ili: ‘you have instituted a complaint, maintaining, “I am a free man.” Show us (the document relating to) your free status.’¹⁶

Faced with the impressive tablet trail relating to the conveyances and pledge of his person, and lacking a document of manumission, Bariki-ili finally gave up the pretence: Bariki-ili answered the following to this: ‘I have succeeded in running away from the house of my master two times and was not discovered for many days. I was afraid and I said: “I am a free man.” I have no free status. I am a slave who was redeemed for silver belonging to Gaga. She gave me (to) her daughter Nupta. Nupta legally transferred me to her son Zababa-iddin and her husband Iddina. After the deaths of Gaga and Nupta, I was sold to Itti-Marduk- balat ụ son of Nabuaḫḫe-iddin descendant of Egibi. I am a slave.’ . . . The Judicial officer, important men, and judges heard his testimony and returned Bariki-ili to slavery.¹⁷

Contrary to the notion popular among some classical historians that ‘servile status’ was a blurred affair among eastern peoples, this document attests to an eminently practical and workmanlike ability to make clear legal distinctions between slaves and free persons.¹⁸ The second document we shall consider relates to a case heard in 558 BC.¹⁹ Finley was in fact aware of the existence and contents of this document.²⁰ It reveals the mathematical sophistication of Babylonian courts in determining whether or not a temple oblate (širku)²¹ named Ina-ṣilli-abulli, serving in debt bondage, had worked off his obligation and could be considered free of debt. His father Aḫušunu contracted a loan for ⅔ mina and 2 shekels from a woman senior in the priestly hierarchy named Aḫata, pledging him as a pawn to work off the debt. Ten years later, Ina-ṣilli-abulli’s father was dead, as was Aḫata; the role of creditor had passed to a woman named Banat-ina-Esagila, and Ina-ṣilliabulli considered the debt to have been finally paid off. He approached a court to determine whether or not this was so: The judges reckoned up the money belonging to Aḫata and the interest on it and increased her silver of two-thirds mina two shekels (i.e. calculated the interest) and established (that the interest had risen to the sum of) one and a third minas four shekels of silver. They also calculated the quitrent of the baker Ina-ṣilli-abulli for six years, in accordance with the admission of Banat-ina-Esagila, and (established that)

¹⁶ Nbn. 1113, tr. Dandamaev 1984: 441. ¹⁷ Nbn. 1113, tr. Dandamaev 1984: 442. ¹⁸ Cf. Van der Spek (1990): 249, writing that Finley’s idea ‘that the one status did shade into another is untenable’; Head (2010): 59: ‘Finley erred, believing that there were “shades of freedom” in Babylon.’ For other cases with similar features, see Jursa, Paszkowiak, and Waerzeggers (2003–4); Baker (2004) no. 183; Magdalene, Wells, and Wunsch (2008); Wunsch and Magdalene (2014): 341–2. ¹⁹ Revue d’assyriologie 12 (1915): 5–10. ²⁰ Finley (1981): 120. ²¹ For širku status, see Chapter 11, p. 237 n. 11.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

88

Greek Slave Systems

Banat-ina-Esagila had received seventy-two kur of barley from him, not counting twenty kur of barley bread. They calculated that (a total of) ninety-two kur of barley (was paid) to Banat-ina-Esagila. Further, they calculated [a total of . . . ] for the four years during which he had served Aḫata, the former sāgittu, that two minas six shekels of silver (had been paid) to Ba[nat-ina-Esagila], the sāgittu, and that he had (also) paid the debt to Aḫata, the former sāgittu. And they took Ina-ṣilli-abulli, the širku of (the goddess) Ištar of Uruk, away from Banat-ina-Esagila and gave him to Ištar of Uruk. Their case has been decided, their case is closed. In order that (this decision) not be altered, the judges have drawn up the document, witnessed (it) with their seals, and given (it) to Ina-ṣilli-abulli.²²

The facts of the case were clear and uncontested; the court’s role consisted here of calculating the amount of the loan plus interest for ten years at 20 per cent, calculating the amount paid back by the debtor, and confirming that the debt had been fully repaid. It was indeed repaid, and the širku was taken from his temporary mistress and returned to temple service. Finley drew quite the wrong conclusion from this document. He asked ‘during the ten years of service was the bondsman . . . a free man or a slave?’ By applying his vague Aristotelian approach wherein freedom is viewed in terms of ‘absence of constraint’, Finley muddled a rather clear-cut case. From the emic Babylonian point of view, the man was a širku, a temple oblate, and therefore juridically free. The Babylonian terminology of freedom bears some resemblance to its Greek counterpart; and the language of free (mār banûti) status can be used in both a juridical and a social sense. In a juridical sense it denotes those who are not the property of another person or institution: as Wunsch and Magdalene have shown in an important recent study, a slave might be manumitted and given a tablet of mār banûti status (t ụ ppi mār banûti), yet become a širku—that is, be legally free in the sense of not being the alienable property of anyone, yet be socially and legally a temple dependant.²³ On the other hand, the free urban citizenry are often referred to as mār banî, a narrower social usage.²⁴ This is not unlike the situation in Gortyn’s laws, where the term eleutheros can be used in a juridical, property sense in contradistinction to doloi, slaves (IC IV 72 I 2, 4–5, 9, 15–16); but also as a technical term for full citizens in contradistinction to free non-citizens such as apetairoi (IC IV 72 II 1–5). In both cases there is a technical legal register for the terminology, and a looser social register; context allows a distinction to be made between the two. Recent studies are beginning to show that emic,

²² Revue d’assyriologie 12 (1915): 5–10, tr. Dandamaev 1984: 542. ²³ Wunsch and Magdalene (2014): 342: ‘širkus are some type of mār banî, that is, persons of free status’. See also van der Spek (1990): 249. ²⁴ Wunsch and Magdalene (2014): 342. My thanks to Cornelia Wunsch (personal communication) for discussing this term with me.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Status Distinctions

89

Near Eastern concepts of freedom did indeed exist; and whilst their cultural significance and articulation differed from Greek concepts of freedom, classical historians have greatly overplayed that difference.²⁵ To sum up so far, then: methodologically speaking, the concept of freedom is not a prerequisite for a clear juridical concept of slavery. Yet not only could the Babylonians make razor-sharp legal distinctions between slaves and nonslaves; they also possessed a conceptual vocabulary that included terminology for freedom. One task remains: an examination of Finley’s thoughts on the connection between the conceptual distinction between free persons and slaves and the economic role of slavery will provide a key to understanding why he paid little attention to slavery as an economic institution outside the Graeco-Roman world.

I V . ‘ THE ADVANCE, HAND IN H AND, OF FREEDOM AND S LAVERY’ ‘It is a fact, I believe, that social and political progress in the Greek poleis was accompanied by the triumph of chattel slavery over other statuses of dependent labour.’²⁶ ‘The cities in which individual freedom reached its highest expression—most obviously Athens—were cities in which chattel slavery flourished.’²⁷ Finley’s model of the growth of slavery in ancient Greece tethered together the economic role of slave labour and the conceptual distinction between slaves and free persons. In the old archaic world, elites relied on ‘other statuses of dependent labour’. As Finley himself wrote: In early Greece and Rome, how did the rich and well-born, the holders of the large estates, obtain and increase their labour force? Both hired labour and chattel slaves are known from our earliest sources, the Homeric poems and the Twelve Tables, but it is clear that they were not the answer. Labour was essentially dependent labour—clients, helots, pelatai or whatever else they were called— and debt bondsmen.²⁸

In this situation, one status shaded into the next, much like the Near East. But a major transformation in labour relations in the sixth century BC ushered in a conceptual revolution as well. Solon’s ban on debt bondage forced the Athenian elite to turn to imports of barbarian slaves as their new source of extrafamilial labour. The muddy waters of archaic dependence then clarified into a new conceptual reality: the ability to make a sharp distinction

²⁵ Snell (2001); von Dassow (2011); Kuhrt (2014); cf. Vlassopoulos (2007a): 101–22. ²⁶ Finley (1981): 149. ²⁷ Finley (1981): 114. ²⁸ Finley (1981): 155.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

90

Greek Slave Systems

between free men and slaves.²⁹ ‘Only when slaves became the main dependent labour force was the concept of personal freedom first articulated.’³⁰ As we shall see in greater detail in Chapter 5, this model of the rise of slavery in ancient Greece has proven extremely influential. Finley’s claim of ‘the advance, hand in hand, of freedom and slavery’ is one of his most quoted sayings.³¹ Raaflaub has taken this evolutionary model to its logical conclusion: in societies where there was no explicit slave–free dichotomy of the Greek sort, slavery cannot have been of any real economic importance. As we noted earlier in this chapter, on pp. 82–3, to use a Marxist metaphor, Raaflaub’s proposal amounts to reverse-engineering information about the economic ‘base’ through the presence or absence of this distinction between free and slave in the conceptual ‘superstructure’.³² Despite its widespread currency, this particular evolutionary model is untenable, for a variety of reasons. First, as we noted in Chapter 2, a pragmatic, clear-cut distinction between free persons and slaves can already be discerned in Homer and is not a sixth-century development. Second, as we shall see in Chapter 5, there was no ‘rise’ of chattel slavery in the sixth century. The basic labour relations wherein Greek elites relied on slave labour for their wealth were already in place at least as early as the seventh century; the strategies of supply were certainly reconfigured during the archaic period, with an increasing reliance on non-Greek imports; and to be sure the slave populations of Greek communities grew in size (as did their citizen populations too). But there was no shift from a reliance on helots, pelatai, and debt bondsmen to a reliance on slaves. Indeed, despite Finley’s certainty, there is not a single mention of helots, debt bondsmen, or any of these other kinds of dependent labour in the works of Homer and Hesiod, a glaring omission that has rarely ²⁹ I cannot improve on Lenski’s pithy description (Lenski 2018b: 115): ‘Finley’s pure ancient slavery was thought to have been refined into its essence in the political centrifuge of democracy, which meant that it was bound to dissipate once that centrifuge had been turned off.’ ³⁰ Finley (1968): 308. On the same page he also wrote that ‘slavery attained its greatest functional significance, and usually its greatest numerical strength, in societies in which other, less total varieties of bondage had either disappeared or had never existed’. Cf. Wiedemann (1987): 13 for a similar view. This model inspires the evolutionary view of Ismard (2015): 58–60. According to Ismard, the rise of democracy and the (allegedly sixth-century) rise of chattel slavery worked hand in hand: ‘Les deux phénomènes engagèrent conjointement une vaste redéfinition des statuts personnels au sein des sociétés civiques du monde grec.’ Hall (2007): 242 holds a similar view: ‘The concept of a slave economy depends on a sharp definition of chattel slavery that seems not to have existed for much of the archaic period.’ On Finley’s model, there is much useful analysis in Lenski (2018b). ³¹ Finley (1981): 115. For citations of the phrase, see, e.g., Willetts (1977): 183; Vidal-Naquet (1986): 164; Morris: (1987): 177; Garlan (1988): 39; Cartledge (1993b): 176; Osborne (1995): 38; Rihll (1996): 90; Joshel and Murnaghan (1998): 16; Descat (2006): 21; Vlassopoulos (2009): 347; Tamiolaki (2010): 19; Wallace (2010): 139; Tordoff (2013): 16; Zurbach (2014): 273; Ismard (2015): 58. ³² See n. 4 above. Against this rigid approach, see the cogent criticisms of Vlassopoulos (2016b): 8–9.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Status Distinctions

91

been confronted in scholarship on archaic Greek slavery. Nor was helotage a fossilized ‘archaic’ form of dependent labour representing an earlier stage in the development of Greek slavery: it was a form of slavery specific to Sparta alone, its features tailored to the institutional structure of Spartan society (see Chapter 6 below). Finley’s model of the rise of slavery in archaic Greece is so riddled with impossibilities that it must ultimately be abandoned.³³ This has implications for his thoughts on the conceptual development (and significance) of the free– slave dichotomy. Above all, it is necessary to uncouple the deterministic link that he proposed between the rising economic importance of slave labour and a putative conceptual development wherein the (allegedly) indistinct morass of archaic dependency resolved into two clear categories, free and slave.³⁴ Finley and Raaflaub posed the following question: why did the same free– slave dichotomy not develop in the ancient Near East? We have seen that this question is based on a false premise, and that the Babylonians did possess terms for freedom and free status. But perhaps it is useful to see the problem the other way round: why did the Greeks feel the need to coin a word for absolutely everyone who was not a slave? We have already noted that in pragmatic terms one does not need a special word to distinguish slaves from everyone else. The answer perhaps lies in a well-known aspect of Greek culture, that is, its strong tendency to pair a concept with a binary opposite. For a society that organized its thought in binary terms to an unusual degree, it should come as no real surprise that they sought out a comprehensive term to serve as the polar opposite for ‘slave’.³⁵ It is far from my intention to deny cultural or ideological differences between Greece and the Near East in the usage of the terminology of slavery and freedom. There were clearly many differences between the two regions, a subject that requires a detailed new comparative study. My concern here, however, is with pragmatic legal status distinctions and with economic realia.

³³ See Chapter 5 for more detailed arguments against Finley’s view of archaic Greek slavery and the evolution of slavery in the sixth century. ³⁴ Kostas Vlassopoulos (2016b: 10) has arrived at the same conclusion: ‘Finley was right that the concept of freedom is largely absent outside the West, but his assumption that the emergence of slave societies is a sufficient cause for the creation of the conceptual language of freedom is clearly wrong.’ ³⁵ I cannot claim credit for this observation, which was suggested to me by Alexandre Johnston. On binary concepts in ancient Greece, see Lloyd (1966). I have since found that Cartledge (1993a): 169 makes a similar observation: From men and de to Men and Women, as has been neatly said, the Greeks resolved basic cultural problems of ethnicity and gender, of political and social identity, into mutually exclusive and jointly comprehensive binary oppositions—Greeks and Barbarians, Men and Women, Citizens and Aliens, Free and Slave. Such a habit of thought was attractive to relatively unsophisticated Classical Greeks because of its simplicity; as we ourselves might say, everything could be seen by them in black or white terms.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

92

Greek Slave Systems

In substantive terms, that is, the everyday realities of law and economic life, the cultural differences relating to ‘freedom’ between Greece and the Near East amounted to naught. * * * Having set out a robust methodology for conceptualizing slavery in law and metaphor in the Greek and Near Eastern worlds, and having documented where similarities and differences lie, it is now time to turn to the economic role of slavery in comparative perspective. Chapter 4 unpacks the various methods for capturing differentials in the economic development of slavery in various regions, highlighting their strengths and weaknesses. The chapters that follow examine slavery in the Eastern Mediterranean world in two phases. Chapters 5–8 analyse slavery in the Greek world, underscoring regional diversity and the need to go beyond viewing Athenian slavery as the sole paradigm on which understanding of Greek slavery is based (it is no accident that Attica is treated at the end of this section). These chapters also contain arguments designed to revise ideas of how slavery developed in the longue durée c.800–400 BC. Chapters 9–13 step back from the Greek world proper and examine a diverse range of regions in the Eastern Mediterranean world (broadly defined to include adjoining regions in the Near East, and Carthage in the West). These investigations of slavery among the Greeks’ Eastern neighbours are set out in dialogue with the Greek case studies, in order for similarities and differences to emerge fully. Chapter 14 progresses beyond the mere observation of similarities and differences, and attempts to explain regional variation through setting out a matrix of variables that shaped the trajectories along which slavery developed in this or that locale.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

4 Slave Societies, Societies with Slaves Capturing the Relative Importance of Slavery to Ancient Economies

One of the chief heuristic tools of twentieth-century scholarship on ancient slavery—especially, but not exclusively, in Britain and America—has been the concept of ‘slave society’ as opposed to ‘slaveholding society’ or ‘society with slaves’.¹ The first term refers to societies in which slavery played a role of great importance (the precise meaning of which we shall turn to shortly). The last two terms refer to those societies that merely permitted the existence of slavery, but in which it was less important, especially in the economy; societies such as these can be seen as something more commonplace, since slavery was legally permissible in a wide range of cultures. The process of distinguishing between these two concepts is a common preliminary step in studies of slavery.² Perhaps more surprising is the longevity of the view put forward by Finley that only five ‘genuine slave societies’ existed in world history: classical Greece (by which Athens is normally meant), Rome, the US South, the Caribbean, and Brazil. Scholars working on ancient slavery have often repeated this mantra.³ There are three basic problems with this approach. First, how do we define ‘slave society’, and what job do we wish this heuristic tool to perform? Scholars have proposed different definitions of slave society: Finley emphasized a qualitative approach based on the location of slavery and contribution of

¹ A recent volume, Lenski and Cameron (2018), provides a great deal of fruitful debate and reflection on the issue discussed in this chapter, but appeared too late to be taken into account fully here; our approaches are, however, complementary. ² e.g. Turley (2000): 4–5; Stewart (2012): 4; Lovejoy (2011): 9. For Finley’s approach to ‘slave society’, see Lloyd-Jones (1991): 98–100; for the invention of the category, see Higman (2001). ³ Finley (1968): 308 (omitting Brazil); (1980): 9, where all five appear. Cf. Hopkins (1978): 99–100; Cartledge (1985): 22; Fisher (1993): 4; Bradley (1994): 12; Garnsey (1996): 2; Andreau and Descat (2006): 18; Joshel (2010): 7–8; Bradley and Cartledge (2011): ix; Forsdyke (2012): 19–20; Cartledge in OCD³ s.v. ‘slavery, Greek’.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

94

Greek Slave Systems

slave labour to elite income; other scholars such as Hopkins have argued for defining slave society on proportional, i.e. quantitative grounds. The second problem lies in the mantra of ‘five genuine slave societies’: even if we use Finley’s definition of ‘slave society’ (which, as we shall see, outperforms its competitors), this brief list seems excessively short, and a strong case can be made for extending it to a larger number of historical societies. Third, the idea of ‘slave society’ is only a conceptual tool, not a primordial fact; to put it in Greek terms, nomos, not physis.⁴ As we shall see, mechanical application of this principle, involving the sorting of ancient societies into the boxes of ‘slave society’ and ‘slaveholding society’, can produce a very distorted picture of reality. This chapter shall not jettison the category of ‘slave society’ altogether, because it still has a (limited) role to play; but it provides a kind of health warning regarding its (over)use.⁵

I. APPROACHES TO DEFINING ‘ SLAVE S OCIETY’ The concept of ‘slave society’ is a comparative yardstick: it aims to make a distinction between those societies in which slavery was economically important and those in which it was not. It enables us to answer questions such as ‘Did slavery play a role in Greek societies comparable at least in terms of economic function and relative scale to slavery in the US South?’ But on what basis should the comparison be made? Three definitions of ‘slave society’ vie for special attention. Let us first consider the statistical definition adopted by Keith Hopkins. It should be stated at the outset that this is probably the least useful method of classifying the importance of slavery to a given society. The essence of this approach lies in setting a proportional figure on a society’s population that must be composed of slaves to qualify it for the rank of ‘slave society’. Hopkins set a figure of 20 per cent.⁶ The ⁴ It is also a comparatively recent one: whereas the legal concepts of ‘slave’ and ‘slavery’ are extremely ancient, ‘slave society’ is a rather more recent concept; we might trace its roots back to Marx’s idea of a ‘slave mode of production’, although the modern category of ‘slave society’ means something quite different from what Marx had in mind. At any rate, if we cast our minds back to the distinction between emic and etic categories outlined in Chapter 2, we may note that ‘slave society’ is purely an etic category designed for modern researchers, and thus (unlike slavery) we have more leeway in deciding how it might be defined. ⁵ Vlassopoulos (2016b): 8–11 argues that we are better off without the category, though at p. 11 writes ‘[t]he distinction between slave societies and societies with slaves might be a useful shorthand for certain tasks’. I agree with his position; all I would add is that one such task is the work of the economic historian seeking to account for differentials in the economic role of slavery between societies, an issue we will turn to in Chapter 14 of this book. For that task, the distinction retains a certain value as a crude rule of thumb; for most other tasks, though, it is relatively devoid of merit. Cf. the comments of Harper (2011a): 37. ⁶ See Hopkins (1978): 99–102.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Slave Societies, Societies with Slaves

95

limitations of this approach are obvious. First of all, it sets an arbitrary limit. Why choose 20 per cent? Why not 25 per cent or 30 per cent? What does this figure in particular tell us about the nature of slavery in a particular society? Second, the proportional definition lacks the requisite flexibility to make the term ‘slave society’ a meaningful category. For example, if the limit were set at 20 per cent, a society of which 19 per cent of its members were slaves would be labelled a ‘slaveholding society’ whereas one in which 20 per cent of the population was composed of slaves would be labelled a ‘slave society’. Prima facie, the formal designations of these two societies could imply vast structural differences, but were one to glance at the figures, they would suggest little real substantive difference. The rigid nature of the proportional model severely limits its use as a heuristic tool. Above all, we generally lack accurate demographic information for the ancient world, so that only a society that significantly surpasses the qualifying figure (to the extent that this is obvious despite the statistical margins of error) can be confidently assigned the title ‘slave society’. One need only look at Hopkins’ demographic ‘evidence’ to grasp the limitations of this approach. In a table comparing slave societies in Athens, Rome, and the Americas, Hopkins cites R. Sargent’s figures relating to slaves and the Athenian population in the fourth century (200,000, of which 60,000 were slaves). But modern estimates (and one must always emphasize that these are merely educated guesses) range from Hans van Wees’ high estimate of 323,000 slaves in 317 BC to Sargent and Finley’s more modest estimates, with various possibilities running the gamut in between.⁷ In invoking this wide range of guesses I do not wish necessarily to endorse the outliers, or claim that nothing can be known about slave numbers at Athens; nevertheless, we simply lack the requisite data to produce an ultimately compelling figure, and with such divergent opinions the problems of the proportional approach are further compounded. In using this approach, Athens might fall within or outside this category depending on which historian’s statistical calculations we rely upon. A much better method of identifying a ‘slave society’ is, as Finley pointed out, to avoid the ‘numbers game’ and to look at the location of slavery within a given social structure. This approach focuses upon two areas: (i) the distribution of slaves: were slaves owned in similar proportions by all classes of society, or were they concentrated amongst the wealthier classes? (ii) what was the economic role of slavery, especially among the elite? Finley’s approach to defining a ‘slave society’ is that any society can be so labelled if its elite derived a significant proportion of their wealth from slave labour. This formulation has, for sure, certain vague elements (How do we characterize an ‘elite’? What constitutes a ‘significant proportion’ of its wealth?), but is ⁷ Van Wees (2011); Isager and Hansen (1975): 17; Sargent (1924); Finley (1981): 102. On van Wees’ figure, see the Appendix to this volume.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

96

Greek Slave Systems

probably more useful for this very reason, since this admits a degree of flexibility; we have seen how a rigid definition such as the proportional version displays limited utility. This approach really does tell us something about the nature of a given society: how do its wealthiest members, among whom the majority of the society’s power is concentrated, create and maintain their position of dominance? Most significantly, Finley’s approach can be utilized even if we lack reliable demographic data. Methodologically speaking, it requires identifying the standard range of slave ownership among the elite, the elite’s sources of wealth, the uses they put that wealth to in maintaining their dominance, and gauging (roughly) whether or not slave labour was a key ingredient in producing this wealth. One final approach is the broader concept of slave society developed by Orlando Patterson.⁸ This essentially accepts Finley’s idea of ‘slave society’ at its core, but opts to extend the category to societies in which slaves might not necessarily form a key element of the elite’s wealth, but nevertheless have an important and conspicuous role in social life. The problem with this conception of slave society, or ‘large-scale slave system’, as Patterson styles it, is that it is too vague, admitting too many variables, and as a result limiting its use as a category of comparison (and, after all, this is the whole point of the category ‘slave society’), since it cannot focus on one shared attribute in the manner of Finley’s concept of slave society. It is Finley’s approach to slave society that is the most analytically useful for the ancient world.

I I . T H E MY T H O F FI V E S L A V E S O C I E T IE S The notion that only five genuine slave societies have existed in world history (Greece, Rome, the Caribbean, the US South, Brazil) has been surprisingly tenacious⁹, but such a claim can no longer be accepted. Even if we focus on the list of five societies itself, problems are immediately apparent: for instance, can we really call the Caribbean a society, given that the region comprised a number of islands with differing local conditions, subject to the rule of several colonial powers? Can Athens stand as representative of the Greek world, given that the latter comprised over 1,000 poleis? What exactly is meant by ‘Rome’? Perhaps the greatest contribution of Orlando Patterson’s Slavery and Social Death has been his insistence that this brief list of five slave societies is inadequate. Although (as we have noted) he does not adopt Finley’s definition of ‘slave society’ and prefers the vaguer term ‘large-scale slave system’, it is ⁸ Patterson (1982): 353–64.

⁹ See n. 3 above.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Slave Societies, Societies with Slaves

97

obvious from his wide-ranging study that earlier scholars had not considered a broad array of societies that would comfortably fit into a strict, Finleyan conception of slave society, or even Hopkins’ quantitative definition. A few examples will help us to realize how slave societies have been far less rare in world history than the traditional view admits.¹⁰ We might begin with Africa and its surrounding islands. As Patterson has noted, ‘Keith Hopkins’ dogmatic assertion that there were only five large-scale slave societies in world history is too absurd to be taken seriously . . . in precolonial and nineteenth-century West Africa and the Sudan alone there were more than a dozen large-scale slave societies.’¹¹ Of special note is Paul Lovejoy’s work on the Sokoto Caliphate in nineteenth-century North Africa, where a large and complex system of plantations was fed by a slave supply with tendrils reaching farther into the African interior.¹² Slave societies in Africa were not just a native affair. During the sixteenth century the Portuguese island colonies of Santiago and Fogo in the Cape Verde islands would qualify as slave societies on any of the current definitions. Following the beginning of Portuguese settlement of the islands in 1462, Europeans began to import and exploit a much larger population of enslaved Africans, who grew cotton and foodstuffs. By 1582 Santiago had a population of 1,100 Europeans and mulattos, 400 free Black Africans, and 10,700 slaves (the number of slave children remains unknown). Slaves, then, accounted for at least 87 per cent of the island’s population. Fogo had a similar slave-to-free ratio, with a population of 2,000 slaves and 300 free persons.¹³ We might also turn to the slave societies of the Indian Ocean. Burton Benedict has highlighted the key importance of slavery in the Seychelles and Mauritius: slaves made up some 77 per cent of the population of the islands by 1735, a level that was maintained and even increased during the next century.¹⁴ Another notable early modern slave society was the Dutch colonial settlement in the Banda Islands, in what is modern Indonesia. These islands were the only source of nutmeg and mace, highly sought-after commodities in world markets. By 1621 the Dutch East India Company (VOC) had undisputed control of the islands, and, having decimated much of its native population, imported slave labour from diverse sources across the Indian Ocean to work its nutmeg plantations. Slaves dominated the populations of these islands: as Winn notes, ‘In 1638, a total of 2,199 slaves constituted some two-thirds of the islands’ total population of 3,843 persons. By 1794, slave ¹⁰ Cf. Lenski (2018a), criticizing the limited horizons of Finley’s list and similarly drawing attention to other slave societies in world history. ¹¹ Patterson (2008): 33 n. 5. It is interesting that Patterson directs this discourteous dismissal at Hopkins, rather than against Finley, whose idea it was. Patterson has long sought to associate himself closely with Finley and downplay the incompatibility of their approaches: see Lewis (2016): 50 n. 11, regarding Patterson (1982): 369 n. 21. ¹² See Lovejoy (1981). ¹³ K. D. Patterson (1988): 293. ¹⁴ Benedict (1980): 137.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

98

Greek Slave Systems

numbers had increased to 4,122 or around three-quarters of the islands’ population at this time.’¹⁵ Slaves thus not only represented a preponderance of the overall population, but also were the key labour sector in producing the island’s much sought-after spices. Slavery played a key role in the elite economies of other corners of the Indian Ocean. Recently, Matthew Hopper has underscored the key role of slave labour in the economy of nineteenthcentury Oman. African slaves were instrumental in two major labour sectors of the Omani economy: pearl diving and date cultivation. As with Bandanese nutmeg and mace, pearls and dates were highly sought-after commodities in world markets, and the Omani slave society survived well into the twentieth century.¹⁶ These few examples are far from an exhaustive list of possible candidates for ‘slave society’ status,¹⁷ but they are sufficient to indicate the problems with retaining the old list of five genuine slave societies. Whichever definition of ‘slave society’ one prefers, it is clear that there were very many more such societies than was once thought. The same principle must be applied to antiquity, where the idea of Athens and Rome as its sole ‘slave societies’ must be completely rethought. Recently, Kyle Harper has convincingly shown the longevity of Rome’s slave society, which persisted well into Late Antiquity.¹⁸ Noel Lenski has demonstrated the key role of slave labour among the Saracens in Late Antiquity;¹⁹ and Egon Flaig (among others) has argued for a better appreciation of the large scale and key importance of slave labour in Carthage’s economy during the Hellenistic period,²⁰ a subject that will be more fully treated in Chapter 13 of this book. If we turn to the Greek world, we can see how inadequate the mantra of ‘five slave societies’ really is. For the archaic period, Hans van Wees and Edward Harris have argued that slave labour played a key role in elite economies from as early as the seventh century, a subject that will be treated in detail in the next chapter.²¹ The Greek world of the classical period comprised around 1,000 poleis with their own individual laws, economies, and constitutions. Athens clearly cannot be treated as representative of them all; but by the same token, it strains credulity to believe that slavery attained a vastly greater degree of economic importance in Athens than anywhere else. What about the ¹⁵ Winn (2010): 371. For a useful sketch of the social structure of Banda under the VOC, see Loth (1995). ¹⁶ Hopper (2013). T. E. Lawrence provides an interesting account of date plantations manned by African slaves very much of the sort that Hopper discusses, but in the district of Wadi Safra near the Red Sea coast at the time of the Arab Revolt during the First World War. See Lawrence (1935): 89–91. ¹⁷ For one thing, I have left out the conspicuous example of Korea in the Koryo period, for which, see Palais (1984). See Lenski and Cameron (2018) for further examples. ¹⁸ Harper (2011a). ¹⁹ Lenski (2011); the same was probably true of the Huns: see Lenski (2014). ²⁰ Flaig (2009): 55–6. ²¹ Van Wees (1992): 49–53; Harris (2012).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Slave Societies, Societies with Slaves

99

island of Chios, a major producer and exporter of wine, which Thucydides (8.40.2) claimed had the largest slave population in Greece after Sparta? (Sparta herself, perhaps the most extreme form of slave society in the ancient world, is treated in detail in Chapter 6 of this book.) And what of poleis such as Corinth and Aegina, which were renowned for their large slave populations (though the precise figures preserved for them are not credible)? When Herodotus (5.31) described Naxos in the early fifth century as having a great many slaves, is it not credible—even likely—that he was describing a slave society? Even this cursory overview demonstrates that continued adherence to the old mantra of five genuine slave societies fits the historian with a very narrow set of blinkers, preventing a better appreciation of the much wider array of societies and contexts in which slavery has constituted a key source of labour for elites. The old list is now no longer tenable, and it is clear that progress on this issue will stall unless we are open to the possibility that slavery played a significant role in many more historical societies than was once imagined.²²

III. S LAVE SOCIETY AND HISTORICAL MYOPIA Even if we accept one or other definition of slave society as a useful heuristic tool (as has been argued in section I, on pp. 94–6 above) and open our minds to the possibility that there were far more than five historical societies that can be categorized under such a definition, it is still vital to highlight the drawbacks of using the category of ‘slave society’ at all. We would be wise to heed Einstein’s observation that ‘concepts that have proven useful in ordering things easily achieve such authority over us that we forget their earthly origins and accept them as unalterable givens’.²³ We may, of course, use the concept of ‘slave society’ to order our thoughts; but it is worthwhile stepping outside this paradigm once in a while to see if a different view of the issue is possible, or if the paradigm is itself restrictive. The basic problem boils down to this: let us take, for example, two of the definitions of ‘slave society’ outlined above, that of Finley, and that of Hopkins. For a society to be classed as a ‘slave society’ under Finley’s definition, we need to be able (i) to identify its elite; (ii) to identify the full range of that elite’s income-generating activities; and (iii) to determine with some degree of certainty the contribution of slave labour to those activities. For a society to ²² Several ancient historians have challenged the traditional consensus and rightly emphasized the inadequacy of the traditional list of five genuine slave societies: see Scheidel (2008): 105 n. 1; Ismard (2015): 21; Vlassopoulos (2016b): 10; and especially Lenski (2018a). ²³ Einstein (1916): 102.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

100

Greek Slave Systems

be classed as a ‘slave society’ on Hopkins’ criteria, one requires accurate statistics. On either definition, only a society for which a large amount of evidence is preserved can be subjected to these tests. We have already noted the impracticality of Hopkins’ definition for the analysis of ancient Mediterranean societies in general, given the large margins of error inherent in computations of population sizes. But even on the more practical, qualitative approach of Finley, we still cannot even begin to analyse most ancient societies. For how many ancient societies have we a full picture of the various sources of elite income and enough evidence to gauge the contribution of slave labour to that income? Very few, and Athens and Rome stand out as the major examples from antiquity (if we leave Babylonia aside for now) of societies from which large enough bodies of evidence survive to permit such analysis. It is no wonder that they are the only ancient societies to make it onto the traditional list. One must be aware that numerous ancient societies might have been genuine ‘slave societies’ according to Finley’s or Hopkins’ definitions, but that too little evidence survives to make the identification possible. One could take the Thracian kingdoms as an example: according to Xenophon, King Seuthes gave the Greek mercenary force that returned from Persia 120 slaves as a gift; presumably he had much larger holdings (Xen. Anab. 7.7.53). Moreover, Thracian social structures have recently been compared to those of Homeric Greek society, which, as we shall see in Chapter 5, was one in which elites leant heavily upon slave labour for their wealth.²⁴ Of course, one cannot use this proxy data to fill in the gaps in our knowledge of the elite economies of the Thracian kingdoms; nevertheless, do we really know for sure that these Thracian kingdoms were not slave societies? Or we might take another Balkan society as illustrative, the Ardiaioi, an Illyrian people. According to Theopompus of Chios (FGrHist 115 F 40) writing in the fourth century BC, the Ardiaioi possessed 300,000 prospelatai like the helots («᾽Αρδιαῖοι δέ» φησί «κέκτηνται προσπελατῶν ὥσπερ εἱλώτων τριάκοντα μυριάδας»). To be sure, Theopompus’ formulation is somewhat opaque: he fails to state the precise grounds for his comparison between the prospelatai and the helots. Nor can we know the significance of his using the hapax legomenon prospelatai rather than, say, douloi; and, like all ancient ‘statistics’, the figure of 300,000 is up for question. But it seems at least possible that this Illyrian society would count as a ‘slave society’, especially since Theopompus

²⁴ See Sears (2013). That Thucydides—whom Archibald (2015: 388) notes ‘ought to be our most informative historical source’ due to his family connections in Thrace and personal experience of the region—thought that the lifestyle of the barbarians of his own day (and he is most likely thinking of the Thracians, the barbarians he knew best) resembled that of the early Greeks (Thuc. 1.5–6) makes ‘Homeric society’ a priori the most promising analogy for the vexed question of the character of Thracian society; on which, see Archibald (2015) passim.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Slave Societies, Societies with Slaves

101

elsewhere labelled the helots as douloi (FGrHist 115 F 122).²⁵ For what it is worth, one might also note Agatharchides’ comments (FGrHist 86 F 17) on the neighbouring Dardanians, the wealthiest of whom, he claimed, owned over 1,000 slaves, thus in the same range as the wealthiest Thessalian aristocrats of the classical period. A further problem concerns questions of the growing integration of slavery into ancient economies. Finley’s account (see Chapter 5 below) identifies a single turning point, Solon’s reforms in 594/3 BC, which supposedly triggered the emergence of a ‘slave society’ in Greece. Other scholars such as Rihll and Descat have opted for a model that emphasizes gradual change.²⁶ We will examine this problem in greater detail in Chapters 5 and 14; but it is clear a priori that the kind of ‘radar’ exemplified by Finley’s approach only detects fully fledged slave societies. Societies in the process of becoming slave societies, or societies with many but not all of the attributes of the ‘slave society’ Idealtypus, are automatically consigned to the ‘societies with slaves’ box. An alternative method, such as that recently proposed by Lenski, is to think of a scale of intensification, which allows different stages of development to emerge: shades of grey rather than black and white.²⁷ (There is a certain irony here; for we have noted already the problems inherent in Finley’s application of the spectrum metaphor to the issue of status in ancient Greece. A spectrum might have been far better suited to the issue of the relative economic importance of slavery in a variety of ancient economies than the binary-choice model for which Finley opted. In other words, Finley applied the spectrum metaphor to the wrong problem.) Although the concept of slave society was not invented to sort societies dogmatically into these two taxonomic boxes, repeated use of this approach over the years, combined with repetition of the list of five ‘genuine slave societies’, has had unfortunate results. This approach engenders—albeit unintentionally—a mindset wherein absence of evidence amounts to evidence of absence: since these other societies cannot be shown to be ‘slave societies’, they are automatically labelled ‘societies with slaves’ and consigned to the

²⁵ Regarding their status, Ducat (1993): 216 writes ‘le verbe ktasthai et la comparison avec les Hilotes indiquent clairement que les Prospélatai sont des esclaves’. In Theopompus’ time the feature that allowed helotage to be compared to other slave systems was the fact that the helots were a monoglot slave population: see Chapter 6 below. (For Theopompus’ knowledge of this region, see FGrHist 115 fragments 39, 40, 129, and 130.) One should note that, like Sparta, the Ardiaioi were renowned for their inward-looking attitudes and lack of contact with the outside world: see [Arist.] Mir. 138. These were perfect conditions for the development of a slave population that depended to a greater extent on natural reproduction than external supply to replenish its workforce, because it was poorly connected to networks of foreign trade. ²⁶ Rihll (1996); (2011): 69–72; Descat (2006). ²⁷ See Lenski (2018a), who points out the rigidity of Finley’s binary approach.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

102

Greek Slave Systems

investigative scrapheap.²⁸ Whilst the concept of ‘slave society’ has its uses, mechanical sorting of ancient societies into the two aforementioned boxes generates major blind spots in our overall picture of slavery in antiquity, and does not take into account the enormous degree of unevenness in our sources for studying slavery in ancient societies in terms of quantity, quality, and generic diversity.

IV. WORKING AT CROSS-PURPOSES: COMPARATIVE APPROACHES TO CLASSICAL AND NEAR E A S T E R N S LA V E RY When we turn to the ancient Near East, the shortcomings of this mechanical approach quickly become clear. In areas where our evidence comes from several genres, such as Israel, it is less than abundant and, furthermore, plagued by methodological uncertainties regarding the dating of various texts. In other areas where the evidence is abundant, for example Babylonia, it is less than generically diverse (deriving in the main from legal documents), meaning that whilst we may be able to view some aspects of the economy and the role of slave labour in fine detail, generalizations can be difficult to establish. In this volume I do not aim at definitively proving that this or that Eastern society was a slave society. Rather, I aim to analyse the economic role of slavery, particularly in elite circles, and to show that the sharp contrast often asserted between East and West in the use of slave labour is greatly exaggerated. Indeed, slave labour played an important if not crucial role in the economic arrangements of elites in several of the societies treated in Part III of the book; in others we find it developed to a high degree, but not high enough to qualify the society in question as a ‘slave society’. In all these cases, Finley’s judgement that slavery played ‘no role of any consequence’ in the Near East is demonstrably mistaken. One crucial point is worth emphasizing at this stage in our investigation: misapprehension over the nature and importance of slavery in antiquity is not simply an error perpetrated by classical scholars alone. A number of influential scholars of the Near Eastern world have also been working at cross purposes with their counterparts in Graeco-Roman history. There are two strands to this problem. First, a number of scholars of the Near East have compared the importance of slavery in the Near East unfavourably with the classical world due to a misapprehension of the scale of slavery in Greece and Rome. ²⁸ Cf. Vlassopoulos (2016b): 10: ‘The concept of societies with slaves was merely a negative catch-all that could only work as long as interest focused on the better known ancient and early modern slave societies.’

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Slave Societies, Societies with Slaves

103

Furthermore, some tend to collapse the distinction between the Greek and Roman worlds—especially the huge differences in scale between these cultures in their utilization of slave labour—and work instead with a notion of ‘classical slavery’ that is every bit as reductive as that used by some classical historians towards the East. The second strand lies in the criterion of comparison: whereas historians of Greece and Rome have for many decades used the notion of ‘slave society’ and its attendant focus on elites to measure the importance of slavery to their own subjects, this category is conspicuously absent from studies of slavery in the Near East, where many scholars hold their subjects to completely different standards of evaluation. A few examples should illustrate the state of the problem. In a highly influential work on the institutions of ancient Israel, Roland de Vaux argued that slaves were a key source of labour on the estates of the wealthy. In other words, he made a good case for Israel being a ‘slave society’, in Finley’s terms, though he was not (to my knowledge) aware of the category. Nevertheless, he also wrote that ‘In Israel . . . there never existed those enormous gangs of slaves which in Greece and Rome continually threatened the balance of social order.’²⁹ Few scholars of Greek slavery today would agree with the idea of huge gangs of slaves (rather more modest holdings seem more credible for a society such as Athens, for which, see Chapter 8 of this volume), and only in a few regions such as Sparta and Thessaly did the slave population pose a credible threat to the social order.³⁰ De Vaux’s contrast is based on a popular but misleading picture of Greek and Roman society as populated with vast numbers of slaves heavily outnumbering the free and posing a constant threat of rebellion. M. A. Dandamaev made a similar point in his important book on Babylonian slavery during the first millennium BC. In a discussion on slave numbers, he wrote that ‘it is possible that the free population [i.e. of Babylonia] was only two or three times that of the slaves’.³¹ Now, for our purposes it is irrelevant whether or not this estimate is entirely off the mark;³² what is important is that this estimate would place the percentage of slaves in Babylonia between 25 per cent and 33 per cent of the population, in other words, comfortably within the margins of Hopkins’ proportional criterion for membership of the select ‘club’ of slave societies. Nevertheless, Dandamaev immediately follows up his comment by writing: ‘In any case, the number of slaves in first-millennium Babylonia was probably far less than in the developed Classical states.’ Once more, the popular image of classical slavery obstructed a fair comparison. ²⁹ De Vaux (1961): 80. ³⁰ Cf. Harper (2011a): 9 in relation to Roman Italy: ‘Based on little evidence, Beloch, Brunt, and Hopkins had produced estimates of the slave population that were fantastically overblown.’ ³¹ Dandamaev (1984): 218. ³² The comments of Jursa (2010): 37–41 on Babylonian demography underscore, to my mind at least, the limitations of the evidence for addressing such an issue.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

104

Greek Slave Systems

A second misapprehension that one finds among several scholars of the Near East is that for slavery to be a structurally integral institution, slave labour must form the basis of a given society’s total production. Thus Faust, citing de Vaux, has recently written that ‘the economy of the kingdoms of Israel and Judah was probably not slave-based’.³³ If we are to engage with the issue on these terms, then the economies of Greek and Roman society were not slave-based either. In the field of classical slavery, the debate on the scale and role of slave labour has not been concerned with its contribution to overall production for the better part of a century. The case for the importance of slave labour does not stand or fall in relation to its contribution to the aggregate production of the entire society under consideration, but rather to its location and contribution to elite income. When Chester Starr published an article in 1958 entitled ‘An Overdose of Slavery’ seeking to downplay the role of slavery in Greek and Roman society—one that he thought to be highly exaggerated in the scholarship of his time—by pointing out that slavery played a negligible role in aggregate production, he was quickly attacked for not understanding that the significance of slavery lay in its location in society.³⁴ By and large, scholars of the Near East have not utilized the ‘slave society’ approach, but viewed the problem in the same manner as Starr: thus, in a valuable collection of essays on Near Eastern slavery published as recently as 2011, we find the standard disclaimer: ‘Even though slavery was never the dominant source of labor in the Near East . . . ’³⁵ Scholars in both disciplines have thus been working at cross purposes, using completely different standards of comparison of the role and importance of slave labour in the societies that they have analysed, each largely misapprehending the other based on popular caricatures. None of this proves that Near Eastern societies were ‘slave societies’. This is something that requires demonstration, not assertion; and at any rate, it is more important to look at these societies in detail using a comparative approach rather than stake everything on positive demonstration of the ‘slave society’ label, which, as we have seen, is difficult to prove in many cases. From such a detailed comparison, many similarities will emerge, whilst importance differences can be brought into sharper focus. But if we are to compare the role of slavery in Greek and Near Eastern economies, it is crucial that we evaluate them by the same standards.

³³ Faust (2012): 15. Cf. Hans Neumann’s entry s.v. ‘Sklaverei, Alter Orient’ in Der neue Pauly: ‘Zu keiner Zeit jedoch waren Sklaven . . . die entscheidenden Produzenten im Rahmen der Gesamtökonomie.’ ³⁴ Starr (1958); cf. Degler (1959); de Ste. Croix (1981): 54. ³⁵ Culbertson (2011): 14.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Part II Epichoric Slave Systems of the Greek World

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

5 The Archaic Greek World If we leave to one side those few scraps of evidence contained in late sources that purport to describe the origins of slavery in early Greece¹, the focus of any general study of slavery in the archaic period must fall primarily on the Homeric epics and Hesiod’s Works and Days.² Studies of slavery in these texts are nothing new, but a good deal of the work undertaken on slavery in Homer during the twentieth century was produced before the current modern methodologies for studying Homeric poetry historically were fully developed.³ The main influence on much of what has been written on early Greek slavery in recent decades (at least, in Anglo-American scholarship) has been M. I. Finley’s The World of Odysseus (1954). In this seminal work Finley argued that the Homeric poems did reflect a real historical world dating to the Dark Ages, that is, the tenth and ninth centuries BC. Slavery, according to Finley, though present in this world, was not economically important to the elite.⁴ He therefore developed a theory to explain how and why slavery had attained its evident importance by the classical period. In Finley’s view, Solon put an end to debt bondage in Attica; the Athenian elite, stripped of its traditional labour force, turned to imported barbarian slaves as a replacement.⁵ This

¹ For a critique of these sources and the theories of early Greek slavery based on them, see Lewis (forthcoming b). ² References to slaves do occur in Lyric poetry, but not in ways that tell us very much: see Asius fr. 14 [West]; Hipponax fr. 26–26a; 27; 40; 115 [West]; Anon. Theognidea 301–2; 535–8; 1211–16 [West]. The Homeric Hymns provide two vignettes of kidnap by raiders (h. Cer. 119–44; h. Bacch. 6–57), but little else. ³ e.g. Lencman (1966); Mele (1968); Wickert-Micknat (1983). ⁴ Finley (1954): 49–56. Cf. Finley (1981): 155, quoted on p. 89, above: ⁵ Finley (1980): 67–92. Harris (2012): 346–51 points out that Finley’s conclusions were remarkably similar to those of Meyer (1910): 169–212, though the two scholars came from very different intellectual backgrounds.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

108

Greek Slave Systems

process sparked off a ‘rise’ in slavery in the sixth century, after which it is possible to talk of ‘slave societies’ in the Greek world. This view remains to this day the standard account in most quarters.⁶ However, a number of developments have shown almost every element of Finley’s theory to be untenable.⁷ Above all, the idea that slavery was insignificant to the production of elite wealth in Homer has been robustly contested on several occasions. In his groundbreaking monograph Status Warriors: War, Violence and Society in Homer and History (1992), which produced a much more thorough and careful reading of the society portrayed in Homer than Finley’s The World of Odysseus, Hans van Wees showed that slavery was the key labour source in the estate of the Homeric hero.⁸ Van Wees’ analysis of slavery was, however, merely a part of a larger study of the economy of the elite oikos and not a specific study of Homeric slavery.⁹ In a book published in 1998, William Thalmann arrived at much the same conclusion as van Wees, but likewise dealt with the issue in passing without drawing out the wider implications of his position.¹⁰ A groundbreaking recent article by Harris has taken up the issue in greater detail, and levelled decisive criticisms against Finley’s theory of the rise of slavery in the sixth century.¹¹ This chapter builds on these studies, providing an up-to-date analysis of slavery in the world

⁶ e.g. Garlan (1988): 38–9; Patterson (1991): 51; Joshel and Murnaghan (1998): 16; Cartledge (2002); Osborne (2007): 300; Hall (2007): 242; Rankine (2011): 36; Ismard (2015): 26; Gehrke in Der neue Pauly s.v. ‘Sklaverei, historische Entwicklung’. For further references see Harris (2012): 350 n. 15. I do not understand Rihll’s claim (2011: 73) that ‘Finley’s hypothesis . . . connecting the growth of slavery to Solon’s seisakhtheia seems to have been quietly abandoned.’ Rihll (1996) and Descat (2006; 2015) also believe in an archaic birth of slave society in Greece, but connect this to growing levels of commerce with the non-Greek world, a view followed in Zurbach (2013). ⁷ For example, Morris (1986) has convincingly refuted Finley’s dating of ‘Homeric society’ to the tenth or ninth centuries, and shown that it must lie somewhere around the late eighth century. Rihll (1996) has pointed out that Finley’s theory of the rise of Greek slavery only works (at best) for Attica, not for the rest of the Greek world, and furthermore does not explain what happened to the liberated debt bondsmen. Harris (2006): 249–69 has shown that Solon did not abolish debt bondage, only enslavement for debt. ⁸ Van Wees (1992): 49–53. ⁹ See now van Wees (2009), pointing out the importance of slave labour for middling farmers such as Hesiod, and van Wees (2013b): 224–6 on elite slaveholdings. Cf. also Murray (1993): 42, who remarks in passing that ‘The early Greek basileus worked his estate with the help of slaves and occasional hired labour,’ but at p. 240 backtracks to the Finleyan position: ‘it does not seem to have been until the sixth century that slaves became important to the economy’. ¹⁰ Thalmann (1998a): 50: slaves in the Odyssey represent the labour on which the more leisured aristocratic style of living is based. It is their work and its products that support the way of life and the activities of the families at the head of the various oikoi: the feasts and sacrifices, the hospitality, and the (primarily horizontal) redistribution of goods in the form of gifts that is the basic mechanism in the functioning of elite society. ¹¹ Harris (2012). In what follows, I owe a great deal to Harris’ study, but I will attempt to avoid repeating his analysis as far as possible, and refer the reader to his important study. Harris’ view is endorsed in Rose (2012): 101 n. 25.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Archaic Greek World

109

of Homer and Hesiod and discussing the historical implications of this revised picture.

I. METHODOLOGICAL CONSIDERATIONS A few words are required on the issue of using Homer as an historical source and the methodology required for such a task. The most important study in this regard is a splendid article by Ian Morris.¹² Morris built upon and refined much prior scholarship to produce a robust methodology for historians seeking to utilize Homer to answer historical questions. First, he reaffirmed that the oral nature of the epics must reproduce the values and assumptions of the audience of the poems at the time when they were fixed in writing regarding the basic verity and logic of social institutions and human action. The institutions, practices, and ethics of the poems did not relate to a period a century or more before the poet’s own time (as Finley thought), but to the time when the epics were frozen in written form. Second, he showed that the fossilization of the epics in writing must date to either the late eighth century BC or not long afterwards, a position that has now been reinforced (with minor modifications) in an important study by Crielaard.¹³ Third, Morris demonstrated that the poems did not reproduce a straightforward representation of ‘real life’, but one filtered through literary and ideological lenses. Even so, the basic institutions, social practices, and values depicted in the poems can be used as historical evidence for the period around 700 BC (give or take fifty years). Historians might endlessly argue about the historicity of ‘Homeric society’, but few can now quibble with the value of the poems for studying institutions, practices, and values. Since slavery falls in this category, we must focus carefully upon the epics to gauge what they might tell us about the institution in this period. In this chapter I follow van Wees’ method in Status Warriors: I begin by essaying a sketch of the status of slaves and the role of slavery in the Homeric poems (particularly the Odyssey) and Hesiod’s Works and Days, paying close attention to the texts themselves. Only once those sketches have been adumbrated in full can we proceed to ask questions of an historical nature.

¹² Morris (1986). An updated and fine-tuned version of this essay appears in Cairns (2001): 57–91. ¹³ Crielaard (1995), who opts for a date in the early seventh century. Van Wees (2002) argues that the basic shape of the text was frozen in its current form (i.e. written down and preserved, subject to only minor later alterations) in a ballpark between 735–640 BC.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

110

Greek Slave Systems

I I . S T A T U S A N D S L A V E–MASTER RELATIONS What might Homer and Hesiod tell us about slave status and the nature of slave–master relations? Here, we must tread carefully. Several scholars have argued that individuals called dmoes and dmoai in the Homeric poems—terms usually translated as ‘slave’—cannot be treated as ‘chattel slaves’ in any strict sense. The reasons for this vary. Kurt Raaflaub has claimed that in this early period the language of law and legal status is inappropriate and views social realities through an anachronistic juridical lens; relationships were based on power and influence, not law.¹⁴ Walter Beringer and Ian Morris, on the other hand, have sought to characterize the relations between master and slave as kindly and paternalistic, and thus closer to serfdom or clientage.¹⁵ There are several reasons to reject these views and to consider the dmoes and dmoai of Homer to be slaves sensu stricto, that is, individuals owned by their masters. To answer Raaflaub’s objection, we should recall our earlier analysis of property. Far from being an advanced concept or an invention of Roman jurisprudence, private property is a concept found in many societies with a much lower level of social and institutional development than the world depicted by Homer.¹⁶ More broadly, scholars such as Cantarella and Burchfiel have shown that legal norms do exist in the Homeric poems, at least in a sense that legal anthropologists would recognize, i.e. enforceable social norms.¹⁷ Raw power is, of course, significant; but it is supposed to be exercised in a way deemed legitimate, i.e. by following accepted norms. Much of the drama in the poems emerges when characters break these rules. We might take as an example Agamemnon’s arbitrary confiscation of the slave girl Briseis from Achilles, which sets the entire plot of the Iliad in motion: this act arouses widespread indignation among the Achaeans (Il. 19.85–9) and is challenged by Nestor (Il. 1.275–6). Agamemnon eventually admits that he had no right to deprive Achilles of his property, blames his decision on Zeus, who had addled his mind, and offers other slave girls as compensation (Il. 19.192–5; 243–6). Such an episode is only intelligible if we recognize that rules of conduct exist in Homer; here we find the interplay between raw power and rule-based social expectations. Even though we are dealing with a largely preliterate society, then, we may speak of law; and this rule-based order that is enforced through non-institutionalized mechanisms of social disapproval constituted the substructure upon which later, institutionalized, inscribed slave laws were built.

¹⁴ Raaflaub (2004): 291 n. 54. Cf. Harris (2012): 353 n. 28 for a riposte to this view. ¹⁵ Beringer (1960), (1961), (1982); Morris (1987): 178. ¹⁶ Here, Finley was quite correct to insist that the Homeric world was one in which private property was standard: see Finley (1981): 101. ¹⁷ Cantarella (1979); Burchfiel (1994). The point is made in Harris (2012): 353 n. 28.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Archaic Greek World

111

The view of Beringer and Morris is more complicated. They are quite correct to note the paternalistic flavour of slave–master relations, especially in the Odyssey. The poems do often describe the relations between masters and slaves in surprisingly warm terms: for example, Telemachus refers to Eumaeus as ‘father’ (e.g. Od. 16.31; cf. 16.17). This is quite the opposite of what we find in some later classical texts: by contrast, one might note Epicrates’ play Hard to Sell, a fragment of which (fr. 5 K-A ap. Athen. 6.262d) has an old slave ordered around by a beardless youth and called ‘boy’ (pais) despite his seniority in age. Seen in this light, a very different character in slave–master relations appears to exist in the Homeric period compared to the classical period, at least at first glance. However, in holding this view Beringer and Morris read the text as a more or less straightforward description of social relations between masters and slaves, and thus ignore important generic distortions.¹⁸ As Ruth Finnegan has written, ‘poetry may reflect certain aspects of society and express ideas and reactions that are of concern to people of the time—but to take literary forms as representing a direct and full reflection, or as a direct source of social history can only be misleading’.¹⁹ Generic distortions of the relations between masters and slaves in Homer occur in at least two ways. First, as Harris has pointed out, Homeric epic is an elevated genre that lends an august tone to the character of interpersonal relations, avoiding the seamier side of slavery in which low genres such as Old Comedy revel.²⁰ Second, Thalmann’s analysis of the presentation of slave–master relations has exposed its ideological underpinnings: instead of the psychological complexities of real social relations between masters and slaves, the latter are boiled down into two ideal types.²¹ On the one hand is the good slave: we might take Eumaeus as a typical example. In a remarkable scene in Odyssey Book 14, Odysseus enters Eumaeus’ dwelling (here we must remember that Athena has altered Odysseus’ appearance and that he has not yet revealed his true identity to his slave Eumaeus). Odysseus is thus able to test whether an outwardly obedient slave is truly loyal behind his master’s back. Eumaeus proves utterly loyal, bewailing his owner’s absence and waxing lyrical about Odysseus’ fine qualities (Od. 14.37–44; 55–71). Eumaeus thus serves as a kind of idealized caricature of what a slave should be like, displaying total loyalty to his long-absent owner. On the other hand we have ¹⁸ Morris’ position on the status of the dmoes and dmoai does not, therefore, live up to the standards of methodological sophistication set in his fine 1986 study. Cf. his remarks in Morris (1986): 115–16; 120–7. Cf. Wiedemann (1987): 12: This . . . is about how the social world ought to be, not a description of what relations between masters and slaves were actually like in Homer’s time. Not that Homer is unaware of how slaves were treated in reality: ‘I fear he will be angry with me, for the scoldings of masters are severe’, Eumaeus says at 17.188f. ¹⁹ Finnegan (1977): 263, quoted in Morris (1986): 115. ²⁰ Harris (2012): 356. ²¹ Thalmann (1998a): 49–107.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

112

Greek Slave Systems

the archetypal treacherous slave. Devoid of any loyalty, slaves such as Melanthius and Melantho attach themselves to their master’s enemies and behave with insolence to his household guests. The fate of both kinds of slave is predictable: the good slaves are rewarded, and the bad slaves meet a sticky end (Od. 22.462–77). Thalmann argues that the poet is presenting a version of slave– master relations that has been filtered through an ideological lens so as to present the institution in the ideal form that the slave-owning classes (who must have made up some of the poems’ audiences and probably sponsored the poets themselves) would wish to see it function. Arbitrary violence is cut out: slavery is a good thing that benefits the slave; good slaves do well out of the system, and bad slaves merit a well-deserved death. Good masters are always fair, rewarding the good slaves and punishing the bad ones. The poems thus have an ideological slant that aims partially at flattering slaveholders and partially at providing didactic exempla of how to be a good slave and a good master.²² In sum, then, one cannot take the depiction of slave–master relations at face value in the manner advocated by Beringer and Morris.²³ Fisher has argued that the status of the dmoes and dmoai portrayed in the epics is one of ‘chattel slavery’ and has strongly criticized Beringer’s approach.²⁴ In recent studies, Harris, Ndoye, and Schmidt have addressed the question in greater detail.²⁵ Dmoes and dmoai in Homer can be bought and sold (Od. 15.425–9; 15.483–4; Il. 7.467–75; 21.34–44), given as gifts (Od. 4.735–6; Il. 19.245; 23.509–13; 23.700–5), and inherited (Il. 19.330–3). They are the permanent property of their owners: Agamemnon, for instance, tells the priest Chryses that he will keep his daughter as a slave and take her back to Greece to serve him into her old age (Il. 1.29–31; cf. Od. 1.428–42). As owners, Homeric slaveholders can beat and kill their slaves at will (Od. 4.244–6; 22.440–73); and the rest of society is expected to observe the owner’s exclusive rights over his slaves and to intervene when they are arbitrarily violated.²⁶ This is why Odysseus’ heart barks with fury when he discovers that the suitors have been sleeping with a number of his slave girls: since they are his property, the suitors have no right ²² Hunnings (2011). ²³ Several of Thalmann’s points are anticipated by Fisher (1995): 49, who points out characters ‘who are mostly drawn in black-and-white terms’. As he points out, ‘in the household and estate of Odysseus there are, on the one hand, idealized and admirable relationships between masters and slaves, full of decency, fair treatment and trust, and on the other hand there are pictures of slavish insolence and betrayal, which meet with condign punishment’. ²⁴ Fisher (1995): 49–50. For criticism of Beringer’s approach, see also Wickert-Micknat (1983): 132. Zurbach (2013): 978 rightly notes that ‘Surtout, les poèmes homériques attestent la présence d’esclaves marchandises dès les débuts de la documentation textuelle.’ ²⁵ Schmidt (2006); Ndoye (2010): 226–300; Harris (2012). My thanks to Hans van Wees for drawing my attention to Schmidt’s article. ²⁶ Vidale (2002): 257 discusses an early sixth-century vase painting depicting, in his view, the punishment of two slaves; these seem to be bound at the ankles and their heads restricted by some kind of branks. For a similar punishment for slaves in classical Athens, see Chapter 1, p. 41 n. 49 above.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Archaic Greek World

113

to sleep with them, and the slave girls have no right to choose with whom they have sex (Od. 20.5–30; cf. 22.35–41).²⁷ A close examination of the epics reveals that slaveholders possess all the rights and responsibilities associated with private ownership.²⁸ Nor should we suppose that these slaves had any special ‘rights’. Several privileges that might potentially be seen in such a light are present in Homer, for example, the gift of a house, a plot of land, and a wife to loyal slaves (Od. 21.213–16; cf. 14.61–7). Eumaeus, who seems to be held in special esteem by his owner, even has a slave of his own, Mesaulios, a man he bought from the Taphians with his own goods (Od. 14.448–51).²⁹ On a technical point, there is no good reason to view these privileges as rights in a formal sense.³⁰ And in a comparative sense, many examples could be given of the use of material privileges in incentive systems to control slaves and maximize their productivity,³¹ whilst the institution of servus vicarius in Rome should warn us against assuming that Eumaeus need have had any ‘right’ at all to Mesaulios or any other item of property in his possession.³² At any rate, despite repeated attempts to blur status distinctions in relation to Homeric slavery,³³ a detailed examination of the poems shows a sharp and pragmatic approach to the distinction between slaves and non-slaves. Slavery is in no way an undefined, hazy concept: it is seen as an abrupt and catastrophic change of status.³⁴ Phrases used in Homer like ‘the day of slavery’ or ‘the day of freedom’ (Il. 6.455; 6.462–3; 16.831; Od. 14.340; 17.320–3) show that the distinction between slaves and free persons was not a hazy one based on fluid notions such as domination, dependency, or acting in a certain fashion, but a sharp distinction based on a violent transition whereby the victim became at a stroke the property of another person. In historical terms, the issue of slave status in Homer therefore presents a number of interpretative challenges. The content of social relations between slaves and masters is highly stylized for ideological reasons; it tells us much more about the history of ideology than it does about the history of archaic slave–master relations. What is clear, though, is that the Homeric poems paint

²⁷ See van Wees (1992): 136; Thalmann (1998a): 72: ‘for the suitors to sleep with them is a blow at Odysseus’ property, an implicit claim of rival ownership’. ²⁸ See Harris (2012): 354–6 for a detailed demonstration of this principle. ²⁹ Dolius may also have had some slaves of his own: see Od. 24.223. ³⁰ See Lewis (2013): 396–402 and 409–11 for a full explanation of why this idea of ‘rights’ is unwarranted. Kamen (2013): 7 writes of de jure and de facto rights; but there is no such thing as a de facto right: see Honoré (1987). They are merely (revocable) privileges. ³¹ On the incentive system in Homer, see Hunnings (2011). ³² See J. P. Lewis (2012): 13 n. 1 and p. 23. ³³ One of the most recent is that of Patterson (2008), who rather remarkably argues for collapsing any distinction between dmos and thes and treating them both as slaves. For rejoinders, see Lewis (2016); Hunt (2016). Cf. the comments of Harper (2011b): 165. ³⁴ Cf. Ndoye (2010): 236: ‘L’esclave est tout d’abord un individu arraché brutalement à sa société originelle.’

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

114

Greek Slave Systems

an incontestable picture of slaves as alienable property, often acquired from outside the community. This fact hints at long-term continuities stretching back into the Bronze Age.³⁵ What economic role, then, might these slaves have played? Again, we will begin by constructing a sketch of the institution based on Homer and Hesiod, and then proceed to ask historical questions.

III. S LAVERY AND THE ECONOMY: ODYSSEUS ’ E S T A T E The most detailed picture we possess of the economic arrangements of an elite household in Homer can be found in the Odyssey in relation to Odysseus’ estate.³⁶ We may first consider the general layout of Odysseus’ estate, the forms of production employed, and the labour force. Odysseus’ house is located in the town, but he owns a good deal of land in the surrounding area (Od. 4.757). Inside the house itself are fifty slave women (Od. 22.421–2). Most of the male slaves inhabit the countryside.³⁷ Many tend Odysseus’ flocks and herds at some distance from the town itself. When Odysseus (incognito) meets with his slave swineherd Eumaeus in Book 14 of the Odyssey, the latter, not recognizing his master, discusses Odysseus’ wealth in terms of flocks and herds: Odysseus has twelve herds of cattle, twelve of sheep, twelve of pigs, and

³⁵ Julien Zurbach (2016: 363), for instance, has noted in the Bronze Age Pylos tablets a group of 750 slave women, ‘registered together with their children, some being defined by their provenance: “Milesian”, “Chian”, “Lemnian”, etc. They are probably the victims of razzias and slave trade along the Anatolian coast.’ ³⁶ Although Odysseus’ estate is described in far greater detail than that of any of his peers, references from elsewhere in the poem suggest that their economic arrangements are essentially similar to his: cf. Od. 4.49–59 (Menelaos); Od. 9.200–7 (Maron); Od. 11.430–2 (Agamemnon); Il. 19.330–3 (Achilles). See van Wees (1992): 49–53. ³⁷ Finley (1954): 49 wrote that ‘mostly . . . there were slave women, for wars and raids were the main source of supply: there was little ground, economic or moral, for sparing and enslaving the defeated men’. He therefore considered there to have been a female-heavy gender imbalance in the Homeric slave system. Lencman (1966): 256, 258–9, Ducrey (2007): 10, and Gehrke in Der neue Pauly s.v. ‘Sklaverei, historische Entiwicklung’ follow this view. Patterson (1991): 50–1 uses this argument as the grounds for his strange theory of the origins of freedom in Western culture, which he asserts was a value first held among women in dread of capture, and (at p. 55) writes that ‘freedom in Homeric society was not a personal or individualistic value for men’. However, as Raaflaub (2004): 44 has pointed out, a more careful study of the texts show that Patterson has greatly underestimated the importance of male slaves. Harris (2012): 360 draws attention to Garlan’s important point (1988: 33) that the alleged female-heavy gender imbalance is an optical illusion created by the poem’s focus on the household. Cf. Wickert-Micknat (1983): 166: ‘die Dmoes haben in diesem Zusammenhang kaum einen Platz. Geht man indessen davon aus, dass zu ihnen auch die nicht selten agierenden Hirten der verschiedenen Sparten rechnen, dann stellt sich die Vermutung ein, dass im Tatsächlichen der Anteil von männlichen und weiblichen Bediensteten sich einigermassen entspricht.’ Ndoye (2010): 265–6 argues for roughly equal numbers of male and female slaves.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Archaic Greek World

115

twelve of goats that are pastured on the mainland opposite Ithaca. On Ithaca itself he also has eleven flocks of goats (Od. 14.100–4). As Eumaeus points out, these groups of livestock are managed by Odysseus’ slaves and xenoi (Od. 14.102). Each group is responsible for droving a specified number of pigs, sheep, goats, or cows into town at regular intervals, where they are consumed in the household. When we meet Eumaeus, one of his fellow slaves has been sent into town with a fattened swine for the suitors (Od. 14.26–8). Later in the poem, shortly before the suitors are slaughtered, we find Eumaeus himself droving three fattened pigs into town for a feast (Od. 20.162–3). Melanthius the slave goatherd and two of his fellow slaves have also brought goats into town for the feast (Od. 20.173–5), and Philoetius the slave cowherd, who looks after a herd of cattle for Odysseus on Kephallania, has brought a cow and some goats across on a ferry to Ithaca for the feast (Od. 20.185–6). As Harris notes, if we assign one work group of this sort to each of the flocks and herds enumerated at Od. 14. 100–4, we come to a figure of several hundred male workers.³⁸ But this figure does not take into account those male slaves working in the fields growing cereals. Despite the prominent role of animal husbandry in Homeric period, a close look at the Odyssey shows that cereal farming and arboriculture are of comparable importance, and form a symbiotic relationship with animal husbandry.³⁹ Odysseus’ household is full of barley, bread, and wine as well as meat (Od. 1.144–9; 2.290): all are produced from his estate. When he approaches his town house in Book 17 of the Odyssey, the disguised Odysseus is saddened to see his faithful old hound Argos sitting atop a manure heap dying (Od. 17.300). This manure has been collected from mules and oxen and will be spread on the fields by Odysseus’ slaves (Od. 17.297–9). The cereals grown in these fields in this labour-intensive manner are ground into flour by twelve slave women in Odysseus’ house (Od. 20.105–8) and then baked into bread. In fact, although Eumaeus and his fellow slave swineherds subsist off piglets (Od. 14.80–1), they also have an ample supply of bread and wine (Od. 16.51). Animal husbandry and cereal farming go hand in hand on Odysseus’ estate. The estate also contains orchards, which are looked after by Odysseus’ aged father Laertes and his small group of slaves. These men work intensively, whether digging (Od. 24.226–7) and tending the trees (Od. 24.244–7) or gathering stones for constructing retaining walls for the orchard (Od. 24.222–5). The orchard itself contains figs, grapes, olives, and pears (Od. 24.245–7). When Odysseus was a child, we are told that Laertes led him through the orchard and gave him a number of the trees to name and keep for himself: thirteen pear trees, ten apple trees, forty fig trees, and fifty

³⁸ Harris (2012): 360; see also van Wees (1992): 49 n. 86.

³⁹ Cf. Bresson (2016): 132–3.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

116

Greek Slave Systems

vines (Od. 24.340–3). These trees represent only a portion of the orchard managed by Laertes.⁴⁰ And if we return to the town house itself, the fifty slave women therein, in addition to grinding cereals, undertake a wide range of domestic tasks (Od. 20.147–54). Of particular note is textile production: they labour at spinning and weaving alongside their mistress Penelope (Od. 21.350–2). Odysseus’ estate thus represents a powerful productive unit and produces a wide variety of foodstuffs and textiles. As van Wees has pointed out, there is no indication at all that the Homeric estate is geared merely towards satisfying domestic needs.⁴¹ On the contrary, all of the productive tasks are aimed at generating a surplus. This swells the household’s wealth and increases its owner’s power. The produce that is gathered in can be exchanged for metals and the products of itinerant craftsmen. The household’s wealth can be redistributed as largesse through the provision of feasts to increase the number of the basileus’ followers and cement his ties of loyalty with the common people and his hetairoi; it can be used to create and maintain xenia relationships, for which expensive gifts are required. It can be used for marriage gifts, and the best (and most expensive) gifts guarantee the most advantageous match. It can also be used for lavish sacrifices and funerals, further enhancing the basileus’ power.⁴² It should be emphasized that not all of the labourers on Odysseus’ estate are slaves. There are some hired workers: we have already noted the xenoi who help to look after some of Odysseus’ herds on the mainland (Od. 14.102). Their lowly tasks must mean that we translate this term not as ‘guest friends’ but as ‘hired labourers’ from outside Ithaca.⁴³ When Odysseus, disguised as a tramp, enters his own house, the suitor Eurymachus mockingly offers him work as a thes (wage labourer), and promises to pay, house, and clothe him in return for his labour (Od. 18.356–64). References to hired labourers and the precariousness of their position can be found elsewhere in Homer (Od. 10.84–5; Il. 21.441–55), but thetes are at best a supplement to the core slave workforce. All the most prominent workers on Odysseus’ estate are slaves.⁴⁴ Another source of elite income cannot be passed by unmentioned: gifts from the demos, perquisites of the basileus, who, in return, must render straight judgments in local legal disputes and defend his community in times of war and against raids. These, though, must be reckoned a secondary

⁴⁰ For intensive farming in Homer, see van Wees (1992): 49–52, who notes (at p. 50) that ‘in the economy of a princely estate agriculture and viticulture are at least as important as animal husbandry’. ⁴¹ Van Wees (1992): 52–3. ⁴² Harris (2012): 362–4. ⁴³ Murray (1993): 44; Schmidt (2006): 121; Ndoye (2010): 266. ⁴⁴ Eurycleia (Od. 1.427–38); Eumaeus (Od. 15.402–84); Philoetius (Od. 21.244), Dolius (Od. 24.209–25), Melanthius (Od. 17.212; 22.159), Melantho (Od. 18.322). For thetes in Homer, see Ndoye (2010): 170–6.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Archaic Greek World

117

form of wealth, insofar as they are granted only to men who have already achieved a position of wealth and prominence through the methods outlined above.⁴⁵

I V . HE S IO D’S WO RKS AND DAYS: S L AVE RY I N SUB-ELITE AGRICULTURE Hesiod’s Works and Days provides us with evidence for the role of slave labour in farms at a lower rung of the social ladder around the same period, and confirms the basic picture already ventured on the basis of the Homeric poems. Although scholars have questioned whether the dispute between Hesiod and his brother Perses is historical or was, rather, concocted for didactic purposes, the poem nevertheless reveals many incidental details and assumptions about life in Boeotia around 700 BC. As with the holdings of the Homeric basilees, the farm described in Hesiod’s poem is privately owned and alienable (Op. 37–41; 341). Just like Solon a century later, Hesiod praises the acquisition of wealth, but warns against riches gained by foul means. Instead, an ethic of hard work and wealth justly earned is exhorted upon Perses (Op. 306–34; cf. Solon fr. 13 [West]). From parts of the poem that warn of destitution, several scholars have argued that Hesiod was an impoverished peasant farmer engaged in a struggle for survival.⁴⁶ However, van Wees has convincingly shown that the farm he envisages running is far from impoverished: it utilizes slave labour (see below) and enables its proprietor to buy Byblian wine (Op. 589) and aim to produce surplus crops that can be sold overseas (Op. 641–5; 663–93).⁴⁷ Hesiod describes the diversified practices of a medium-sized mixed farm: it has cattle and mules (Op. 46; 405–6; 435–6; 451–4) as well as sheep (Op. 308; 786–7), arable land used for growing grain (Op. 441–7; 571–81; 597–600), and a vineyard (Op. 571–3). This mix of arable farming with animal husbandry is thus similar to what we find described in Homer, only on a much smaller scale.⁴⁸ Hesiod does not propose that the farmer should work this land by himself. Several passages show that, just like the Homeric basilees, Hesiod presupposes ownership of slaves and access to hired workers. He begins his instructions on farming by advising Perses to buy a slave woman (Op. 405–6).

⁴⁵ On these gifts, see van Wees (1992): 85–6; Harris (1997). ⁴⁶ e.g. Millett (1984); Patterson (1991): 57 calls Hesiod a hardworking, ‘barely coping’ farmer. ⁴⁷ Van Wees (2009): 445–50; see also Harris (2012): 361. ⁴⁸ Agro-pastoral farming should also be noted in sixth-century Attica: Solon passed one law against dung larceny (see fr. 64a–b [Leão and Rhodes]) and set a bounty on wolves (which harassed livestock): fr. 81 [Leão and Rhodes].

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

118

Greek Slave Systems

For ploughing and sowing, he recommends men of forty years of age who will not be distracted by their peers (Op. 441–7). When ploughing, he advises Perses to have a slave with a mattock follow the sower and cover up the seed (Op. 469–71). In midsummer, the slaves are to build shelters in which they are to dwell in the winter months (Op. 502–3). When harvest season comes, he exhorts Perses to rouse his slaves and sharpen the sickles (Op. 571–3), and later the slaves are to winnow the harvested grain (Op. 597–9). The slaves are to be apportioned monthly rations (Op. 765–8). Once the harvest has been stored away, Hesiod recommends turning out one’s hired man (thes) and getting hold of a serving girl (erithos) who does not have a child in tow (Op. 602–3). The poem thus describes a medium-sized mixed farm run with an eye to profit, on which the smallholder has at his command a staff of several slaves augmented by hired labour according to the demands imposed by the rhythms of the agricultural year. Hesiod provides good evidence that slave labour was utilized beyond the confines of the early archaic elite: he does not imply that all farmers will own slaves, for some find themselves in dire straits, have no oxen, have to beg from their neighbours, and might have their land bought by others (Op. 341; 399–400; 451).⁴⁹ These poor farmers presumably do not have enough substance to acquire slaves of their own. Hesiod and Perses, however, belong to a ‘middling’ stratum of smallholders somewhere in between this dire position and the relative affluence of the elite. Having set out the picture of labour conditions in Homer and in Hesiod’s Works and Days, we may now proceed to questions of historical interpretation. What aspects of these fictional works might be grounded in historical realities? What generic features obstruct a clear view of the historical situation? What continuities and changes can we posit between the early archaic period and the classical era based on these texts? And how far can we use them to generalize about economic conditions in the early Greek world, given they derive from specific locales?

V. ECONOMIC STRUCTURES First, then, the question of what aspects of the elite estate in the Odyssey may have some bearing on historical realities. On the one hand, clear distortions are present in the sphere of scale. Homer’s heroes are fabulously wealthy. Odysseus, as we have noted, owns several hundred slaves as well as a vast ⁴⁹ Furthermore, the existence of some manner of tenancy at this period is likely, though impossible to quantify in relation to smallholdings of the Hesiodic sort: see Hom. Od. 11.489–91, when the shade of Achilles tells Odysseus that he would rather be a thes working for some ἄκληρος man with a modest livelihood than be king over the dead.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Archaic Greek World

119

estate. The inflation of the figures is better appreciated when seen in relation to the size of slaveholdings of the classical period. The richest Athenians and Spartans might have owned over a hundred slaves, and it is only in the fabulously wealthy dynasts of Thessaly that we can see holdings approaching four figures.⁵⁰ Yet the Greek world of the eighth and seventh centuries had a much smaller population and was diminutive in overall economic scale compared to that of the fourth century. Thucydides (1.10.3) was right to point out that the poets exaggerate: we cannot take a simple headcount of Odysseus’ slaves as historical information on early archaic slaveholdings.⁵¹ On the other hand, the basic structure of labour is perfectly credible and in line with what we find depicted in later periods. First, the reliance of the elite on a core of slave labour supplemented by a limited amount of hired labour is in line with conditions in classical Greece; here, a basic continuity in practices is the most likely answer. Second, the symbiotic mixture of intensive arable farming with animal husbandry, combined with arboriculture, is exactly what we find in fifth- and fourth-century Greece, a subject that we will treat at greater length in Chapter 8. (That is not to posit identical agricultural practices; only that there are some basic similarities.) Furthermore, the use of female slaves in domestic tasks and textile production in the classical period is also best seen in terms of continuity from archaic practices. To reject the evidence of Homer for the structure and organization of slave labour as merely a collection of literary devices not only requires excessive scepticism, but also requires believing that Homer invented a fantasy system of land tenure and labour that by amazing coincidence bears a strong resemblance to real-life practices known from the Greece of several centuries later. It is far more credible to read the presentation of slavery in the poems as an exaggerated version of real historical conditions in the years around 700 BC. The fact that substantially similar pictures of production and farming derive from the Eastern Greek Homeric epics and the central Greek didactic poetry of Hesiod suggests that these conditions were probably widespread across the early archaic Greek world (at least in a very crude sense, and allowing for much regional diversity at the level of fine detail). This means that, pace Finley, the Greek world was already a world of ‘slave societies’ long before the archonship of Solon.⁵²

⁵⁰ See Hodkinson (2008): 315. Critias (88B F 31 D-K) called the Thessalians the richest of the Greeks. ⁵¹ As pointed out by Harris (2012): 360. For the inflation of the size of estates in Homer, see also Morris (1986): 89; van Wees (1992): 55; Ndoye (2010): 194. This observation is not a new one: Trevelyan (1952): 32 n. 1 wrote (in relation to the Anglo-Saxon poem Beowulf) that ‘the poet would naturally tend to exaggerate the hero’s wealth and munificence, like Homer describing the shield of Achilles or the hall of Alcinöus’. ⁵² Harris (2012): 361.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

120

Greek Slave Systems

VI. LA TE ARCHAIC DEVELO PMENTS It is necessary, therefore, to reject wholesale Finley’s view of agrarian labour in the early archaic period. The forms of dependent labour that Finley assured his readers dominated production on elite estates are conspicuous by their absence. We may, therefore, set to one side the task of charting the ‘rise’ of slavery in archaic Greece: at the dawn of archaic history as revealed in Homer and Hesiod we find societies in which slavery is already firmly established and occupying an economic role of critical importance by undergirding the dominant position of elites.⁵³ What we must account for, then, is not a ‘rise’ in the relative importance of slavery to the elite between the eighth–seventh centuries and the classical period, but alterations in its character, scale, and organization. Archaeological surveys reveal a broad trend of development across mainland Greece from small, nucleated settlements towards much larger settlements, combined with a spread of farmsteads across the countryside by the late sixth century.⁵⁴ Whilst there was much continuity in agricultural practices, then, settlement patterns changed. So too did the population and economy. The tempo of change was different from place to place, and the nature of development also varied. In Attica, by the classical period, the slave population was overwhelmingly comprised of non-Greeks.⁵⁵ We, of course, find trade in barbarian slaves in the Homeric world (Hom. Od. 24.389; 15.417; Il. 6.289–92), but it is clear that the relative importance of this supply strategy had increased greatly by the classical era.⁵⁶ On the other hand, privately organized sea raiding had become much less common (at least, among the Athenians), and enslavement for debt—a problem at the end of the seventh century—had all but disappeared from some regions such as Attica.⁵⁷ The Athenians also managed to disburden

⁵³ Cf. Harris (2012): 364. ⁵⁴ For a recent summary, see Foxhall (2013); see also Bintliff (2012): 207–51. Davies (1981): 38 suggests Homeric-style slaveholding in Solonian Attica. ⁵⁵ Garlan (1987); see also Lewis (2011) on the ethnic make-up of the Athenian slave population. ⁵⁶ On slave trading in the Homeric poems, see Collombier (2002). On the growth of trade and contacts with the barbarian world, see Rihll (1996). Descat (2006) places much emphasis on Theopompus FGrHist 115 F 122 as echoing the ‘turning point’ when esclavage marchandise expanded massively in the archaic world due to the introduction of silver bullion as a means of exchange; Descat’s theory is followed by Zurbach (2013). Whilst Descat is surely right that the use of silver facilitated transactions and catalysed the slave supply, it should be noted that Theopompus does not mention silver (the fragment is from Athenaeus Deip. 6.265b–c, and it is Athenaeus’ introduction to the passage that uses the term argyronetos, not Theopompus). ⁵⁷ On sea raiding, see the superb study of van Wees (2013a), and for the practice viewed in the longue durée with attention to regional differences, see Gabrielsen (2001): 223–7, showing that the old ways persisted into the Hellenistic period in regions such as Crete and Aetolia. On enslavement for debt as a widespread problem, see Lewis (2017), and for its eradication in Attica, Harris (2002a).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Archaic Greek World

121

themselves of the rent-seeking strategies of the rich, i.e. strategies of upward redistribution of the resources of the demos through systems of compulsory gifts: the system of payment of one-sixth of the crop to one’s local lord was abolished by Solon; and the Peisistratid system of a payment of a tenth (or perhaps a twentieth) of the crop to the tyrant disappeared with the fall of the tyranny at the end of the sixth century.⁵⁸ Thereafter, the Athenians managed to turn the tables on their elite, and through the liturgy system effected a downward redistribution of resources, harnessing elite wealth for the good of the community. Essentially, then, the main strategies of slave supply were reconfigured during the archaic period, with strategies such as raiding and debt losing ground to trade with the barbarian peripheries of the Aegean world, though we should not ignore regional variations. Economic developments too changed the character of slavery in Attica. The growth of commerce and manufacturing opened up a variety of new ways of exploiting slave labour, as well as ensuring that the Athenians were firmly connected to a network of trade routes and foreign contacts that fed the Attic slave market with a supply of cheap foreign workers.⁵⁹ Athens’ use of slavery in the silver mines of Laurion is an obvious example of one method of exploitation that could not be practised in all regions of the Greek world. Other Greek economies did not develop along the same trajectory as that of Attica. In the southern Peloponnese, the sixth century witnessed the meteoric rise of Spartan power and the development of her distinctive institutions.⁶⁰ One such institution was helotage; its origins are shadowy and much debated, but it at least had more in common with the agrarian slave systems of the early archaic world exemplified by Homer and Hesiod, albeit with a number of distinctive local innovations.⁶¹ To the south, the numerous poleis of Crete developed several institutions during the archaic period that attracted remarks from later Greek authors regarding their similarity to Spartan institutions (Hdt. 1.65; Ephorus FGrHist 70 F 149; Arist. Pol. 1271b; Plu. Lyc. 4.1–2). The Cretans, too, remained firmly rooted in a simple agrarian economy and did not develop commerce or manufacturing to anywhere near the degree of their Athenian counterparts, nor did they follow the Athenians in giving up the old

⁵⁸ See Harris (1997). ⁵⁹ See Chapter 8, section I, pp. 169–70 below. One might note here the shepherd Libys, whose graffito, scratched on a patch of rock on Barako hill in Attica, has been discovered by Merle Langdon: see Langdon (2015): 53. We cannot be certain, but the name probably points at foreign origins and slave status. ⁶⁰ See Hodkinson (1997); Nafissi (2009). ⁶¹ On comparison between Homeric and Helotic slavery and a new theory of development, see Lewis (forthcoming b); cf. Ducat (2015): 193–4.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

122

Greek Slave Systems

practice of privately organized sea raiding.⁶² Slavery in Thessaly, though, remained far closer to its Homeric antecedents.⁶³ These regional variations in socio-economic development shaped their individual slave systems into distinctive epichoric variants.⁶⁴ The principal point, though, is that slavery developed organically in these centuries. As Ducat has recently put it, ‘no exceptional event, no rupture, is required to explain how Homeric slavery could evolve into the different forms of slavery practised in classical Greece’.⁶⁵ Similar processes of regional variation can be seen in relation to the issue of slave status, and here a few concluding remarks may be in order to see how our analysis might fit into broader debates over status in archaic Greece, which have been especially lively in France in recent years (though mainly in relation to elites and citizens rather than slaves). One strand of thinking is exemplified by Alain Duplouy’s important work on archaic elites, which has challenged the traditional view of archaic ‘aristocratic’ status. These elites, according to Duplouy, were not rigid castes like the aristocracies of Medieval Europe; rather, membership of their ranks was determined through recognition by the ‘in-crowd’ and achieved via performative behaviours such as martial and athletic prowess, material display, religious observances, commensality, and so on. Conversely, to maintain elite status one needed to constantly (re-)enact these behaviours. The archaic elite, then, was not a legal status group, but a social class whose membership was contingent on—in Bourdieusian terms— behaviour fuelled by sufficient kinds of ‘capital’, social, economic, and cultural.⁶⁶ In terms of archaic elites, Duplouy’s view has much to recommend it.⁶⁷ However, there is a danger in viewing all aspects of status in the archaic period in these terms. Here, we must again draw attention to the distinction made in Chapter 2 between institutionalized and non-institutionalized aspects of status, and their corollaries, legal status groups as opposed to social classes. These cannot be compressed into a single ‘spectrum of statuses’ but must be viewed as distinct, yet interactive aspects of social life. ⁶² Best attested for the Hellenistic period: see Brulé (1978); but even fifth-century writers like Herodotus (1.2), when confronted with early stories of raiding and kidnap, tended to blame the Cretans; and perhaps the finest vignette of archaic sea raiding, that in Hom. Od. 14.199–359, concerns a Cretan raider. ⁶³ Ducat (2015): 194. ⁶⁴ Cf. Murray (1993): 46: ‘the various forms of obligated servitude found later in Dorian communities like Sparta . . . are not individual survivals of a general phenomenon, but special developments conditioned by the history of each area’. ⁶⁵ Ducat (2015): 194: ‘aucun événement exceptionel, aucun rupture, ne sont nécessaires pour expliquer que l’esclavage homérique ait pu évoluer vers les diverses formes d’esclavage pratiquées dans la Grèce classique’. ⁶⁶ See Duplouy (2006) passim. ⁶⁷ See also van Wees and Fisher (2015) for a critique of the term ‘aristocracy’. For some specific and well-taken caveats on Duplouy’s view, arguing that something resembling a true ‘aristocracy’ (or at least aspects thereof) existed in relation to the Bacchiads at Corinth and the Eupatridae at Athens, see van Wees and Fisher (2015): 6 and Pierrot (2015) respectively.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Archaic Greek World

123

Julien Zurbach has criticized the fashion in French scholarship for rejecting legal categories tout court and viewing social life solely in terms of fluid performative dynamics. In summarizing much recent work, he writes: ‘the concept of legal status is seen as an excessively philological, legalistic, and traditional approach that stands accused of various shortcomings and a few obvious errors: anachronism, imperialistic use of concepts drawn from Roman law, and ignorance of archaeological sources and recent research’.⁶⁸ But when we turn to slavery, we need to include precisely these institutionalized aspects of status in our purview, for these formal features also have a history. Statuses like ‘slave’ or ‘helot’, Zurbach points out, ‘can only be understood from a legal and collective point of view—in other words, they define groups whose members are characterized by the same bundle of rights and duties’.⁶⁹ Zurbach’s point is well taken. We are likely to misunderstand the dynamics of archaic society if we opt for either extreme, trying to explain everything either in terms of a rigid legal taxonomy or, alternatively, in terms of fluid ‘performative’ social dynamics. Both elements need to be accommodated in our understanding, and their complex interplay properly appreciated.⁷⁰ The ‘elite’ is certainly not a legally defined status group, and we are likely to seriously misrepresent its nature by analysing it in legal terms. How does this debate relate to archaic slavery? On the one hand, one does not fall inside or outside the category of ‘slave’ by acting in a certain way, and it is here that paying attention to formal criteria is more important. Ownership of persons, signified by the ability to buy and sell them, is what distinguishes slavery from other forms of lowly circumstance (Anon. Theognidea 1211–16 [West]). On the other hand, we can see the complex interaction between the dynamics of social behaviour and institutionalized status in early laws that aim to freeze and ring-fence, in institutional form, privileges or behaviours associated with elite citizens that some slaves, at least, must have been trying to imitate: note, for instance, the Solonian law forbidding slaves to exercise in the gymnasium, have a free boy as a lover, and have an oil rubdown (Solon fr. 74 [Leão and Rhodes] = Hermias Alexandrinus ad. Pl. Phdr. 231e; Plu. Sol. 1.6;

⁶⁸ Zurbach (2013): 968: Les classes ont disparu, le rejet hyper-critique des sources textuelles permet de renvoyer la notion de statut juridique à une approche excessivement philologique, juridique et traditionnelle, accusée de divers torts, voire de fautes évidentes: anachronisme, impérialisme de notions issues du droit romain, ignorance des sources archéologiques et des développements récents de la recherche. ⁶⁹ Zurbach (2013): 984: ‘Ces statuts ne se comprennent que du point de vue juridique et collectif, c’est-à-dire qu’ils définissent des groupes dont les membres sont caractérisés par un même faisceau de droits et d’obligations.’ ⁷⁰ Cf. van Wees and Fisher (2015): 34–41 for sensible comments on the polarizing debate between ‘class’ and ‘status’, both of which are important (and not mutually exclusive) concepts in social history.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

124

Greek Slave Systems

Conv. Sept. Sap. 152d; Amat. 751b; Aeschin. 1.138–9; John Chrysostom, ap. Migne, Patrologia Graeca 60.418–19; 62.693). A similar law existed in Crete, and banned slaves from possessing arms or attending the gymnasium (Arist. Pol. 1264a20–2). These went beyond the banal legal distinction between slaves and free persons, and aimed to forestall the former from gaining the social capital accrued by engaging in behaviours appropriate for citizens alone. As the curmudgeonly Old Oligarch observed, though, slaves in Athens during the classical era managed to find ways to do just that ([Xen.] Ath.Pol. 1.10–12). * * * In the chapters that follow, we will look at several regional slave systems of the classical era, analysing slavery in Sparta (Chapter 6), Crete (Chapter 7), and Attica (Chapter 8). In each case I shall try to gauge both local idiosyncrasies in rules regarding slave status and the place of slavery in the wider economy; and I shall endeavour to explain these features in their local and regional contexts.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

6 Helotic Slavery in Classical Sparta ‘Those who are the most enslaved are to be found in Lacedaemon, and those who are the most free.’ Critias, late fifth century BC¹

Spartan helotage represents perhaps the most peculiar slave system studied in this book. In light of the variety of scholarly opinions on the helots, the bulk of this chapter will be devoted to analysing their legal status in comparative perspective, because this is one of the most fraught questions relating to this group. Much traditional scholarship holds that the helots were serfs² or that they occupied some shady status ‘between free men and slaves’; even our standard Greek lexicon defines the helots and other analogous groups using the language of serfdom.³ Scholars have long sought to reclassify the helots as something other than slaves simply because in toto, helotage was rather different from the Athenian form of ‘chattel slavery’. Whilst significant differences between helotage and the Athenian form of slavery are undeniable, it does not automatically follow that the helots were not slaves. This chapter aims to analyse the relevant sources afresh with a source-critical method informed by recent work on Spartan history and historiography. Once we are better able to appreciate the legal position of

¹ Critias 88B F37 D-K = Liban. Or. 25.63: Κριτίας φησὶν ὡς μάλιστα δοῦλοί τε ἐν Λακεδαίμονι καὶ ἐλεύθεροι. The saying is also paraphrased at Plu. Lyc. 28.5; cf. Isoc. 6.96, who calls the helots ‘those whose slavery was harsher than any other’ (τοὺς χαλεπώτερον μὲν τῶν ἄλλων δεδουλευκότας), whilst Theopompus (FGrHist 115 F 13) says that the ethnos of the helots is in a cruel and bitter state (ὁ δὲ τῶν εἱλώτων ἔθνος παντάπασιν ὠμῶς διάκειται καὶ πικρῶς). The helots, according to Xenophon (Hell. 3.3.6) wanted to eat the Spartiates raw (ὠμῶν ἐσθίειν), a Homeric expression (cf. Hom. Il. 4.35: ὠμὸν βεβρώθοις) used elsewhere by Xenophon (Anab. 4.8.14). ² A view that has a long history and continually crops up in basic textbooks over the generations; see Grote (1862): 496; Kitto (1951): 90; Bury (1962): 82; Snodgrass (1980): 88; Davies (1993): 39; Chrimes (1949): 300; Michell (1952): 76–7; Kiechle (1963): 161. The idea that the helots were ‘bound to the soil’ already appears in literature of the late eighteenth century: see Hodkinson and Hall (2011): 81–2. ³ See LSJ s.v. Εἵλως. Cf. the entries for κλαρῶται, πενέστης.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

126

Greek Slave Systems

the helots, we may reintegrate this population into the core study of Greek slavery rather than relegating it to the sidelines as some manner of strange, irrational Spartan practice about which little can truly be known. This chapter will analyse Sparta’s slave system in three phases. First, we shall look at the issue of helot status. What were the legal contours of helotage in the classical period? Does their status and the emic, Greek understanding of helotage align with our cross-cultural, etic definition of slavery in terms of private ownership? Second, what was the contribution of helot labour to the Spartan citizen class, and how does this compare to contemporary slave systems such as that at Athens? Finally, we will close the chapter with an exploration in the history of concepts. In the fourth century BC, Plato, Aristotle, and Theopompus compared helotage to other systems of slavery in the Greek world. What was their criterion or criteria of comparison? What feature (or features) of helotage did Plato and Aristotle believe that these other systems shared? Was it a peculiar status that they had in common? Or was it something to do with the structural character of the systems themselves? Here, we revisit the evidence for this category, if it is indeed anything as clear-cut as a category, and highlight the conceptual discontinuity between the fourth-century sources and the statement of Pollux in the second century AD (upon which many historians have relied) that helots and analogous groups lay ‘between free men and slaves’ (metaxu eleutheron kai doulon).

I . L E G A L ST A TU S In an important essay on the historiography of Spartan history entitled ‘The Invention of Tradition in Classical and Hellenistic Sparta’, Michael Flower deftly sketched the forces that shaped and reshaped images of Sparta from the classical period onwards.⁴ An enduring image of Spartan society is the notion that its institutional structure and social practices were devised ex novo by one man, Lycurgus, long before the classical period, and remained fixed and immutable thereafter. This picture is now widely rejected: Spartan institutions were dynamic and evolved continually, but change was generally represented in a ‘Lycurgan’ garb, as a return to an earlier set of Lycurgan practices that had fallen into abeyance. During the past three decades few areas of Greek history have witnessed such a sea change as Spartan studies. The basic principle that late sources are often untrustworthy when it comes to reconstructing the institutions of the classical period, and where they conflict with classical sources, the earlier sources should be taken seriously, is now axiomatic.⁵ One ⁴ Flower (2002). ⁵ For a classic study in scepticism, see Starr (1965). Kennell (2010) provides a useful summary of the current state of the field based on a more critical appraisal of the sources.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Helotic Slavery in Classical Sparta

127

might cite the issue of Spartan land tenure as illustrative. Plutarch gives two variant models of the ‘Lycurgan’ system of land tenure (Plu. Lyc. 8.3, 16.1; Agis 5.1–2), neither of which stands up to scrutiny when compared to the picture of Spartan land tenure described in Aristotle’s Politics (1271a15–b6), never mind the fact that Plutarch’s two models are themselves mutually incompatible (and in practical terms rather far-fetched). Following the more reliable picture in Aristotle’s Politics, which is not (unlike Plutarch) contaminated by the propaganda and social remodelling of Sparta’s third-century revolution,⁶ most scholars now agree that land ownership was predominantly private in character during the classical period.⁷ The idea that all land was centrally owned and controlled by the state is rightly seen as a later arrangement anachronistically projected backwards on Sparta’s archaic and classical past by late authors. This revolution in scholarly approaches to Spartan history must set the basic methodological parameters of our study of helotage. Since Spartan society was not static, this chapter will not attempt to reconstruct helotage in its original form, but will focus upon the one century for which we have a number of good sources: the period from the Peloponnesian War down to the time of Aristotle. This approach privileges the classical sources as a much safer guide to classical realities than the approach that has remained typical of many scholars of helotage, giving undue credence to later sources and ignoring or explaining away the classical material. Just as land is represented as state-owned and administered in the late sources, so do late sources present the helots as a group collectively owned by the Spartan state.⁸ Discussion of the status of the helots must begin, therefore, with the question of private or public ownership. Detlef Lotze, for example, viewed helotage as ‘Kollektivsklaverei’, in which the helots were enslaved as a group to the state, though assigned to individual citizens.⁹ Yvon Garlan’s categorization of the helots and helotic groups under the rubric ‘communal servitude’ does not differ greatly in substance from this formulation.¹⁰ In an elaborated form of the public ownership argument, G. E. M. de Ste. Croix, followed in a number of studies by Paul Cartledge, has categorized the helots as ‘state serfs’. On this interpretation, the state owned the helots en masse, but they were individually assigned to a Spartiate master to serve on his

⁶ On the ‘revolution’, see Cartledge and Spawforth (1989): 35–53; Hodkinson (2000): 43–5; Kennell (2010): 159–80. ⁷ See Hodkinson (1986), with updates in Hodkinson (2000): 65–112. A notable dissenter from this consensus: Figueira (2004). ⁸ Pausanias (3.20.6) calls them ‘slaves of the community’ (δοῦλοι τοῦ κοινοῦ), whilst Strabo (8.5.4) writes ‘for the Lacedaemonians held them in a certain manner as public slaves’ (τρόπον γάρ τινα δημοσίους δούλους εἶχον οἱ Λακεδαιμόνιοι τούτους). ⁹ Lotze (1959): 27–47. Lotze later changed his mind on this designation and referred to helotage as a form of ‘Unfreiheit’. See Lotze (1967): 14. ¹⁰ Garlan (1988): 85–8.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

128

Greek Slave Systems

land; because that master did not own the helot, he could not sell him or her, effectively binding the helot in question to the land of the Spartiate master for the lifespan of the latter. According to de Ste. Croix and Cartledge, the helots were a sort of public slave, at least in formal legal terms; but in terms of their social condition, they were more like serfs, insofar as they had to render ‘tribute’ to their master and were tied to the soil, but not owned by him.¹¹ M. I. Finley took a slightly different tack. He rightly criticized the tendency of many historians to categorize any subordinate group that did not neatly fit the categories of slavery or debt bondage as ‘serfs’. In his view, the helots could not easily be fitted into any of the usual taxonomic boxes; he therefore proposed viewing their status as sui generis and locating it on his spectrum of servile statuses ‘between slavery and freedom’, a label, as we noted in Chapter 2, paraphrased from the scholar Julius Pollux, writing in the second century AD.¹² A close look at Finley’s argument, however, shows that like Lotze, Garlan, de Ste. Croix, and Cartledge, he too followed the information in Strabo and Pausanias that presents the helots as collectively owned by the state.¹³ The sheer weight of authority that these formidable scholars have exerted on the debate, paired with the long reign of the serfdom/public ownership view, has cut a deep rut of intellectual path dependence from which it is difficult to extricate oneself. Nonetheless, thinking away this long tradition of thought and approaching the sources afresh is precisely what is required if we are to make sense of the status of the helots during the classical period. The French historian Jean Ducat has exhibited a particularly admirable ability to set long-held certainties to one side and re-examine key texts afresh with critical acumen. He has pointed out an insuperable problem with the traditional view of state ownership: it conflicts with a much older source, Ephorus, writing in the fourth century BC.¹⁴ According to Ephorus (FGrHist 70 F117), the helots were slaves serving under special conditions. The master of a helot could neither manumit him nor sell him ‘beyond the boundaries’

¹¹ De Ste. Croix (1972): 89–93; (1981): 149: ‘I shall treat them as the “State serfs” they undoubtedly were.’ The term ‘state serf ’ is present already in Michell (1952): 75. Cf. de Ste. Croix (1988); Cartledge (1979): 142; (1988). The state serf label is followed in Harvey (1988): 47; Hall (2007): 238; Golden (2008): 52; Tordoff (2013): 5–6; cf. Kyrtatas (2011): 92. Schumacher (2001): 13 considers the helots to have been ‘unfree’ but not slaves. ¹² Finley (1981): 116. Oliva (1971): 43–4 also believes that the helots ‘can hardly be placed in any of the existing categories of bondage’ and argues that helotage should be viewed as ‘an undeveloped type of slavery’. ¹³ See Finley (1981): 266 n. 34. State ownership of the helots is asserted in Jones (1967): 9; Forrest (1968): 31; Toynbee (1969): 198; Hooker (1980): 118–19; Fitzhardinge (1980): 156–7; Wiedemann (1987): 41; Papazoglu (1995): 370; Hunt (1998): 18 n. 82; Flaig (2009): 39; Thommen (2014): 39. ¹⁴ See Ducat (1990): 19–29. On Ephorus’ interest in Sparta, see Christesen (2010).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Helotic Slavery in Classical Sparta

129

(ἔξω τῶν ὅρων).¹⁵ Various feats of intellectual gymnastics—all of them unconvincing—have tried to elide seamlessly the information of Ephorus with that of Strabo and Pausanias: some are untenable on a practical level;¹⁶ others are untenable in terms of imputing hidden meaning to Ephorus’ phrase;¹⁷ yet others simply bypass the incongruity and focus on the late sources, thus not confronting the jarring incompatibility of the two pictures.¹⁸ Clearly, the only acceptable solution is Ducat’s: that in Ephorus’ day, Spartiates could sell their helots, but not outside Spartan territory, and could not manumit them.¹⁹ The right of Spartiates to sell their helots shifts the entire discussion away from public ownership towards private slave ownership, and thus closer to the mainstream practices—at least from a legal point of view—of the rest of the Greek world. The whole question in many ways parallels the debate over land tenure, which was once thought to be public in character, but is now rightly recognized by most scholars as private, as was the norm elsewhere in Greece. Given the now well-known problems with late sources in reconstructing classical Spartan society, the most methodologically sound approach is to see whether or not a coherent picture of helotage can be constructed from the classical (fifth- and fourth-century) sources. If not, doubts may linger about the exact meaning of Ephorus’ statement. If the classical sources do provide a coherent picture, however, we can then reject the notion of state ownership as a late intrusion anachronistically projected onto earlier periods, much like the ideas in Plutarch about state ownership of Spartiate landholdings. ¹⁵ Morrow (1939): 19 erroneously claims that the succeeding lines, in which Strabo describes the helots as public slaves, form part of Ephorus’ quotation, but the quotation ends before this point. See Hodkinson (2000): 117 for the correct interpretation. Ando (1988): 323 commits the same error as Morrow. ¹⁶ MacDowell (1986): 35 argued that the ‘boundaries’ are the boundaries of individual farms, but see the criticisms of Ducat (1990): 22 n. 13. ¹⁷ Jeanmaire (1939): 478 argued that, alongside a total ban on manumission, we should suppose that Ephorus implied a complete ban on sale. Cf. Thommen (2014): 39, who simply asserts that they could not be sold, without discussing the Ephorus passage. But as Luraghi (2002): 228–9 rightly observes: only preconceived ideas about Helotry can explain how some scholars have been able to interpret this clause as if it meant that it was forbidden to sell Helots altogether. A quick look at the text shows that, in order to convey that meaning, it would have been enough to conclude the sentence with πωλεῖν, without mentioning the borders. ¹⁸ e.g. de Ste. Croix (2002): 191, using Ephorus to point out that individual Spartiates could not manumit their helots, but ignoring the clause on sale. Cf. Papazoglu (1995): 370 n. 3. On the claim that the helots had special rights, see the decisive criticisms of Luraghi (2002): 228–33; see also Lewis (2013): 396–402 for methodological discussion on the idea of slave ‘rights’. ¹⁹ Ducat (1990): 21–2. Ducat’s understanding of Ephorus’ phrase is the same as Klees (1975): 146; this position is endorsed by Luraghi (2002); Nafissi (2009): 121; Kennell (2010): 75–88; Lewis (2013); Zurbach (2013): 973; Paradiso (2013): 49–50; Ismard (2015): 26. Kennell (2003) convincingly argues that the state ownership of helots is a feature of Hellenistic Spartan society that has been retrojected onto earlier periods. Hodkinson (2000): 113–17 takes a very similar position on private ownership but is less certain on the capacity of owners to sell helots (at 119).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

130

Greek Slave Systems

I propose to examine this issue in three steps: first, an analysis of the classical sources that relate to the status of helots and the rules that governed their treatment and management. Second, an explanation of why these rules were formulated and how they relate to Sparta’s other institutions, forming a logical and coherent whole. Finally, a comparison of the legal position of helots to slaves more generally, using the cross-cultural formulation of ownership developed earlier in our analysis: this will allow us to assess rigorously the validity of our application of the term ‘slave’ to helot status.

The Legal Status of the Helots: An Anatomy of the Classical Sources Ephorus does not stand as a lone voice in characterizing the helots of the classical period as privately owned, yet subjected to forms of what Hodkinson calls ‘communal intervention’.²⁰ Xenophon and Aristotle describe a similar practice. Xenophon (Lac. Pol. 6.3) writes: He [i.e. Lycurgus] devised it that someone could make use of another’s slaves if he were in need of them. And he likewise made hunting dogs available in common, so that those without them invite the owner to the hunt, and if he lacks the free time to join them, gladly sends along the dogs; and they use horses the same way. For if a man falls ill or needs a mule cart or wishes to go somewhere quickly, and should see a horse, he takes it and having made use of it later restores it properly.

In a similar vein, Aristotle (Politics 1263a35–7) argues that property should be privately owned but used in a fashion that enables borrowing and a certain level of communal use. He points out that in the Sparta of his own day such a use was made of private property: ‘and in Sparta they use the slaves of others as if they were their own, and likewise horses and dogs, and if they should be in need of supplies they use the produce of the fields throughout the countryside’.²¹ The fact that Aristotle placed Spartan slaves in the category of private property subject to some level of communal use dovetails with the notion of helot status found in Ephorus, i.e. of privately owned slaves subject to unusual state controls. Is it possible that these slaves could simply be state-owned and shared out among the citizenry, along the lines of Strabo and Pausanias’ description of helot status? This is hardly likely. For one thing, it is irreconcilable with the very specific implication of Ephorus that helots could be sold by their masters, albeit within Spartan territory alone, implying private ownership. For another, we should have to suppose that dogs and horses too

²⁰ Hodkinson (2000): 115. ²¹ Hodkinson (2000): 200–1 believes that Aristotle here repeats and modifies Xenophon’s comment rather than deriving his information from an independent source.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Helotic Slavery in Classical Sparta

131

(which are grouped alongside slaves by Xenophon and Aristotle) were stateowned but shared out among the Spartiates, a scenario that strains credulity too far. Besides, the idea of a state-controlled ‘pool’ of helots, who were shared among individual Spartiates, implies a bureaucracy of some elaboration, for which we have no evidence and which seems extremely unlikely during this period.²² What we might call ‘communal intervention’ went much further, however, than merely the occasional borrowing of a helot without the permission of his owner. The Old Oligarch, writing in the 420s, laments the inability of Athenian citizens to strike capriciously any slaves within their reach, a practice that he fondly notes is possible in Sparta ([Xen.] Ath. Pol. 1.10–12). This ability of non-owners to beat slaves is further attested by the Hellenistic historian Myron of Priene, known for his pro-Messenian and anti-Spartan attitudes.²³ Myron writes that the Spartans forced the helots to wear degrading clothes, prescribed for them an annual whipping (to remind them they were slaves), and permitted magistrates to execute any helots whose physical size and vigour they thought inappropriate for a slave—fining the helot’s owner as well for letting him reach such girth (FGrHist 106 F2 ap. Athen. 14.74. 657c–d).²⁴ This might sound sensational, but the rationale of such measures— that is, targeting large, powerful helots who might be potential rebel leaders—is observable in fifth- and fourth-century sources as well. Aristotle’s description of the krypteia is perhaps the best example of this. According to his account, from time to time youths would be sent into the countryside armed with daggers and carrying only the most meagre rations. They would hide during the day, emerging at night to kill any helots whom they found abroad on the roads (surely some manner of curfew is implied). They would also infiltrate the fields and dispatch those who were particularly large or vigorous (Arist. fr. 538 [Rose] ap. Plu. Lyc. 28. 1–4; cf. Arist. fr. 611.10 [Rose] ap. Herakleides fr. 10 [Dilts]).²⁵ Aristotle further noted (ap. Plu. Lyc. 28.4) that the ephors declared war on the helots each year, thus absolving those Spartiates (surely for the most part the krypteia) who murdered helots from miasma, effectively granting them a

²² Cf. Hodkinson (2000): 72–5. ²³ I include Myron for the sake of completeness, but of all the sources discussed in this section this writer should be treated with the most caution. ²⁴ Alain Bresson has drawn my attention to Steph. Byz. s.v. Λακεδαίμων, who notes a special kind of whip distinctive to Sparta (‘εἰσι καὶ μάστιγες Λακωνικαί’), although it is, perhaps, unlikely that this was in the first instance a slave whip. Hornblower (2000) convincingly shows how the Spartan staff (bakteria) was a popular instrument of corporal punishment. ²⁵ This second method—that is, of infiltrating the fields—falls outside Rose’s Aristotelian quotation (or paraphrase). Rose ends with ἀπέσφαττον, and he is supported by Ducat (2006): 287; but I see no reason why the sentence in Plu. Lyc. 28.5 (πολλάκις δὲ . . . κτλ.) should not too derive from Aristotle. There is no contradiction in the two practices, provided one does not follow Ducat’s overly mechanical interpretation that the krypteia operated exclusively at night. See n. 35 below.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

132

Greek Slave Systems

‘licence to kill’. All of these sources tally with the remark of Critias (in the epigraph to this chapter) that at Sparta were to be found the most enslaved and the most free of men, which must refer to helots and Spartiates: helots—of all Greek slaves—were subject to the most oppressive and unusual means of control; Spartiates—of all free Greeks—were those most liberated from the need to perform manual labour.²⁶ Let us sum up so far. The picture painted by our contemporary sources is that individual Spartans were able to own helots, and this included the power to sell them, though not outside Spartan territory; they could not manumit them in any way whatsoever. Furthermore, their helots could be commandeered by their fellow citizens for minor tasks, and could be disciplined by them as well. Most remarkably, the state could manumit or even kill these apparently privately owned slaves. It is now necessary both to explain why these rules existed and then to determine whether or not such measures of communal intervention push helotage outside our classificatory scheme of private ownership.

The Rationale behind the Rules These rules, at first glance highly unusual, all make a good deal of sense when contextualized in terms of Spartan society more generally. Every last one of them is necessary and rational, and none can be passed off as mere eccentricities of the Spartan property system. It is vital to grasp the fact that helotage represented the most volatile element of Spartan society; de Ste. Croix was not exaggerating when he characterized the Spartan citizen body as perched atop a human volcano.²⁷ In a seminal article published in 1985, Paul Cartledge drew on the research of Eugene Genovese to study the factors that increased the likelihood of this human volcano erupting, and showed that of all Greek societies, Sparta was in particular danger. Genovese’s work focused on the New World, but, mutatis mutandis, Cartledge adapted the variables for a

²⁶ As Luraghi (2009): 261–2 rightly notes, this formula is almost the opposite to that of Pollux (3.83), who claimed that the helots lay ‘between free men and slaves’. Another fifth-century source of interest is a play entitled Helots (perhaps by Eupolis), to which Aristophanes (Eq. 1225) possibly refers: see schol. ad loc. with Sommerstein (1980): 51–2. Actors posing as helots must have constituted the play’s chorus: Storey (2003): 176 suggests that they represent the helot suppliants from the notorious episode mentioned at Thuc. 1.128.1. ²⁷ See de Ste. Croix (2002). By claiming that helotage was the most volatile element of the Spartan system, I do not mean to imply that all of Sparta’s institutions were geared to security against the helots. Several revisionist studies have tried to paint the relations between Spartans and helots as relatively free from tension: Roobaert (1977); Talbert (1989); Whitby (1994). It is tempting to draw the wrong conclusion—that the helots’ condition was relatively mild—from the fact that they were relatively free to run their own affairs and were not micromanaged by their owners. The arguments that follow will clarify why I think the approach of these scholars is mistaken.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Helotic Slavery in Classical Sparta

133

Greek context. He identified eight factors that were particularly important: (i) absenteeism, i.e. the physical absence of the slave owner in the day-to-day life of his slaves; (ii) the occurrence of economic distress or famine; (iii) slave units approaching 100–200 in size; (iv) splits in the ruling class; (v) slaves outnumbering the free; (vi) foreign-born slaves outnumbering those born into slavery; (vii) the social structure of the slaveholding regime permitting the emergence of slave leaders; and (viii) the geographical profile of the region facilitating the formation of runaway colonies. In classical Sparta, a confluence of most of these variables (especially nos. i, iii, v, vii, and viii) made the threat of slave revolt particularly acute. An everdwindling citizenry controlled a much larger helot population; and to make the disparity in numbers worse, the citizen lifestyle of the Spartiates revolved around day-to-day presence in the town of Sparta itself, making absenteeism a basic fact of life. The sheer distance of many estates from Sparta was also remarkable, since she controlled some 8,500 km² of territory, much larger than any other contemporary polis.²⁸ Some Spartiates must have owned over a hundred helots,²⁹ and the rugged terrain of parts of Messenia afforded potential strongholds for rebels. It is no wonder that a helot rebellion erupted after a major earthquake in 464 BC. It is vital to grasp the scale of this problem for the Spartans, and remember the inadequacy of their resources to tackle it by conventional means. The helots could not be managed like any other slave population in Greece, and the methods of control the Spartans employed are a monument to brutal creativity and an empirically mastered understanding of effective means of oppression. Lacking the resources to micromanage helot labour, the Spartiates resorted to what Cartledge aptly calls ‘state terror’.³⁰ Most of the practices noted above through which the state or citizenry could punish or kill the helots of individual owners can be explained in terms of this function. The chief weapon of this battery of measures was the krypteia, which has been alternately explained as either a rite de passage, a means of policing the helots, or a means of schooling Spartiates in the tactics of guerrilla warfare.³¹ However, the most credible

²⁸ Hodkinson (2008): 285. ²⁹ Hodkinson (2008): 315 n. 125. ³⁰ Cartledge (2011b): 57. ³¹ On the krypteia, see Ducat (2006): 281–331. For the krypteia as a rite de passage, see Jeanmaire (1939); Vidal Naquet (1986): 112–14; but Aristotle’s discussion makes it clear that the krypteia involved only the most intelligent of the young men, not all of them, and was sent out irregularly, not at fixed intervals, so the practice does not make sense as a general rite of passage for all Spartan youths (see Whitby 1994: 105–6). Ross (2011) sees the krypteia as a form of training in guerrilla warfare, in line with the emphasis in Pl. Leg. 633b9–c4. This view is not incompatible with the Aristotelian emphasis on terrorizing the helots, for surely the activities of the group acted as an apprenticeship in subterfuge and scouting. The two traditions are easily reconcilable so long as one does not assume that either Plato’s or Aristotle’s description tells the full story.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

134

Greek Slave Systems

function of the institution is as an agent for spreading paranoia among the helots.³² The effect of such an institution is more palpable if we consider a passage from the Life of Frederick Douglass. Here, Douglass recounts the methods of a supervisor named Mr Covey, under whose authority he fell in 1833: There was no deceiving him. His work went on in his absence almost as well as in his presence; and he had the faculty of making us feel that he was ever present with us. This he did by surprising us. He seldom approached the spot where we were at work openly, if he could do it secretly. He always aimed at taking us by surprise. Such was his cunning, that we used to call him, among ourselves, “the snake”. When we were at work in the cornfield, he would sometimes crawl on his hands and knees to avoid detection, and all at once he would rise nearly in our midst, and scream out, “Ha, ha! Come, come! Dash on, dash on!” This being his mode of attack, it was never safe to stop a single moment. His comings were like a thief in the night. He appeared to us as being ever at hand. He was under every tree, behind every stump, in every bush, and at every window, on the plantation.³³

We need not suppose that the krypteia operated exactly in this fashion, but this passage conveys very strongly the psychological effect that such methods of supervision could have on the slaves thus observed. We might call the krypteia—to borrow an engineering term—a ‘force multiplier’. In other words, it was a way of achieving maximum effects from a minimal input.³⁴ Such a system would not require micromanagement of helot labour, which in any case the Spartan citizenry had neither the time nor the resources to implement; rather, the belief that at any time one might be under observation—combined with occasional and sensational acts of extreme brutality to show that the Spartans meant business—could instil the kind of paranoia that Douglass felt when under the eye (or not) of Covey.³⁵ The high level of ingrained paranoia among many helots is clear from a story related by Plutarch that is set following the Theban liberation of Messenia from Spartan control in 369 BC. According

³² Aristotle is quoted in Plutarch’s Life of Lycurgus 28.1–3. Nobody knows the date of the krypteia’s introduction. For a variety of views: Flower (2002): 206–7; Azoulay (2006); Nafissi (2015). ³³ Gates (2002): 385, my italics. ³⁴ Much, then, as modern terrorist groups create a climate of paranoia vastly out of proportion to the threat they pose to the average citizen going about his or her business. For the psychology behind the effectiveness of such tactics, see Kahnemann (2011): 144; 322–3. ³⁵ This interpretation, I think, makes the best sense of Aristotle’s description of the krypteia in Plu. Lyc. 28.1–5. Ducat (2006): 287–8 finds the two methods (in 28.4 and 28.5) of dispatching helots to be contradictory, but the underlying logic works and is eminently practical: the krypteia would kill any helots abroad at night, and would murder in the fields (obviously by day; the reference to ‘night’ only refers to the method described in Plu. Lyc. 28.4) those they had observed as overly large and powerful. As for the ἄλλως in 28.3, this surely indicates that they were sent out with no specific target, that is, no named individual, but could exercise their discretion in the selection of victims.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Helotic Slavery in Classical Sparta

135

to the account, the Theban troops asked the liberated helots to sing to them the songs of Alcman, Terpander, and Spendon, but they refused, claiming that their masters forbade them to do so (Plu. Lyc. 28.5). This anecdote describes precisely the effects one would expect on the psyche of many of the individuals subjected to the unconventional methods of brutality and bullying practised in Sparta.³⁶ The most famous act of state brutality of this ilk is the alleged massacre of 2,000 helots narrated by Thucydides (4.80), who says that a proclamation was made to the helots inviting those who thought they had served the state well in war to come forward and receive their freedom. Those who answered the summons processed around the temples, but then promptly disappeared; nobody knew how each individual died. Several scholars have advanced reasons to reject the historicity of this episode. Paradiso believes that the elimination of so many helots would have been logistically impossible, but Harvey has shown that we should credit the Spartans with sufficient imagination and expertise to carry out such an enterprise.³⁷ Whitby, on the other hand, has made a far more compelling objection: that such an action would deter helot enlistment for military service abroad, effectively harming rather than aiding the Spartan war effort. Since Thucydides, after discussing the massacre, goes on to say that 700 helots went north with Brasidas to serve in his campaign in 424 BC, the massacre itself makes no sense and should be seen as apocryphal.³⁸ Critics of the episode’s historicity have overlooked one key point: the ideological dimension of the unprecedented act of the Spartans asking the helot veterans if they felt entitled to be free. Cartledge has rightly written that state manumission of helots ‘was practised by the Spartans with considerable managerial art’.³⁹ The Spartans did on occasion explicitly offer freedom at the outset to helots if they performed a specific service (Thuc. 4.26; Xen. Hell. 6.5.28–9); and Andrewes has argued that one group of manumitted helots, the neodamodeis, were manumitted at the point of recruitment.⁴⁰ But that was not the case with the helots who followed Brasidas northwards: for them, freedom was to be granted after services were rendered, and the impetus came from above, not from demands from below. Thucydides tells us just this: that after three years of service the helots who had served with Brasidas were freed by a decree of the Spartans (Thuc. 5.34). What Thucydides describes in the context of the massacre should be seen as something completely different and utterly unprecedented. Instead of identifying those helots who had merited manumission and passing a decree to make them free, the Spartans asked the helots to identify themselves ³⁶ ³⁸ ³⁹ ⁴⁰

Cf. Hornblower (2000). ³⁷ Paradiso (2004): 186; Harvey (2004): 204–5. Whitby (1994): 98–9. Cartledge (2003): 18. On helot manumission, see Paradiso (2008). Gomme, Andrewes, and Dover (1970): 35–6.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

136

Greek Slave Systems

if they thought they merited freedom. Ideologically speaking, this ran contrary to the whole ethos of Spartan domination of the helot class; and indeed Thucydides himself points out that the Spartan declaration was a test (πεῖρα). The message was clear: the Spartans decided when helot veterans should be freed, not the helot veterans. By rooting out those helots who felt entitled to freedom, the state killed two birds with one stone: they removed those of the helots who were potential rebels or rebel leaders (and possessed some military experience⁴¹); and they sent a clear message to the helots in general, that the Spartans would crush any resistance without hesitation.⁴² This action did not send mixed messages to any helots contemplating volunteering in the army, but to those who might presume that military service entitled them to be free or that they should have any say in the matter. This was a shock-and-awe tactic enacted during a period in which Sparta experimented with the recruitment of significant numbers of helot soldiers, and needed to make it perfectly clear who was in charge and who made the decisions about rewards for service. Though the figure of 2,000 victims—like almost any ancient ‘statistic’—may be open to question, the arguments for believing in the historicity of the episode outweigh those (in my opinion) that see it as a kind of bogeyman story concocted by the Spartans for propagandistic reasons. The message that all of these measures sent to the helots was unambiguous: if they kept their heads down, worked hard, and sent the required portions of produce to their owners in Sparta, no harm would come to them, and they could go about their lives relatively free from interference. But if they fomented sedition, went abroad at night (presumably to communicate with other helots, whether for seditious purposes or not), or rose beyond their humble station in life, they could expect a brutal response. Ducat has claimed that the selection of helot victims for state assassination was random,⁴³ but this does not tally with the evidence of Thucydides, Aristotle, and Myron, all of whom stress the consistent, targeted nature of state violence and its focus on rooting out potential rebels or exceptionally strong individuals. In other words, state violence towards the helots was not random at all, but specifically punished certain kinds of behaviour. Not only are such measures credible: they are the only sort of measure that a state with limited manpower like Sparta could resort to in order to keep in line a large and widely distributed slave population.

⁴¹ It comes as no surprise that Critias noted a practice of removing the porpax from hoplite shields to render them useless to opportunistic helots (88B F 37 D-K), and Xenophon noted the care taken over keeping helots away from weapon dumps (Lac. Pol. 12.4). For removing the porpax from a shield to render it useless, see Ar. Eq. 847–57. ⁴² One should note that in 425 the events at Pylos and Sphakteria were arousing helot resistance in Messenia (Thuc. 4.41). The massacre was, from the perspective of the Spartans, a timely and appropriate response. ⁴³ Ducat (2006): 306.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Helotic Slavery in Classical Sparta

137

Other measures, such as the right to borrow helots, hounds, and other items must be understood not in terms of security measures, but rather as a reflection of the culture of communalism and sharing that prevailed in Sparta.⁴⁴ Even the fact that Spartiates could beat the helots of other owners appears rather less singular once it is contextualized in terms of Sparta’s broader culture of corporal punishment: Spartiates could also corporally punish the children of other fathers (Xen. Lac. Pol. 2.10; 6.1–2), break up fights between groups of youths (Xen. Lac. Pol. 4.6), violently penalize breaches in sobriety amongst one another (Pl. Leg. 1. 637a; cf. Dion. Hal. Ant. Rom. 20.13.2), and strike the tresantes—a group comprising shamed Spartiates who had displayed cowardice in the field—if they caught them comporting themselves in a way that did not befit their disgraced status.⁴⁵ Private property may have been ubiquitous in Athens and Sparta, but very different cultural attitudes towards its use prevailed in the two regions.⁴⁶ Finally, we must account for the ban on the external sale of helots and the complete ban on private manumission. These should be understood in the context of Sparta’s economy and levels of foreign trade. Unlike large commercial poleis such as Athens, Corinth, or many of the other Aegean polities, Sparta practised a comparatively isolated form of economic life that was cut off from foreign trade to a high degree, a fact that conservative philosophers in Athens admired and attempted to build into the economies of their ideal states (e.g. Pl. Leg. 4.704–5). We are thus faced with an unusual fact: whilst Sparta was more heavily reliant on slave labour than any other Greek society (see section II below), she could not have imported many slaves to ‘top up’ the helot population, which instead relied more or less entirely upon natural reproduction.⁴⁷ It was thus vital to plug any gaps that might represent a ⁴⁴ Hodkinson (2000): 187–208. Cf. the Spartan attitude to theft of foodstuffs by youths: Xen. Lac.Pol. 2.6–9; Anab. 4.6.14–15; Isoc. Panathenaikos 211–12, with Hodkinson (2000): 201–5— clearly a mitigation of the normal rules on respecting private property; cf. Luraghi (2002): 238. On this culture of communalism, see Christesen (2004). ⁴⁵ On the tresantes, see Ducat (2006). On Sparta’s culture of physical violence, see Hornblower (2000), who illustrates its distinctive character, very much running against broader Greek norms. ⁴⁶ For an Athenian view on the impropriety of breaking up street fights between strangers, see Theophr. Char. 13.5. One might also contrast the rather different attitudes to entering private land in Athens and Sparta. As we noted, Spartan rules gave citizens the right to enter private land and help themselves to food or to borrow the horses and slaves of other citizens if the occasion called for it. In Athens, though, a more cautious attitude to private property is reflected in the exhortation of Xenophon (Cyn. 5.34) not to interfere with crops or watercourses on private land if hunters had to enter it in pursuit of their quarry. In fact, a law of Solon clarified the rights of citizens with no access to water on their land to use the wells of their neighbours in a very specific fashion: see Plu. Sol. 23.6 = fr. 63 [Leão and Rhodes]. ⁴⁷ In many ways, then, like the slaves of the nineteenth-century US South: see Fogel and Engerman (1974): 126–44. The fact that helotage relied upon natural reproduction goes a long way towards explaining why the helots appear to have dwelt in family groups. As Harper (2008): 113 rightly notes, ‘the reproductive rates seen in North American slavery were the product of a

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

138

Greek Slave Systems

potential drain on servile manpower. Since the provision of an adequate slave labour force was so crucial to the whole edifice of Spartan society, the state took unusually intrusive measures to curb the potential of slaveholders to alienate their helots and thus shrink the slave workforce.⁴⁸ Similar measures were undertaken in at least two other slave systems that were compared to helotage by Plato and Aristotle: the Penestai of Thessaly (Archemachus FGrHist 424 F1) and the Mariandynoi of Heraclea Pontica (Poseidonios FGrHist 87 F8). Hodkinson⁴⁹ believes that the explicit parallel on an external sale found in Poseidonius is unreliable, since it is associated with an obviously apocryphal ‘contract of servitude’. It makes more sense, however, to view these ‘contract’ stories as charter myths that seek to account for the origins of genuine historical bans on sale. In other words, they are aetiological inventions: the rationale that lay behind the original enactment of these bans was forgotten, and Archemachus and Poseidonius (or their sources) then advanced more fanciful explanations couched in the events of a mythical past to explain their existence. These two cases thus parallel the Spartan rule and were perhaps enacted for similar reasons. It is worth noting that this form of state intervention regarding the export of certain resources was hardly unique in the Greek world. In Attica, Solon enacted a law that prohibited the export of agricultural products other than olive oil beyond the Attic frontiers, which guaranteed sufficient foodstuffs in circulation to feed the population (Plu. Sol. 24.1 = fr. 65 [Leão and Rhodes]). The extent to which this was followed may be debated, but there was certainly a rigidly enforced ban on the export of grain (Dem. 34.37; 35.50–1; 58.8–9). This, like the Spartan rule, was a ban on exporting certain items of property ἔξω τῶν ὅρων, to use Ephorus’ phrase. One could view the Athenian impositions on Keos regarding the export of ruddle to areas outside Attica in a similar light (IG II² 1128), or the ban on exporting certain items of naval equipment (Ar. Ran. 364); these were simple, hard-nosed measures taken by rather simple recipe. Space, physical independence, privacy, and opportunity, rather than deliberate breeding policies, were behind the fertility of the American slave population.’ For a superb comparative demography of slavery, see Higman (2011). ⁴⁸ As Ducat (1990): 23 notes: Quant aux interdictions de la vente à l’extérieur et de l’affranchisement privé, elles correspondent moins à d’authentiques pratiques communautaires qu’au souci de maintenir dans son intégrité le cheptel servile: comme il n’était alimenté que par sa propre reproduction et qu’à la différence de ce qui se passait pour les esclaves-marchandises on ne pouvait importer des Hilotes de remplacement, il était essential d’éviter toute déperdition. Van Wees (2003): 70 and Luraghi (2002): 229 adopt a similar view of the ban. Cf. Zurbach (2013): 973, who sees the ban as an archaic feature analogous to the restrictions placed on alienating or mortgaging certain amounts of landed property in other poleis. Another possibility is that the ban was designed in part to keep Spartiates at arm’s length from commerce: Luraghi (2002): 234. ⁴⁹ Hodkinson (2000): 118–19.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Helotic Slavery in Classical Sparta

139

the state to ensure an adequate supply of necessities.⁵⁰ Similar practices occur in many historical societies: centuries earlier, the Assyrians banned the free export of items such as cedar wood and metals, guaranteeing themselves a monopoly of supply.⁵¹ Just a few years ago (August 2010, to be exact), the Russian government temporarily banned the export of grain due to poor harvests. We do even possess a parallel ban on the external sale of slaves from Ptolemaic Egypt (C.Ptol.Sklav. 1.13–14), though the reasons behind this ban are obscure.⁵² Thus, extraordinary circumstances called for extraordinary measures. The ostensibly unusual legal position of the helots was, rather, a tailoring of the standard legal contours of slave ownership to fit the specific local conditions that prevailed in Sparta. Helotage was calibrated to work in tune with Sparta’s broader institutional framework.

Helotage and Slavery: Emic and Etic Views Our Greek texts use the same vocabulary of slavery for the helots as they do for other Greek slave systems.⁵³ When Thucydides (8.40) commented on the large numbers of oiketai at Chios and Sparta, he evidently did not think he was comparing apples and oranges. In the following century, when Theopompus (FGrHist 115 F 122) wrote that ‘The Chians were the first Greeks after the Thessalians and Lacedaemonians to have used douloi,’ the contrast he was striking was not between slaves at Chios and serfs or some kind of peasant dependency at Sparta and Thessaly, but between a system that drew on trade for its slaves and others that did not. For the Greeks of the classical period, the helots were without question slaves. They satisfy an emic understanding of douleia. But does this align with the etic framework of private ownership set out in Chapter 1 of this book? Some historians have doubted it and put the Greek usage of slave terminology for the helots down to juridical naivety.⁵⁴ Potentially problematic are, first, the inability of Spartan citizens to manumit their helots and, second, the broad range of measures that allowed the state as ⁵⁰ Cf. Bresson (2000): 129 n. 80. On the decree concerning Kean ruddle, see now Lytle (2013). The comic poet Alexis drew on the ubiquity of such practices to joke that Lesbos should be given tax exemptions for exporting her wine to Attica, but anyone who re-exported it should have their property confiscated (fr. 278 K-A). ⁵¹ Deszö and Vér (2013): 354–9. ⁵² See Bieżuńska-Małowist (1974): 56. It is possible the Assyrians also banned the external sale of slaves: see SAA V 150 with Deszö and Vér (2013): 357–8. ⁵³ See Ducat (1990): 45–8 for a list of references. ⁵⁴ Cartledge (1979): 139; (2011a): 78–9; Harvey (1988): 48. It should be noted, though, that this accusation of juridical naivety is based on the view that the helots were not slaves, and therefore that the Greeks were clumsy in labelling them as such. The criticism unravels, though, if we accept that the helots were indeed slaves.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

140

Greek Slave Systems

well as average citizens to interfere with the helots of individual owners. The first problem relates to the principle of Right to capital, the second relates to the principle of Right to security. Let us begin with the rules on sale and manumission. Recently, Cartledge has objected to revisionist work on helot status on the grounds that the right to manumit helots rested with the state, claiming that this is in formal conflict with the notion of private ownership.⁵⁵ But is this really so? As we have seen, owners of property must be able to alienate that property; yet there is nothing in this principle that implies that the forms of alienation might not be restricted. Spartans could sell their helots, albeit in a limited area; and a fortiori they must have been able to transfer their helots via gift as well. They did have a Right to capital: a heavily restricted one, to be sure; but as we have seen, these restrictions existed for good reasons. As for the right of the state to manumit privately owned helots, Ducat pointed out almost thirty years ago that the Athenian state did much the same thing in special circumstances, such as offering freedom to slaves who informed against their owners if the latter had committed sacrilege (Lys. 5.3–5; 7.16; And. 1.28), or in the case of emergencies during wartime, when manpower was in urgent demand.⁵⁶ To Ducat’s examples we could add Diod. 20.100.1–4, which describes how the polis of Rhodes freed slaves during a siege, and [Arist.] Oec. 2.1350a in which the polis of Olynthus forced the sale of privately owned slaves during wartime to raise money.⁵⁷ Honoré has rightly written: it is, perhaps, a characteristic of ownership that the owner’s claims are ultimately postponed to the claims of the public authority, even if only indirectly, in that the thing owned may, within defined limits, be taken from the owner in order to pay the expenses of running the state or to provide it with essential facilities.⁵⁸

A glance at the Aristotelian Oeconomica provides many examples of this principle in action across the classical Greek world. Even in the most wellknown slave society in history, that of the United States, a number of states enacted blanket bans on private manumission, reserving the right to manumit for the state alone. This fact has not led anyone to suppose that the slaves of the US South should be recategorized as ‘state serfs’.⁵⁹ Pace Cartledge, the

⁵⁵ Cartledge (2003): 17–18. Cartledge (2011a: 82) claims that central control over manumission ‘is in formal conflict with the possibility of private alienation of Helots’. Cartledge (2003): 19 (cf. 2011: 90) claims that the private ownership argument implies that there were helot markets in Sparta and that helots were frequently sold, but this misrepresents the arguments of Ducat and Luraghi: cf. the statements of Luraghi (2002): 233 and Ducat (1990): 21–2. ⁵⁶ Ducat (1990): 26–7. ⁵⁷ See also Kamen (2005): 15–22. ⁵⁸ Honoré (1961): 124. Cf. Birks (1985): 11–14. ⁵⁹ See Klebaner (1955); Kolchin (1993): 89–90. A law enacted in Virginia in 1723 prohibited manumissions of ‘Negro, mullatto or Indian slaves . . . except for some meritorious service, to be adjudged and allowed by the governor and council’.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Helotic Slavery in Classical Sparta

141

Spartan rules on manumission are not in formal conflict with the notion of private ownership. The issue of Right to security should be seen in a similar light. The fact that ownership is normally fenced around with rules protecting the owner’s enjoyment of his property never means that this enjoyment is absolute or exclusive. For example, in classical Athens, the state was able to punish privately owned slaves with whippings in special circumstances,⁶⁰ and under exceptional conditions could manumit slaves without the consent of their owners. Athenian citizens could also punish the slaves of other owners if they trespassed on their property in order to steal their goods.⁶¹ In Sparta the same principle obtained, but to a far greater degree, making truly intrusive inroads into the rights of owners. In Teos towards the end of the fourth century, we find that the state could commandeer privately owned oxen for public works (SEG 2.79, lines 5–6); in Sparta, the same principle was extended to a larger number of items of private property (including helots) and with a broader definition of the public interest (including the immediate needs of fellow citizens). These exceptional measures were, again, necessary and are not simply an eccentricity of Sparta’s property system. We might better understand this case if we consider some extreme forms of property restriction in modern British society. For example—and at the liberal end of the scale—the owner of a garden shed is free to do as he or she wishes (within reason) to that shed, even to reduce it to kindling. Yet an owner of a house faces multiple restrictions if he or she wishes to alter its appearance, and must apply for council permission with plans of proposed alterations. At the extreme end of the scale, an owner of a listed building—particularly those of a high listing grade—has almost no freedom at all to make alterations to his or her home. Honoré is surely correct to write that ownership is the greatest legal interest permissible under a mature system of law; but such individual interests are often trumped by the interests of the state and the broader interests of society at large. Recall our earlier discussion of Parisian property owners during Baron Haussmann’s remodelling of Paris, or property owners whose dwellings lie in the way of the proposed high-speed rail link between London and Birmingham. These individuals had and have no way of stopping topdown policies from riding roughshod over their private interests. With such thoughts in mind, Sparta’s rules regarding helot ownership amount to merely another example of the triumph of collective concerns over individual rights. Helotage still fits in our etic framework of private ownership, but it is a form of private slave ownership more heavily subject to public restrictions than any other that is known to me.

⁶⁰ See Chapter 1, p. 45 n. 69.

⁶¹ See Chapter 1, p. 45 with n. 67.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

142

Greek Slave Systems

II. HELOT ECONOMY AND S LAVE SOCIETY AT S PARTA Here is not the place to reconstruct the full workings of the helot economy in detail: Hodkinson’s fine recent comparative study does just that, and readers who are interested in how helotage worked ‘on the ground’ should consult his essay.⁶² Instead, this section will analyse the degree of dependence of the Spartan citizen body upon helot labour, as well as sketching the disparity in holdings between those Spartiates merely hanging onto their citizen status by a whisker, and those belonging to the most wealthy and powerful families in the state. The main fact to emphasize is that, unlike Athens, citizenship at Sparta was predicated upon wealth. All male citizens belonged to a mess group (phidition—called a syssition by non-Spartan writers: Arist. Pol. 1272a3; Plu. Lyc. 12.1), and continued membership of the citizen body depended on that Spartiate’s ability to contribute—monthly and without fail—set rations to his mess (Arist. Pol. 1271a26–37). This is not a controversial issue, nor is the fact that the agricultural products contributed to the mess group were the result of helot labour (Spartiates did not generally partake in production).⁶³ We can, at any rate, sketch a chain linking the production and movement of these key resources running from the fields of Laconia and Messenia to the messes of the Spartiates in Sparta itself. At the first stage is primary production, most of it agricultural. This must have been managed in some way, though our sources are not clear on how; but what is clear is that the helots kept some of this produce for their own consumption,⁶⁴ and the remainder was sent to Sparta to their owners’ households (by what means and routes remains unclear). It was from there that the owner could furnish his mess dues and participate in the life of the citizen community. Fundamentally, then, citizen status rested upon the ownership of enough land and helots to produce the regular mess contributions. Spartan citizen society was a society of slave owners. But as was the case with the ownership of land and other goods, there was a degree of disparity in Spartan society between those citizens with just enough to maintain their contributions and those with much larger holdings. Spartan society was far less egalitarian than the ‘mirage’ created by later writers would imply. Hodkinson has estimated that the richest Spartiates

⁶² See Hodkinson (2008). On Hodkinson’s argument that the term μονονόμοι at Hescychius μ 1626 should be emended to μνῳονόμοι (as suggested by Wilamowitz), i.e. ‘helot overseers’, see now Schmitz (2014). Even if Schmitz’s argument against Hodkinson is correct, it is surely the case that some helots acted as overseers, whatever they were called. ⁶³ See Hodkinson (2000): 177–9. If Xen. Lac.Pol. 7.1–2, Oec. 4.3, and Hdt. 2.167 do not amount to a rigid ban on citizens practising crafts or chrematismos, they at least attest a strong social disdain for such activities. ⁶⁴ Hodkinson (1992a); cf. Luraghi (2002): 229–30.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Helotic Slavery in Classical Sparta

143

may have owned well over a hundred helots, probably more on average than their Athenian counterparts.⁶⁵ Let us put this in perspective. Athens, as we shall see in Chapter 8, was a slave society in the Finleyan sense: its elite derived a substantial proportion (but by no means all) of its wealth from slave labour. Sparta’s upper class was also a slave-owning elite that owed its wealth and position to slave labour, but to a much greater degree: all wealthy Spartans derived practically the entirety of their wealth from helot labour. If we look further down the wealth spectrum in either society, we see a fundamental difference between Athens and Sparta: many Athenian citizens had no slaves at all, and their citizen status depended on nothing more than having two Athenian parents (at least, after 451 BC). In Sparta, all sub-elite citizens were reliant on helot labour, and their continued citizen status rested on the ownership and exploitation of helots. Sparta, in other words, represents the most extreme example of slave society anywhere in the ancient world, perhaps even world history. The only reason why this has not been fully realized is the mistaken belief of many scholars that the helots were something other than slaves.

I II . W H A T I S H E LO T IC S L A V E R Y ? Classical-era writers working in Athens were well aware of the unusual character of Spartan slavery, at least when compared to the form of slavery with which they were familiar. They noticed other slave systems in the Greek world that somehow resembled helotage, and drew some rudimentary comparisons between them. The groups they compared to helotage were the penestai of Thessaly, the variously labelled dependent populations of Crete, the mariandynoi of the Black Sea colony of Heraclea, and the kallikyrioi of Syracuse (Pl. Leg. 776c-d; Arist. Pol. 1264a32–6; 1269a37–b5; 1271b40–72a2; Arist. fr. 586 ([Rose]). We may group these systems provisionally under the rubric ‘helotic slavery’. Yet it is crucial to be clear and explicit about what sort of category this label denotes, and the precise grounds for comparison between helotage and these other systems. To understand this, it is perhaps best to begin with the approach shared by most twentieth-century scholars, one that is fundamentally problematic, indeed methodologically back to front.⁶⁶ It begins with the statement of Pollux (3.83)—writing in the second century AD—that ‘between free men and slaves are the helots of the Lacedaemonians, the penestai of the Thessalians, the klarotai and mnoitai of the Cretans, ⁶⁵ Hodkinson (2008): 315 note 125. For the upper limits of slave ownership at Athens, see Chapter 8, section I, pp. 171–2. ⁶⁶ For recent examples of this approach, see van Wees (2003); Cartledge (2011a): 78–82.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

144

Greek Slave Systems

the Mariandynians’ dorophoroi, the gymnetes of the Argives and the korynephoroi of the Sicyonians’.⁶⁷ Evidently, the factor that Pollux thought united these systems was one of status: they were neither slaves nor free men, but something in between. Latching on to this issue of status, much twentiethcentury scholarship focused on putative differences in legal status between helotic groups and the form of so-called ‘chattel slavery’ evident in Athens and elsewhere. Scholars such as Lotze and Finley believed that these legal-status differences included communal rather than private ownership, as well as special legal rights enjoyed by the groups in question; and they ransacked earlier sources to confirm details of the picture of ‘unfree’, but not-quite-slave status they had derived from Pollux. There are several problems with this approach. First, it begins from a very late source and works backwards— precisely the opposite direction that scholars should take when analysing evidence; second, it presumes rather than demonstrates the accuracy of Pollux’s claim; and third, it presumes that Pollux’s criterion for comparing these systems was identical to that of Plato and Aristotle. Yet if one proceeds from the early sources to the late ones without the blinkers imposed by faith in Pollux’s statement, a very different picture emerges. There is no sign in the early sources that any writer considered the helots or analogous groups to share some mysterious status ‘between free men and slaves’. Plato, for example, has only one thing in mind when he compares the helots to groups such as the Mariandynoi or Penestai: they all speak the same language, and are thus quite distinct from the style of slavery familiar from Attica, where the slave population comprised a hotchpotch of barbarians imported from all over the periphery of the Aegean world. Aristotle, too, compares the helots to other slave systems, and his criterion is precisely the same as Plato: language, not status (Arist. Pol. 1330a25–33; cf. Pl. Leg. 777b–d). ⁶⁷ ‘Μεταξὺ δὲ ἐλευθέρων καὶ δούλων οἱ Λακεδαιμονίων εἵλωτες καὶ Θετταλῶν πενέσται καὶ Κρητῶν κλαρῶται καὶ μνῷται καὶ Μαριανδυνῶν δωροφόροι καὶ Ἀργείων γυμνῆτες καὶ Σικυωνίων κορυνηφόροι.’ On the phrase metaxu eleutheron kai doulon, see Ducat (1990): 45–51. Finley (1980): 116 claimed in relation to Pollux: it is of 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. Here, Finley represents Swoboda’s suggestion that Aristophanes might lie behind Pollux’s claim (Swoboda 1905: 252 n. 1, followed by Lotze 1959: 1) as if it were a fact, lending authority and greater antiquity to the slogan; many scholars have subsequently followed Finley’s ipse dixit. But there is no good evidence that the slogan is any earlier than Pollux, and the link to Aristophanes is pure speculation. Indeed, Athenaeus (6.84.263d–e) cites Aristophanes’ student Callistratus’ claim that the terms helots, penestai, klarotai, and dorophoroi were euphemistic names given to these populations to take the sting out of the fact that they were slaves (oiketai). This only increases the unlikelihood that Aristophanes had anything to do with the metaxu slogan. To their credit, Westermann (1945) and Hunt (2016) cite Pollux without invoking Aristophanes. I treat this issue at greater length in Lewis (forthcoming b).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Helotic Slavery in Classical Sparta

145

Aristotle’s contemporary Theopompus of Chios also compared the helots to the Penestai of Thessaly (FGrHist 115 F122). His criterion is subtly different from that of Plato and Aristotle, for he believed that the helots and Penestai were descended from indigenous groups enslaved en masse in long-ago wars by their masters upon entering the territory where they would settle. This he contrasted with the commercial origins of other slave populations, which were supplied with slaves through trade rather than conquest, the earliest practitioners of which he believed to be his own people, the Chians.⁶⁸ Note, however, that, as with Plato and Aristotle, there is no hint that Theopompus considered the helots or Penestai to be anything other than slaves, and therefore directly comparable with the slave population of Chios (cf. Thuc. 8.40.2). His discussion, however, does shift the grounds of comparison from one of language to one of origins; indeed, theories about the helots’ origins go back to the fifth century.⁶⁹ These two factors—language and origins—are closely linked and represent a field of inquiry that we can also find in Ephorus’ writings. Whereas Plato and Aristotle merely noted that helots and analogous groups all shared a common tongue, Ephorus and Theopompus, proceeding from this observation, asked the following question: why do these groups share a common tongue, and are thus unlike the slaves of regions such as Athens or Chios? The answer, for them, lies in the origins of these systems: the helots speak a single tongue because they are a single people whose ancestors were enslaved en masse long ago. This aetiological activity begins innocently enough, with speculation and etymological hypotheses, but ends up with charter myths accounting for the origins of these groups in terms of mass enslavement in war in the distant past, often accompanied by a ‘contract’ between the enslaved population and their masters that secured them certain privileges not enjoyed by slaves in more conventional systems. The work of Nino Luraghi has shown that these late accounts cannot be taken seriously as hard evidence of the origins of helotage. And as these systems gradually became extinct, and aetiologies of original ‘contracts’ of servitude more common, it is no wonder that a late scholar like Pollux could speculatively label them as ‘between free

⁶⁸ Scholars often accept Theopompus’ statement as a fact about early Greek slavery and ignore its status as a piece of ‘intentional history’ (e.g. Cartledge 2011a: 80; Braund 2011: 121; Kyrtatas 2011: 92–3). But the buying of slaves from barbarians is already found in Homer (Il. 6.289–92; Od. 15.417; 24.389); besides, would anyone today take seriously the claim of Critias (88B F 2 D-K) that the Athenians invented pottery or that the Carians invented the merchant ship, or Herodotus’ claim (6.112) that the Athenians at Marathon were the first Greeks to face men in Persian dress in battle? Clearly, we should view Theopompus’ statement with the same scepticism as we would apply to these statements. For Greek tendencies of attribution to a single protos heuretes, see Kleingunther (1933). ⁶⁹ i.e. Antiochus of Syracuse FGrHist 555 F13. See Luraghi (2003): 117–35.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

146

Greek Slave Systems

men and slaves’ and see in them a real substantive difference in status from the slave systems of his own day.⁷⁰ The implications of these facts for our understanding of helotic slavery are significant. Above all, we should privilege the more abundant classical evidence for the helots being slaves ahead of the dubious claim of a late writer such as Pollux. By following Pollux uncritically, many scholars of the previous century removed helotage and comparable systems from the category of slavery sensu stricto, sometimes even excising it from whole books on ancient slavery.⁷¹ If we divest ourselves of this misconception and look at the early sources afresh, however, we can see that there is no reason to suppose that helots and analogous groups were anything other than slaves. Helotage was certainly different in many ways from the form of slavery found in Attica; not least, its linguistic uniformity implies major differences in structural organization and strategies of slave supply. This is only problematic to the historian who thinks of Athenian slavery as somehow characteristic of Greek slavery in general. But an approach that is informed by comparative study of slavery should be more aware of the myriad forms that slavery can take in different periods and places, and should not be dogmatically tethered to familiar paradigms such as Athens, Rome, or the US South, viewing everything else as odd or aberrant. For Sparta, though, we only have a handful of sources that tell us about the details of helot status. The ‘serfdom’ or ‘between free men and slaves’ argument can be tested much more rigorously against a richer body of evidence when we turn to classical Crete, the subject of Chapter 7.

⁷⁰ Luraghi (2002; 2003; 2009). ⁷¹ e.g. Schumacher (2001); cf. Andreau and Descat (2006): 11–13, who refer to the helots as ‘dépendants ou serfs’ and ignore helotage (and Cretan slavery) entirely in their subsequent discussion.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

7 Classical Crete Crete constitutes the best testing ground for the issue of whether the economies of various regions of the Greek world were dominated by ‘serf ’ populations or whether these populations instead constituted privately owned slaves whose labour was simply sourced and organized in a different manner from the Attic slave population. The reason why Crete makes such a good test case lies in the relative riches of its epigraphy, far outstripping that of any other region where ‘helotic’ slavery is held to have existed, and providing a good deal of fine-grained data on specific legal issues. Despite the relative abundance of evidence for slavery in Crete, the island rarely features in general works on Greek slavery: some works devote a page or two to Cretan slavery, but elsewhere it has simply been written out of the textbooks.¹ Meanwhile, assumptions that Crete was peopled with ‘serf ’ populations continue to abound.² This chapter treats the issue at length, because Crete belongs at the heart of any detailed discussion of slavery in the classical Greek world that aims at going beyond merely extrapolating from the Athenian situation; and a study of Crete’s inscriptions allows us to peer into the legal inner workings of one specific ‘helotic’ slave system, Gortyn. We shall begin with the literary sources that discuss Cretan slavery; these provide us with some useful generalizations about the subject and serve to set the scene, but they must be used with caution. From ¹ For example, Fisher (1993): 32–3 and Bradley and Cartledge (2011): 146–7 both provide a paragraph on Crete, whilst Garlan (1988): 99–101 provides a page and a half. But Crete is entirely left out of the introductory textbooks of Schumacher (2001) and Andreau and Descat (2006). Gortynian slavery has been treated in recent years by Link (1994): 30–48 and (2001), but his work has not received the attention it deserves. ² For example, Zelnick-Abramovitz (2012): 106 translates Ephorus FGrHist 70 F 29 (ὁ ῎Εφορος δ᾽ ἐν γ´ ῾Ιστοριῶν «κλαρώτας» φησί «Κρῆτες καλοῦσι τοὺς δούλους ἀπὸ τοῦ γενομένου περὶ αὐτῶν κλήρου») thus: ‘Ephorus in the third book of his Histories says that the Cretans call their slaves (τοὺς δούλους) klarōtai from the lots (klaroi) to which they were attached.’ But this translation presupposes a system of serfdom. A less forced translation would be ‘Ephorus in the third book of his Histories says that the Cretans called their slaves klarōtai after the lots cast for them.’ Cf. also the context in which this fragment of Ephorus is discussed in Athen. 6.263f, especially the preceding sentence, which makes it clear that we are dealing with captives being allocated to private owners by lot, a traditional practice (on which, see van Wees 1992: 299–310; cf. Ferone 1997: 145–8).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

148

Greek Slave Systems

there, we zero in on the polis of Gortyn, whose famous law code—analysed in conjunction with some lesser-known contemporary legal inscriptions—allow us to determine whether or not the trend we have observed at Sparta and elsewhere (i.e. private ownership of slaves and the standard incidents of private ownership) holds for Crete as well.

I . T H E L I T ERA R Y SO U R C ES Aristotle’s Politics contains several interesting observations on Cretan slavery as it functioned in his own day. He notes that the Spartan and Cretan forms of organization (taxis) resemble one another and that the Cretan form of slavery resembled helotage.³ As we have noted, this comparison implies only a loose association at best, and beyond the factor of linguistic uniformity, we should expect diversity. Besides, Aristotle was not blind to the differences between helotage and Cretan slavery. These differences fall in several areas; one is the lack of rebellions in Crete compared to the frequency of revolts among the helots and penestai: It is a fact that everyone agrees upon that if a polis is going to be well governed there must exist leisure from necessity; but it is not easy to ascertain the manner in which this is to be made manifest. For the Thessalian penestai often attacked the Thessalians, and the helots did the same against the Lakonians (for they are like enemies continually lying in wait for their misfortunes). Among the Cretans nothing of this sort has come to pass. The reason for this is probably that the neighbouring poleis, even when at war with each other, never ally with the deserters since it is not profitable as they themselves have their own perioikic population, whereas the neighbours of the Lakonians are all enemies: Argives, Messenians, and Arcadians. With the Thessalians, the penestai were originally rebelling because the Thessalians were still fighting with their neighbours: Achaians, Perrhaiboi, and Magnesians. It seems that, other issues aside, the issue of supervision is difficult, that is, the manner in which it is necessary to ³ Aristl. Pol. 1271b41–1272a1: ἔχει δ᾽ ἀνάλογον ἡ Κρητικὴ τάξις πρὸς τὴν Λακωνικήν. γεωργοῦσί τε γὰρ τοῖς μὲν οἱ εἵλωτες τοῖς δὲ Κρησὶν οἱ περίοικοι ‘The Cretan organization resembles that of Sparta. At Sparta the helots do the farming; in Crete, the perioikoi do the farming.’ Cf. Arist. fr. 586 [Rose] = 603, 1–2 [Gigon]: ὡς Ἀριστοτέλης ἐν Συρακοσίων Πολιτείᾳ, ὅμοιοι τοῖς παρὰ Λακεδαιμονίοις Εἵλωσι καὶ παρὰ Θετταλοῖς Πενέσταις καὶ παρὰ Κρησὶ Κλαρώταις ‘As Aristotle says in his Constitution of the Syracusans, they [the Kallikyrioi] are like the Spartans’ helots, the Thessalians’ Penestai and the Cretans’ Klarotai.’ (On the Syracusan Kallikyrioi/ Killyrioi and the aristocracy of the Gamoroi, Morakis 2015 contains much useful discussion.) Luraghi (2009): 289 n. 13 follows Lotze (1959): 8–9 in believing that Aristotle’s use of unitary terms (klarotai, perioikoi) for Cretan slave populations stems from his knowledge that there was a variety of local terms on Crete (on which, see n. 10 below), which necessitated simplification for his readers. The Cretan perioikoi are not to be confused with the free status group of the same name in Sparta.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Crete

149

deal with slaves: for if they are unrestrained, they are insolent and think of themselves as worthy of being on a par with their masters, whereas if they live laborious lives, they plot against them and hate them. It is therefore clear that those who deal with their helot system in this manner have not discovered the optimal approach.⁴

Aristotle returns to the matter of rebellions (or the lack of them) in the Cretan slave systems at Pol. 1272b15–22, after having explained the problems of stasis generated by the upper classes of Cretan cities: Having things this way is hazardous for a polis, for those who wish to attack it have the power to do so. However, as has been said, it [i.e. Crete] is saved on account of its location: distance achieves the same as xenelasia.⁵ And on account of this the perioikoi remain placid towards the Cretans, whereas the helots frequently rebel. For the Cretans take no part in overseas empire, and it is only recently that foreign war has crossed over to their island, which made clear the weakness of the customs there.

Not only were Cretan slave systems better managed in terms of preventing revolts; there were also, allegedly, structural differences in the organization of labour (Pol. 1272a13–21) which we shall consider in more detail in section VIII below. Finally, the character of Cretan slavery was seen as less oppressive than helotage. In his criticism of Plato’s Republic, Aristotle writes: For if the farmers have everything in common, what difference will there be between them and the Guardians? Or what advantage lies in it for them to remain under their rule? Or why on earth would they remain under their rule, unless the Guardians devise something clever like that which the Cretans have? For these have permitted their slaves everything apart from the gymnasia and the ownership of weapons. But if these things are to work exactly as they do in other poleis, what kind of communalism will there be?⁶

These brief comments, however, need to be viewed in light of Aristotle’s broader discussion of Cretan institutions. In an important article published over twenty years ago, P. Perlman showed that Aristotle’s picture of Crete is reductive and homogenizing, presenting the island as having a single politeia and uniform institutions.⁷ By comparing Aristotle’s account with epigraphic evidence from the individual poleis, she demonstrated a much greater degree of diversity than one would assume by depending on Aristotle alone. Subsequent study has in part confirmed Perlman’s conclusions, but shown that she ⁴ Pol. 1269a34–b12. ⁵ That is, periodic expulsions of foreigners at Sparta: see Figueira (2003). ⁶ Politics 1264a17–24. See also Ephorus FGrHist 70 F 29, who mentions a festival in Cydonia where the free leave the town to the slaves; during the festival these slaves had the right to flog free persons (presumably if they entered the town). Carystius of Pergamum (FHG IV 358), writing in the second century BC, reports something similar among the Cretans. ⁷ Perlman (1992).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

150

Greek Slave Systems

has somewhat overstated her case against Aristotle and the other literary sources: there were differences between the Cretan cities, true; but there were also a number of common institutions attested across the island, and it is better to see Aristotle as generalizing rather than deliberately misleading his readers or being poorly informed.⁸ Some forty-nine archaic and classical Cretan poleis are attested,⁹ and each will have had its own slave population and its own slave laws. We should think of Cretan slavery, then, in terms of a patchwork of slave systems arrayed along the length of the island. We know nothing of most of these groups, aside from a few local names preserved in inscriptions or attested in literary sources.¹⁰ But we are in a different position when it comes to Gortyn, for which a large amount of inscribed law that deals with slavery survives.

II. GORTYN: TERMINOLOG Y FO R SLAV ES Most of the laws of Gortyn that deal with slavery derive from the ‘Great Code’, IC IV 72, a twelve-column collection of rules, dating to c.450 BC, that is, around a century before Aristotle’s time.¹¹ Several other laws dating to the same period help to flesh out the picture. In reading the Gortyn code one is quickly struck by a peculiarity in the terminology for slavery. The code uses two words for slave: dolos and woikeus.¹² Many scholars have supposed that the use of two terms must reflect two ‘unfree’ statuses, and it has proven attractive to

⁸ See Chaniotis (2005); Link (2002; 2008). See also Gehrke (1997) and the essays in Seelentag and Pilz (2014). ⁹ Perlman (2004a): 1144–5. ¹⁰ The term aphamiotai is attested in Athenaeus 6.85.263F. Cf. an inscription from Eleutherna in the west (IC II XII 16Ab: apamiais, probably meaning land where slaves worked), and aphamian somewhat later in Hierapytna in the east (SEG 26.1049.72, second century BC). Athenaeus also mentions the terms chrysonetai (‘gold-bought’) and klarotai, a term used also by Ephorus (in relation to Kydonia: FGrHist 70 F29) and Aristotle (F 586 [Rose]). The term oiketeia is also found at Eleutherna and Lato (IC I XVI 17.16) and oikeis at Lyttos (IGT 117). Sosicrates (FGrHist 461 F4) claims that the Cretans call public slavery mnoia, private slaves aphamiotai and the subject population perioikoi. Late literary sources like these mentioned by Athenaeus preserve, in an albeit garbled form, several local Cretan words for slave populations, but they try to impose an artificial order and streamlined meaning upon them which reflects the taxonomic interests of these writers more than the original meaning of the terms (cf. Link 2001: 101–10). ¹¹ On Gortynian law, see Davies (2005a). For the methodological principles needed when treating this material, see Lewis (2013): 392–8. In the following discussion I shall refer to IC IV 72 as ‘the code’ for reasons of convenience and convention. Text, translation, and commentary on all of the laws discussed in this chapter can be found in the important collection of Gagarin and Perlman (2016). ¹² As with the term οἰκέτης, it is a mistake to translate woikeus as ‘domestic slave’. See the Appendix to this volume.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Crete

151

equate these with the two forms of servile status often postulated for the Greek world in traditional scholarship, that is, ‘chattel’ slaves imported from abroad and indigenous ‘serf ’ populations comprising local Greek inhabitants ‘bound to the soil’. For instance, several scholars have argued that the woikeis were an aboriginal group of ‘serfs’ who were bound to the soil and yet in possession of special legal privileges, to which an influx of ‘chattel’ doloi were later added.¹³ Yet there are major problems with this picture. The first cracks in this approach were exposed over a century ago by Lipsius, who showed that in terms of their usage in the code, dolos and woikeus were synonyms, and to read them as anything else produces confusion. Building on this work, S. Link has recently shown that Lipsius’ earlier analysis was indeed correct, and provides a fuller analysis of the problem.¹⁴ The reasons why dolos and woikeus must be words referring to the same status can be briefly summarized. When the code discusses individuals generally, it discusses them in terms of free persons and slaves; that is, two—not three—basic groups (e.g. IC IV 72 I 2, 4–5, 9, 15–16). Likewise, there is only one term for the master of a dolos or woikeus: pastas, a term that simply means ‘owner’.¹⁵ Furthermore, one never finds a rule stating that ‘if a dolos or woikeus’ does a certain thing, then certain consequences will follow: one term or the other is used, but never both side by side in this manner. In other words, they are never set up as alternative groups in the code. The equivalency of dolos and woikeus is most clear from the provisions on rape and seduction in column II of the code, for here the terms are used interchangeably. If we suppose them to be synonyms, what is produced is a logical and complete list of rules. If they relate to different groups, we are forced to conclude that the individuals responsible for drafting the code produced an illogical set of rules on seduction and rape full of major gaps. The ratios of the fines related to these offences further cement the conclusion that we are dealing with two terms for one status. The principle is obvious from Tables 7.1 and 7.2,¹⁶ where the proportions of the fines are internally consistent and logical if we proceed on the assumption that dolos and woikeus are two words denoting the same status.

¹³ Vinogradoff (1922): 208; Willetts (1967): 14–15; Lévy (1997); Gagarin (2010a). A succinct example of this classic position: Champion (1997). ¹⁴ See Lipsius (1909): 397–9; Lotze (1959): 18; Maffi (1997a): 119–21; Link (1994): 30–48; (2001); Welwei (2008): 2–3. Useful discussion in Finley (1981): 135–7, with whom I mostly agree, apart from the issue of slave ‘rights’, for which see Lewis (2013). ¹⁵ The word is derived from paomai. There is a different word for the (temporary) master of a debt bondsman: katathemenos. ¹⁶ Adapted from Gagarin (2010a): 17. All fines are converted into drachmas to enable ease of comparison.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

152

Greek Slave Systems

Table 7.1 If the slave is a victim, the penalty is ¹/₄₀ of that exacted were the victim a free person Reference (IC IV 72)

Offence and offender

Victim

Fine (drachmas)

ratio

II 2–4

Rape by an eleutheros Rape by an eleutheros Rape by a dolos

eleutheros

200

Slave victim: 1/40 penalty

woikeus

5

eleutheros

400

Rape by a woikeus Seduction by a dolos Seduction by a dolos

woikeus eleuthera

10 400

dolos

10

II 8–9 II 5–7 II 9–10 II 25–6 II 27–8

Slave victim: 1/40 penalty Slave victim: 1/40 penalty

Table 7.2 If the slave is the offender, twice the penalty is exacted than if the offender were a free person Reference (IC IV 72)

Offence and victim

Offender

Fine (drachmas)

Ratio

II 2–4

eleutheros

200

Slave offender incurs × 2 penalty

dolos

400

eleutheros

200

dolos

400

II 8–9

Rape of an eleutheros Rape of an eleutheros Seduction of an eleuthera Seduction of an eleuthera Rape of a woikeus

eleutheros

5

II 9–10

Rape of a woikeus

woikeus

10

II 5–7 II 20–3 II 25–6

Slave offender incurs × 2 penalty

Slave offender incurs × 2 penalty

As these tables show, in different situations where a slave is the victim of an offence, the penalty imposed is consistently one-fortieth of the penalty imposed in otherwise identical situations where the victim is a free person. Likewise, in situations where the offender is a slave, twice the penalty is imposed as would be the case were the offender a free person. This trend makes best sense if we believe that dolos and woikeus are two terms for a single legal status. J. K. Davies has rightly pointed out that of the various arguments used by opponents of the view that dolos and woikeus denote the same status, the ultimate cause of their doubt boils down to a single issue: the fact that two terms, rather than one, are used.¹⁷ It is, therefore, important to explain why

¹⁷ Davies (2005a): 317.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Crete

153

this should be the case. This seemingly illogical mixture of terms is best understood in the context of the compositional prehistory of the code.¹⁸ Since the code collects and combines earlier legal material, it seems probable that the use of multiple terms for slaves reflects the language of the earlier individual enactments, which was retained when they were incorporated into larger units. The result may look very strange to us today, and not at all in line with our own views on legal draughtsmanship which would demand standardized terminology, but if both terms simply meant ‘slave’ to the mind of a Gortynian reader (just as a document blending the terms doulos and oiketes would appear to an Athenian, e.g. Thuc. 3.73; Pl. Resp. 578e–579a; Dem. 49.55–7; cf. [Xen.] Ath.Pol. 1.11; Lys. 7.16–17), there need not have been any confusion created for the code’s original audience. Choosing to be conservative, those responsible for the writing of the code retained the original language of the earlier constituent enactments, rather than smoothing and harmonizing the terminology.¹⁹

I II . S L A V E S T A T U S AN D S L A V E OW N ERS H I P AT GORTYN Dolos and woikeus are, therefore, terms denoting the same legal status. A detailed analysis shows that this status was one of slavery, not serfdom. In fact, Gortyn’s laws on doloi/woikeis fit neatly in the scheme of private ownership that we have been employing so far. Gortyn’s laws recognize the owner’s title to his slaves and regulate the boundaries between status groups (Right to possess). We come across this issue at the very beginning of the code. IC IV 72 I 15–18 governs a case concerning a man who is contended by one party to be a slave, the other to be free, giving the benefit of the doubt to the latter party (presumably if there were no other evidence to facilitate a clear decision either way).²⁰ The text then turns to the case of an individual who is beyond doubt a slave, but whose ownership is contended between multiple parties (IC IV 72 I 18–35). It provides guidance for the judge in deciding the case and lays down rules for the return of the slave to his true owner, as well as penalties for any delays in the return. An exception is the man serving as kosmos (chief

¹⁸ On which, see Davies (1996); Kristensen (2004b). ¹⁹ Lewis (2013): 392–3. For the term oiketes and its relation to doulos in classical Greek, see the Appendix to this volume. ²⁰ On the judging procedure here see Gagarin (2010b): 132 n. 10; cf. (1988). For a comparison of these laws with their Attic equivalents, see Maffi (1997b).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

154

Greek Slave Systems

magistrate), who has to wait until his magistracy has ended before litigating on these issues (IC IV 72 I 51–5).²¹ Some scholars have argued that Gortynian slaves were able to own (rather than possess) money, claiming that the rules on rape and seduction in column II of the code prove that slaves had the right to own money from which they would pay any fines imposed upon them. If this is true, it raises the possibility that if they were the victims of an offence, they also received the fines imposed on perpetrators (rather than their masters).²² On this view, slaves had formal title to the money in their possession, which, if true, would be a startling concession by their masters and would contradict the general principle we have observed so far, i.e. that owners, not slaves, had a Right to income generated by slave labour. At first glance the code seems to support this view, insofar as it uses the verb katastasei at lines IC IV 72 II 4, 7, and 12–13. But if one examines the rules on debt bondage found in IC IV 47, it is clear that the owner, not the slave, was legally entitled to damages.²³ These rules attest to the practice of owners placing their slaves at the disposal of other free persons to whom the owner owes money, so that the debt can be worked off. This inscription describes two scenarios. First, if a slave placed by his owner in debt bondage with the owner’s creditor commits an offence against a third party, the third party is to proceed against the creditor if the creditor gave the slave the order, but against the slave’s original owner if the slave committed the act on his or her own initiative. The slave is not held personally liable for the act in the eyes of the law.²⁴ Second, in situations where a slave given into debt bondage was the victim of an offence by a third party, both the owner and the temporary master (katathemenos) could agree to litigate, and if successful, they split the damages. If one of them did not wish to litigate, the other one could do so unilaterally, and if successful, kept all of the damages to himself. In either scenario, the slave received nothing (IC IV 47 10–16).²⁵ If we read the section on sexual offences in column II of the code against this background ²¹ The rule on the kosmos seems to refer to disputes over free men and slaves: Kristensen (2004a): 75 with note 8. ²² Lévy (1997): 34; Willetts (1967): 14; Davies (2005a): 316; Gagarin (2010a): 26. ²³ On debt bondage in Gortynian law, see the important study of Kristensen (2004a). ²⁴ IC IV 47 A 1–8: [?] κατακείμενος αἰ κ’ ἀδικήσει δ͂ολος ἢ δόλα, ὄτι μέν κ[α κατα]θ̣εμένο κελομένο ἀμάρτηι τ͂οι καταθεμένοι τὰν δίκαν ἤμην, ὄτι δὲ κ’ αὐτὸς πρὸ ϝιαυτ͂ο τ͂οι ἀρκαίοι πάσται τὰν δίκαν ἤμην τ͂οι δὲ καταθεμένοι μή. ‘If a (slave) debt bondsman or a (slave) debt bondswoman wrongs (someone), in that he does wrong on the orders of his temporary master, the case is to be brought against the temporary master, but in that he himself (does wrong) on his own, the case is to be brought against his original owner and not his temporary master.’ ²⁵ IC IV 47 10–16: αἰ δέ κα τὸν κατακείμενον ἀδικήσει ἄλλος, αἰ μέν κ’ ἀνπότεροι μολίοντες νικάσοντι, τὰν ἠμίναν ϝεκάτερος ἐκσίοντι· αἰ δέ κ’ ὀ ἄτερος μὴ λῆι, ὀ ἄτερος μολίον αἴ κα νικάσει αὐτὸς ἐκσεῖ. ‘But if someone else wrongs the (slave) debt bondsman, if both [i.e. the original owner and the temporary master] bring suit and win, they shall each have half (of the penalty). And if one of them does not wish (to bring suit), if the other brings suit and wins, he shall have the penalty (himself ).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Crete

155

and with this principle in mind, we can see that the notion that slaves had a right to receive fines or were personally liable to be fined is unwarranted.²⁶ Clearly, the rules on sexual offences rely on the assumption of the Gortynian reader that fines went to the master of the slave victim, and that if slaves perpetrated the offence, their masters were liable to be fined (cf. p. 156 below). The code does not state this principle in so many words because it was too obvious to a Gortynian reader to require elaboration. The principle that slaves could not own property is further demonstrated by IC IV 72 VI 56–VII 10, a law that deals with a situation in which a sexual union is established between a free person and a slave. If a free woman bore both slave and free children, only the free children inherited her property; and if she had no free children, her kin were next in line to inherit (IC IV 72 VII 4–10). Again, the slaves received nothing. Yet another sign that the fruits of slave ‘labour’ accrued to slaveholders alone are those rules that apportion the children of slave couples: in situations where slave couples belonged individually to different owners, there were elaborate rules on which owner gained property rights to the child; but slaves never gained rights to the child (IC IV 72 III 52– IV 23).²⁷ Slaveholders at Gortyn had a Right to capital and could thus sell their slaves (IC IV 72 VII 10–15; cf. IC IV 41 IV 5–17²⁸). Sales were not unrestricted, however: slaves who were serving as debt bondsmen to pay off the obligations of their owner could not be sold by that owner until the debt service was completed (IC IV 72 X 25–9), essentially suspending the owner’s right to alienate his slave for the duration of the debt service. Private manumission appears to have been possible, as may be indicated in IC IV 78.1 (if the restoration of ton apeleu[theron] is correct).²⁹ The protection of owners’ property rights (Right to security) is evident in the rules on debt bondage that we have just considered, for these provide legal redress for an owner whose slave has been damaged by a third party (IC IV 47 10–16). Owners also had guarantees that if they placed their slave in debt bondage, the temporary master (katathemenos) could not kill or sell the slave, and had to alert the owner if the slave ran away to a temple (IC IV 47.16–33).³⁰ Other rules buttress the owner’s right to hold onto his slaves: for example, in IC IV 72 I 39–49 we find rules regarding slaves who take sanctuary in a temple,

²⁶ Cf. Link (1994): 37–8. I discuss this at greater length in Lewis (forthcoming e). ²⁷ Lewis (2013): 402–4. ²⁸ This rule bans individuals from selling woikeis who have run away from their masters and taken refuge in a temple; it sets an expiry date on the sale ban of one year from when the slave escapes. Implicit throughout is the assumption that woikeis can normally be sold. ²⁹ Willetts (1954); Chaniotis (2005): 188 n. 63; cf. Seelentag (2017). Gagarin is probably correct in rejecting the idea of some scholars that IC IV 62 concerns manumission: Gagarin and Perlman (2016): 328–9. ³⁰ Kristensen (2004a): 74.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

156

Greek Slave Systems

in which case anyone with knowledge of this is obligated to inform the owner of the slave’s whereabouts. IC IV 41 IV 5–17 complements these rules by preventing anyone from selling a slave who has fled from his master and taken shelter in a temple until a year has passed.³¹ Fines pertained in cases where a man in illegal possession of a slave was tardy in returning him to his owner (IC IV 72 I 1–14; 24–35), and also in cases where a slave was raped or seduced (cf. pp. 154–5 above). In a separate law (IC IV 43Ab) we find rules against unjustly seizing slaves in restitution for an unpaid debt or taking their clothes or ornaments. In this rule, the penalty imposed was a fine half the amount that would be exacted were a free man unjustly seized; and if it were the slave’s clothes or ornaments taken forcibly, the penalty was a third of that exacted were the victim a free man.³² The aspect of duration, comprising Absence of term and Transmissibility, can be seen in IC IV 72 V 39, where the property inherited by the children or heirs of the deceased is called tnaton or mortal property, i.e. livestock and slaves.³³ The assertion of some scholars that slaves had a place in the order of succession is rightly rejected by most specialists: the principle that slaves could not inherit property is clear in IC IV 72 VII 4–10.³⁴ The owner’s liability for his slave’s actions (Prohibition of harmful use) is evident if we turn again to the provisions relating to slaves held in debt bondage. We noted earlier in this section that if a slave held in debt bondage caused damage to a third party, the responsibility for the action lay with the slave’s owner. The owner was not responsible, however, if the actions of the slave were carried out on the instructions of the katathemenos (IC IV 47.1–10).³⁵ The same principle is evident in the provisions on rape and seduction in column II of the code, setting fines for sexual offences perpetrated by slaves which were, as we have seen, the responsibility of their owners. And in the law on selling a dolos, the buyer had a sixty-day grace period within which he could return the slave to the vendor; after that, he became liable for the misdeeds of the slave, whether committed before or after the sale (IC IV 72 VII 10–15; cf. IC IV 41 VII³⁶). State interference in the use of privately owned slaves is indicated in a fragmentary inscription, IC IV 79. This document mentions a payment in kind, made up of barley, figs and must (pressed grapes), which is paid to both

³¹ See Naiden (2006): 373–4 for the practice of slaves running away to a temple and demanding to be sold to a new owner. ³² Edward Harris points out to me a problem in the translation of ἐνεκυράδδο in Gagarin and Perlman (2016). Gagarin translates this in terms of taking an object in pledge, but it should be translated in terms of seizing an item in restitution for an unpaid debt (cf. Att. ἐνεχυράζω with Harris 2008). ³³ See Lévy (1997): 32; Brixhe and Bile (1999): 92. Lévy (1997): 32 n. 22 points out the opposite formulation athanata kremata in IC IV 76 B 8–9. ³⁴ See Lewis (2013): 408–9. ³⁵ Kristensen (2004a): 74. ³⁶ See Davies (1996): 46.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Crete

157

free persons and slaves dwelling in the city as recompense for certain labour services rendered, with penalties for those who refuse to work administered by the kosmos for foreigners. Given the specific magistrate charged with imposing the fine, we should perhaps view this document as detailing some manner of compulsory labour services on subaltern, non-citizen statuses, but it is difficult to be certain.³⁷ Some scholars have argued that Gortynian slaves had special rights to marry and own property that did not exist in other poleis. However, a close look at the texts that they use to support this view shows that the notion of ‘rights’ is mistaken: Gortynian slaves—like their Homeric predecessors—certainly could form ‘marriage’ relationships, possess property, and have children; but the law was concerned with clarifying the property rights of slaveholders, particularly in complex scenarios where two slaves belonging to different owners might ‘marry’, entangling the property interests of the individual owners.³⁸ The ‘rights’ of the Gortynian slave—just like the much-vaunted rights of helots— turn out to be a chimera. Despite the common appellation ‘serf ’ of Cretan slave groups, a detailed examination of Gortynian slave status shows that it displays the same basic property features that we can observe in both Homeric and Attic slavery. We are therefore dealing with slaves and slave systems in Crete, not serfs or some hazy form of ‘unfree’ labour. It is now time to ask whether we are also dealing with ‘slave societies’ in the sense that Finley intended the concept.

IV. S LAV ERY IN THE E CONOMY OF GO RTYN AND CLASSICAL CRETE Our earliest source for the importance of slavery in Cretan society is the skolion of Hybrias preserved in Athenaeus 15.695f–696a. Its message is fully consonant with the picture of early Greek slavery we have already observed in Homer, though it probably dates to the end of the archaic period (Ducat even suggests a fifth-century date³⁹); and it resonates with the archaic values that we can see also in Archilochus (Elegy 2 [West]) and Homer (Il. 12.310–21; Od. 1.397–8). In the song, an aristocratic warrior asserts that his social rank and power derive from his arms and his ability to use them. It is these that bring him substance and sustenance, and that ensure his mastery over the mnoia,

³⁷ The document is discussed in Davies (2005b): 162. ³⁸ The issue is too complex and lengthy to develop here; my position is more fully set out in Lewis (2013); see also Lewis (forthcoming e). ³⁹ See Ducat (1990): 74 with note 15; but cf. Tedeschi (1986) on the epithet μέγας βασιλεύς.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

158

Greek Slave Systems

an old Cretan word for slaves.⁴⁰ For those lacking martial accomplishments, the only course of action possible is to bow down and recognize him as master and Great King. Here, there is a clear division of roles: the warrior fights; his slaves work; and the fact that the warrior fights justifies his domination over the slave workers. The next time we catch a glimpse of Cretan slavery in the literary records is in the fourth-century writings of Aristotle and several other nonCretans. As we have noted, the observations of non-Cretans of this period on Cretan society must be treated with caution, for they usually gravitate towards pan-Cretan generalization, often obscuring local idiosyncrasies. Nonetheless, these nuggets of literary evidence should not be dismissed out of hand, and may provide clues to understanding the epigraphic texts if employed with due caution. Let us turn back, then, to Gortyn’s inscriptions, in order to see how far they can take us towards a reconstruction of the place of slave labour in economic life. In a recent article, J. K. Davies has deftly sketched the contours of Gortyn’s economy in the classical period, and we may summarize his findings before proceeding to the matter of slave labour.⁴¹ By the fifth century BC—the century to which both the Great and Little Codes belong—Gortyn dominated the lower Mesara plain, including several of the nearby poleis, right down to the coast. This gave her access to fertile arable land as well as to the sea and the surrounding foothills, the latter providing pasturage. Inscriptions mention all of those things that we would expect in a normal system of Greek polyculture: livestock (cows, goats, sheep, and pigs) as well as the paraphernalia associated with cereal production and processing (ploughs, yokes, mills, and millstones) and arboriculture (figs, grapes, olives).⁴² Land ownership followed the general Greek pattern, whereby alienable private land owned by the citizenry and (perhaps) other free individuals was interspersed with some public landholdings and the property of temples.⁴³ Of Crete’s common institutions, the most central (and one related to citizen militarism) was the system of common messes (andreia), in some (but not all) respects like Sparta’s citizen mess groups, a matter to which we shall turn in due course. The existence of this feature of Cretan life depended on having an adequate body of slaves to work the land and provide the citizens with the leisure and substance with which to practise and perpetuate their militaristic lifestyle. We should not envisage Gortynian society, however, as a simple division between a small body of citizens and a mass of slaves; it was more complex and layered than that. A stratum of (presumably poorer) free men called apetairoi existed as well—the name implies that such individuals did not ⁴⁰ The word is cognate with δμώς: see LSJ⁹ s.v. μνοΐα. ⁴¹ Davies (2005b). ⁴² Animals and tools: Davies (2005b): 164. Wine and figs: IC IV 79; IC IV 72 X 34–9. Olive oil and wine in the Cretan economy are discussed in Hadjisavvas and Chaniotis (2012): 163–70. ⁴³ Davies (2005b): 160–1.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Crete

159

belong to the mess groups that denoted full citizenship.⁴⁴ They lacked full citizen rights (like the hypomeiones at Sparta and atimoi at Athens), but there is no reason to suppose that they could not own land. To this mix we may also add foreigners and freedmen, who came under the jurisdiction of a specific magistrate (the xenios kosmos).⁴⁵ We may thus envisage Gortynian society as constituting a slave-owning citizen class, economically differentiated to be sure, and surely not, at least at the lower end of the scale, entirely leisured like their Spartan counterparts; living alongside them were sub-citizen statuses, foreigners, and a comparatively numerous class of slaves. We may tease a few further items of information from our sources in relation to the functioning of slavery at Gortyn.

V. MAN AGEMENT Aristotle, as we observed earlier, noted the lack of slave revolts among the Cretan cities, a fact that stood in contrast to the situation in Sparta and Thessaly (Pol. 1269a34–b12). His explanation for this difference is not wholly convincing: he claims that the Cretan cities had a kind of unwritten rule of conduct in war whereby they refrained from enlisting the support of an enemy’s slave population in times of conflict. The hostile neighbours hemming in the boundaries of Spartan and Thessalian territory, he writes, did not share this rule of conduct. Perhaps he is referring to a genuine practice, but it is equally possible that his explanation is conjectural.⁴⁶ A more convincing explanation may lie in the realm of geography. We have seen that Sparta controlled a vast swathe of territory which teemed with helots, though her citizens were restricted to spending most of their time in Sparta itself: this created quite a headache in terms of the logistics of slave supervision. Crete, by contrast, presents quite a different profile. Here, with a patchwork of microterritories arrayed along the length of the island, absenteeism will not have been so large a problem. The sheer physical distance between Cretan slaves and their owners was far smaller than between the helots and their Spartan masters. Cretan systems of slavery, therefore, did not require any of the elaborate state measures of discipline that existed in contemporary Sparta.⁴⁷

⁴⁴ Differences in wealth also existed within the ranks of the citizens: see Link (2009). ⁴⁵ See Seelentag (2017). ⁴⁶ Note the ἴσως, ‘probably’ or ‘perhaps’, that introduces his explanation. ⁴⁷ Cf. Luraghi (2009): 279 ‘none of the poleis of Crete or Thessaly controlled a territory comparable for complexity and extension to that of Sparta, and this may explain why those systems appear to have been more stable than helotage.’ Luraghi also points out (2009: 294 n. 94) that the average territory of a Cretan city was around 167 km² compared to Sparta’s 8,400 km².

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

160

Greek Slave Systems

VI. F AMILY L IFE The major factor in which Cretan slavery resembled helotage lies in the ethnic uniformity of the two slave populations. This is why both regions appear in the lists of slave systems in which the slaves all spoke a single language; and such a picture dovetails well with our understanding of Cretan foreign trade, which even on an optimistic reading was low in volume and localized in extent, and therefore presents an unsuitable context for the kind of steady high-volume system of slave imports that could be found in Attica.⁴⁸ It was the consequent need to foster the slave family as a social institution that explains the attention given in Gortynian law to the clarification of slaveholders’ property rights in complex scenarios where slaves belonging to two different owners might ‘marry’ and start a family, or where slaves and free (though probably lowstatus) individuals might intermarry. I have argued elsewhere that these arrangements should not be viewed in terms of slave ‘rights’, but evidently there was some de facto latitude for slaves to find partners of their own choosing.⁴⁹ If this analysis is correct, we should expect relationships between slaves belonging to two different owners to have been relatively common; we do not know the average number of slaves owned by Gortynian citizens, but it is unlikely to have been as high as the Spartan average (estimated at 25 by Hodkinson), and comparative evidence suggests that even in holdings larger than this, slaves might normally prefer to look further afield for marriage partners.⁵⁰

VII. RESIDENTIAL P ATTERNS A difficult but revealing law in the Great Code mentions the dwellings of slaves in the countryside and their associated livestock (IC IV 72 IV 33–6).⁵¹ One of the similarities between Spartan and Cretan social life seems to have been the focus on the town itself, where the dwellings of citizens were located and where mess groups met. The klaroi belonging to the citizens, however, were spread throughout the town’s hinterland. This implies that any klaroi that were too far from the town to be reached easily by foot on a daily basis could not have been worked by slaves residing in their owners’ town

⁴⁸ See the essays in Chaniotis (1999) and Davies (2005b). Cf. Perlman (2004b) and Erickson (2005) for a more optimistic picture. See Hadjisavvas and Chaniotis (2012): 165 n. 49 for Chaniotis’ criticism of the optimistic view. For Attica’s slave supply in contrast to Sparta and Crete, see Lewis (2015a). ⁴⁹ Lewis (2013). ⁵⁰ Hodkinson (2008): 315 n. 125; Lewis (2013): 411–14. ⁵¹ For the interpretation of this difficult text, see Lewis (2013): 406–8. Cf. Maffi (1997a): 64–70; Probert (2015): 374–7.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Crete

161

houses;⁵² these slaves must have dwelt in country dwellings of the sort that the aforementioned law describes. In the case of Gortyn, some of the foothills suitable for pasturage are around 20 km from the town itself. The closest we can get to garnering a sense of what life might have been like for slaves out in the foothills is to turn to the Odyssey, where a similar residential pattern is envisaged for those slaves involved in the raising of livestock. Eumaeus, if we recall from Chapter 5, dwells out in the countryside with four other slaves in a small hovel and manages a herd of 360 pigs (Od. 14.20–5). When Odysseus approaches Eumaeus’ dwelling, one of the slaves is absent: he has been sent by Eumaeus into town to deliver a pig to the suitors for sacrifice and consumption (Od. 14.26–8). This is a regular practice: each of the managers of Odysseus’ numerous flocks and herds sends an animal to his town house daily, the animal being presumably driven into town by one of the slave herders (Od. 14.105–6). Some of Odysseus’ slaves work even farther afield, on the mainland opposite Ithaca (Od. 14.100–4). The lives of Gortynian slaves living at the farther reaches of Gortynian territory and managing the livestock of their town-dwelling owners cannot have been greatly different from the scenario described by Homer.⁵³ This raises another issue: the dwellings of female slaves. Since Gortyn’s slave population was replaced through reproduction, female-to-male ratios must have been about equal. The Great Code makes it clear that women were involved in the domestic manufacture of textiles (IC IV 72 III 24–31), and one of the tasks of female slaves in Gortynian society must have been assisting their mistresses in textile manufacture. When envisaging the distribution of slaves within the Gortynian landscape, then, we must take into account the different gendered production roles and their physical locations. That is not to say that we should imagine a simplistic residential pattern for slaves, with the males in the countryside and the females in town: it would have made sense for any males slaves involved in the working of klaroi within walking distance from the town to be resident there; and there is no reason to suppose that the countryside was devoid of female slaves, nor that the work of females was limited to indoor work, nor that textile production need have taken place exclusively in town. But even in cases where male slaves worked some distance away from the town, it is quite feasible that they might have ‘commuted’ on occasion to see their urban-dwelling wives and children, an arrangement that can be paralleled in other slave systems.⁵⁴ The importance of slave families and the nature of Gortyn’s economy and territory thus present us with a complex situation, one whose details are often obscure because of a

⁵² The same goes for any out-of-the-way public land on which animals were pastured, something that the literary sources mention (Arist. Pol. 1272a18). For the Ackerbürger model of farming and settlement in the Greek world, see Hansen (2006a): 93–4. ⁵³ For upland pasturage in Crete, see Bresson (2016): 136–41. ⁵⁴ West (2004); Lovejoy (1981).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

162

Greek Slave Systems

paucity of evidence, but whose basic contours can be mapped out with a fair degree of confidence.

V I I I . P U B L I C L A ND AND THE ANDREIA It would appear that the common messes in Crete were organized somewhat differently from those in Sparta, for Aristotle mentions a major structural difference between the two regions.⁵⁵ In Sparta, a citizen had to bring without fail his own contributions to the mess, but in Crete the state and its officials were responsible for the collection of produce, which was then redistributed to the andreia: The arrangments for syssitia among the Cretans are superior to those of the Lakonians. For in Lakedaimon each person contributes a fixed amount per capita, and if he does not, a law prevents him from having a share in the citizenship, as was said earlier;⁵⁶ in Crete things are more communal. For from all of the crops and cattle produced from the public land, and from the tribute which the perioikoi bring, one part is earmarked for the gods and public use, the other for the syssitia, with the result that everyone is supported by the common funds: women, children, and men.⁵⁷

As we have already noted, Aristotle’s generalizations should be handled with caution and, where possible, tested against the epigraphic evidence from specific poleis for goodness of fit. At first glance, it appears that Aristotle’s picture of the Cretan messes fails to fit Gortyn’s epigraphic evidence, for IC IV 75B provides a list of property that cannot be seized as pledge, including ‘whatever the arkos (= Att. kyrios) provides for the andreion’. This would appear to contradict Aristotle insofar as it suggests that Gortynian andreia received contributions from individual households (as in Sparta) without the state acting as an intermediary. But Davies has pointed out other evidence that, in combination, suggests that the picture in Aristotle for the most part (though not in every respect) corresponds to Gortynian arrangements. First is a law that refers to a magistracy of ‘harvest collectors’, implying that the state did make exactions on the produce of individual households rather than just public land (IC IV 77B). Second, and more suggestive, is a later treaty with Kaudos that commands that the inhabitants of Kaudos pay a tenth of their produce (cereal, not animal) to the Gortynians, just as the Gortynians pay a ⁵⁵ Haggis et al. (2004): 380–2 suggest that a large building in Azoria in eastern Crete was probably an andreion. The identification is endorsed in Hadjisavvas and Chaniotis (2012): 164. On the andreia, see now Seelentag (2015): 374–443. On the development of Cretan drinking practices, see Rabinowitz (2014). ⁵⁶ Arist. Pol. 1271a26–37. ⁵⁷ Arist. Pol. 1272a13–21.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Crete

163

tenth to the state (IC IV 184. 8–11). Third is the system of public messes at Lyttos described by Dosiadas (FGrHist 458 F 2), in which individual households contributed one-tenth of their produce to the state, each slave contributing as well an Aeginetan stater per head; this income was then redistributed by the state to the mess groups, to which all citizens belonged. This arrangement dovetails well with the information relating to Gortyn and Kaudos. Davies writes the following: If classical Gortyn and third-century Lyttos shared the same system, and if the statements of Aristotle and Ephoros can be read as incomplete rather than false, one can create a consistent picture of procedure. The basic step is that a tithe of all agrarian produce is levied (with provision against concealment to be enforced by public officials) and is supplemented by levies in kind or in coin from noncitizens and from public lands. After some has been top-sliced for cult and public administration, these resources are then redistributed by the city to the andreia and can in turn be supplemented by members.⁵⁸

The rule in Gortyn mentioning contributions to the andreion could, therefore, mean produce earmarked for payment to the state for redistribution to the andreia, or perhaps produce earmarked to supplement the standard contribution (cf. Xen. Lac.Pol. 5.3 for Sparta), though the latter possibility seems less likely, as it appears in a list of items that cannot be distrained upon by creditors, which is unlikely to include optional extra contributions to a mess. This picture is far from certain, but makes the best sense out of the evidence we have. If Davies is correct in supposing that, as at Lyttos, Gortynian slaves had to pay a monetary contribution to the state to support the andreia, it is a further sign of the high degree of independence in day-to-day activities enjoyed by Gortynian slaves, who (as other texts suggest) experienced some degree of family life and were able to possess and utilize a certain amount of property, including coinage.

IX. S LAVE SOCIETY AT GORTYN AND ACROSS CRETE Once again we need to highlight a contrast between Sparta and Crete. At Sparta, the state did not guarantee its citizens security of status: this was left up to private initiative, and diminution in the productiveness and size of estates through mismanagement, crop failure, or fragmentation due to female inheritance allowed the whittling-down of some citizen estates to the point where their owner produced too little to maintain his citizen status. The fact that Sparta’s ‘lawgiver’ failed to foresee this eventuality and prevent it was heavily ⁵⁸ Davies (2005b): 167.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

164

Greek Slave Systems

criticized by Aristotle (Pol. 1270a15–b6). By contrast, both the literary and epigraphic sources from Crete show a marked intervention of the state to prevent this kind of eventuality from occurring. The literary evidence on andreia shows that rather than a fixed level of produce determining citizen status, citizens paid a tithe on whatever they produced, so that in practical terms all citizens will have paid different amounts to the state. This was then redirected to the andreia, an egalitarian practice (if we restrict that term to the citizen body alone) that will have enabled all citizens and their male descendants to attend, and to maintain the level of wealth and leisure required to fulfil the duties of a citizen.⁵⁹ The epigraphic evidence from Gortyn points in the same direction: the rules on objects that cannot be pledged (IC IV 75B), those on debt bondage (see section III, p. 154 above), the rules on divorce (IC IV 72 II 45–III 16), widows (IC IV 72 III 24–30), the patroiochos (IC IV 72 VII 15– IX 24), and adoption (IC IV 72 X 33–XI 23) all aim to protect citizen households from becoming impoverished, either by minimizing the transfer of property away from males to females or by protecting citizen households from the extremes that indebtedness can inflict. As Davies notes: the regulations are reacting to, and are trying to minimize the effects of, processes of economic and demographic instability, presumably so that the body of male citizens can perpetuate itself and thereby be strong enough to protect its hold on its territory and to retain control over its dependent labour.⁶⁰

Although there was variation from one Cretan city state to the next in terms of structural detail, some broad similarities can be noted for the island as a whole. As Gehrke has shown, the warrior lifestyle, upbringing, and brigading of the male citizenry that focused on the andreion was a pan-Cretan practice.⁶¹ We need not suppose that citizens of the Cretan city states were prevented from engaging in any occupation other than warfare, as was the case at Sparta; but their lifestyle did require a high degree of leisure, and scholars are right to connect this with the existence of slave populations that bore the brunt of productive work.⁶² It is highly likely, then, that the forty-nine Cretan poleis known to modern scholars were all, in a Finleyan sense, ‘slave societies’. Like Sparta, the reason why this appellation has not been applied to Crete does not lie in disagreement over the role of dependent labour in its economy: scholars have long agreed on the fundamental role of dependent labour in maintaining the position and strength of citizen classes in both regions. Rather, the failure of scholars to notice that the Cretan cities (and Sparta) were slave societies ⁵⁹ For the education of Cretan youths and their incorporation into the citizen body, see Link (2009). ⁶⁰ Davies (2005b): 169. ⁶¹ Gehrke (1997). ⁶² Davies (2005b): 161. Of course, Gortyn’s rules on slavery cannot be extrapolated wholesale to other Cretan societies, but, that said, there is no reason to suppose that any of the other dependent populations on Crete were ‘serfs’ or anything other than privately owned slaves.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Crete

165

stems from the common tendency to label their dependent populations as ‘serfs’. The meagre evidence we possess for classical helotage, as we saw in Chapter 6, provides us with no grounds to suppose that we are dealing with anything other than a slave population; the much more detailed rules from Gortyn show that, from a comparative perspective, we are certainly dealing with slaves, not serfs or some other mysterious status ‘between free men and slaves’. Although slavery in Sparta and Crete differ profoundly on a structural level from Attic slavery (and from each other), all three regions present slave systems in which the legal status of the enslaved as property displays fundamental similarities, albeit with local differences where the standard rules on slave ownership were tailored to meet the needs imposed by local imperatives. The findings of this chapter undermine the preconceived notions of serfdom that abound in much modern scholarship. To be sure, Crete’s slave populations were overwhelmingly engaged in agricultural production. So were the slaves of the US South. The fact that the social position of these labourers can be described in terms of ‘dependent peasantry’ should not obscure the clear fact that their legal status was one of slavery. * * * A few words will serve to summarize the chief conclusions of this chapter and Chapter 6. When Plato and Aristotle compared helotage to other Greek slave systems such as Thessaly’s penestai or Heraclea’s mariandynoi, they made their comparison on one criterion: in these systems, the slaves spoke the same language. They (rightly) did not see a major categorical distinction between slavery in Attica or Chios and the slave systems of Sparta, Crete, Thessaly, or elsewhere. In each of these regions, private ownership of slaves prevailed, and in each of these regions the rules governing slave ownership and the economic structures that dictated their usage differed. Sparta and a few other regions depended overwhelmingly upon natural reproduction to maintain their slave workforces, but this has nothing to do with legal status. Scholars have been correct to emphasize the many differences between helotage and Attic-style slavery, but in so doing they have often wrongly focused on legal status, rather than on the more important variables of slave supply and the specific details of their labour arrangements.⁶³ It is time to rebalance our perspective and treat helotage (and comparable slave systems) as alternatives to the Attic model, not categorically different, but simply epichoric slave systems attuned to local economic and institutional needs.

⁶³ Cf. Zurbach (2013): 975: ‘on ne peut ni ignorer leur diversité, ni les opposer radicalement aux esclaves marchandises’.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

8 Classical Attica The relatively rich evidence for Athens and Attica in the classical period enables us to reconstruct the most complete picture of any ancient Greek slave system. Even if we limit ourselves to the economics of Athenian slavery, sufficient material survives that it would require a series of volumes to do full justice to the topic. The limits of a single chapter, therefore, require some selectivity in regard to subject matter. I aim to focus on three basic issues in order to draw out the main points of contrast between Athens’ slave system and those discussed in Chapters 5–7. First, we will look at the distribution of slave ownership across the wealth spectrum, which extended well below the elite. Second, we will look at elite sources of wealth and the contribution that slavery made to their lifestyle and dominant position in Athenian society. Third, we will explore sub-elite slaveholding. This will allow two important features of Athenian slavery to emerge more clearly. On the one hand, the Athenian elite, though leaning to a sufficient degree upon slave labour to comfortably qualify as a ‘slave society’, was not as absolutely dependent upon slave labour as its oligarchic rival Sparta, whose citizenry lived wholly off the back of helot labour. This was, to a significant degree, a function of Attica’s more developed and diverse economy, which presented the wealthy with a wider range of income-generating activities to choose from. On the other hand, their economy, enmeshed in a web of trading routes that penetrated deep into peripheral zones such as Thrace and Anatolia, connected the Athenians to an abundant source of enslaved labour available at remarkably low prices (comparatively speaking). This opened up the possibility of slaveholding to a broad swathe of the citizenry and metics, meaning that if Athens’ elite was less reliant on slave labour than elites in Sparta, Crete, or Thessaly, slavery permeated the lives and households of average free residents of Attica to a greater degree than was probably the case in these other regions.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

168

Greek Slave Systems

I. PATTERNS OF S LAVE OWNERSHIP AMONG THE CITIZENRY By the fourth century there were some 1,200 Athenians who were liable for liturgies, with fortunes of at least 3–4 talents.¹ We may designate these men the Athenian elite in an institutionalized sense: their involvement with political leadership in Athens amounted to a near monopoly, and they dominated the most important magistracies, especially the generalship.² Roughly commensurate with this body of men was the elite as a social class:³ in an emic sense, texts written by men from this class project their identity as kaloi k’agathoi and hoi plousioi (‘fine and good men’, ‘the rich’), a social class marked out by a specific leisured lifestyle, and distinguished from those who had to work for a living, whether through manual labour or in retail and service occupations, who are called poneroi, ‘wretched’ and penetes, ‘poor’.⁴ (It is far from probable that those so designated accepted this label or built their identity upon it.) It is crucial at the outset that we avoid conflating this term for poor with our own preconceptions of poverty: it is all too easy to go astray through the casual supposition that sub-elite Athenian smallholders were largely poor subsistence farmers struggling to make ends meet. To be sure, there were Athenian citizens who lived in abject poverty, but most did not, and the Greek term penia merely denotes the necessity of working for a living rather than having enough substance to support a life of leisure (cf. Ar. Plut. 552–4; Alexis fr. 167 K-A), and covers a broad swathe of the citizen population ranging from the indigent to the reasonably affluent.⁵ Despite the class language of the wealthy, Athens’ citizenry was not divided into two clear wealth strata: we are looking at a situation wherein wealth was distributed along a curve, with a large number of citizens in the comfortably off middle; the cut-off was the point at which an Athenian citizen no longer

¹ Canevaro (2016a): 47–63, contra Davies (1981): 15–37. ² That is not to say that this elite wholly monopolized political power: some individuals from humbler backgrounds (e.g. Aeschines) could achieve fame and power in Athenian politics. However, such men were the exception, not the rule: Harris (1995): 21–9. ³ I mean this in the sense set out by E. P. Thompson (1963): 9, ‘class happens when some men, as a result of common experiences (inherited or shared), feel and articulate the identity of their own interests as between themselves, and as against other men whose interests are different from (and usually opposed to) theirs’. ⁴ See Davies (1981): 10–14; Harris (1995): 19–21 and 180–1 n. 2. For a useful summary of the phrase kalos k’agathos, see de Ste. Croix (1972): 371–6. ⁵ Cf. Davies (1981): 4. Regarding living standards: on diet, see Lagia (2015); on stature, see Kron (2005) and O’Connor (2013); on wages and subsistence costs, see Scheidel (2005). These studies show the dangers of aprioristic notions about the sort of material resources at the disposal of a ‘poor’ Athenian citizen: ideas of poverty need to be contextualized in comparative-historical terms. Even poor Athenians were taller, better fed, lived in better houses, and enjoyed more material wealth than the lower classes in most premodern societies.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

169

had to labour with his own hands, and could afford the lifestyle of a man of leisure.⁶ Within the Athenian citizenry a deeply entrenched attitude militated against the possibility that one citizen might serve another as his hireling, especially for an extended period of time (e.g. Xen. Mem. 2.8; Dem. 57.35). Unlike in Rome, where clientage was an embedded feature of social relations, the Athenians openly abhorred such practices.⁷ Theophrastus sketches a revealing caricature of the odious, arrogant sort of man who treats his fellow citizens like clients: this man orders anyone working for him, or salesmen eager to sell him their wares, to attend his house first thing in the morning; he tells citizens needing to speak to him to attend him at his own convenience when he is having his evening stroll; and when holding a dinner he does not show up in person but delegates one of his cronies to entertain the guests (Theophr. Char. 24). This caricature portrays the opposite of the ideal democratic citizen. Citizens—even rich ones—were expected to conduct themselves modestly, and not strut around surrounded by hangers-on, which was the mark of a would-be tyrant (cf. Dem. 21.158). These attitudes were not a mere quirk of Athenian society, but had deep roots in the social conflicts of the archaic period, and were the product of a lengthy process of historical development. At any rate, social stigma against the kind of conduct exemplified by Theophrastus’ arrogant man, on the one hand, and against working as the long-term employee of a fellow citizen, on the other, had a significant effect on labour relations in the Attic economy. This fact is relevant to our study, because in business the vertical organization of workshops, farms, banks, and so on tended to involve slaves rather than free wage labourers.⁸ Athenian labour demands were fed by a well-organized system of slave supply from territories on the periphery of the Aegean world. The typical Athenian slave was non-Greek. A classic example is the Macronian who had served as a slave in Athens and later accompanied the Ten Thousand on their failed expedition into the Persian Empire; as the returning army passed along the southern coast of the Black Sea, he negotiated passage successfully on behalf of the Greeks with the local Macronians, whose language he recognized as his mother tongue (Xen. Anab. 4.8.4–8). Like many other foreign slaves, this

⁶ See, e.g., Davies (1981): 34–7; slight differences in Kron (2011). This curved distribution is true regardless of whether one favours Davies’ or Kron’s interpretation. ⁷ See Millett (1989). One should not conflate social attitudes towards working for another as hirelings (which were widespread) with the more specific (and elite) disdain for banausoi, which is a qualitatively different set of prejudices. Jones (1952): 18–19 rightly underscores the distinction. ⁸ Silver (2006) claims that these social attitudes are mere ‘aristocratic snobbery’ and did not matter for most Athenians, but he ignores the institutional context in which they are voiced, which in the case of Dem. 57 Against Euboulides was a courtroom. These views must have chimed with popular (and not elite) sentiments. See Canevaro (2016b).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

170

Greek Slave Systems

man had evidently been enslaved as a child and transported to Attica.⁹ This individual’s striking story must represent one of hundreds of thousands of similar tales now lost to us.¹⁰ At any rate, from an economic perspective, two facts are particularly noteworthy. First, the sources of these slaves were remarkably close: unlike American slaveholders whose slaves had to be imported from across the Atlantic Ocean, Athenian slaveholders were virtually next door to their suppliers, resulting in relatively short supply chains and low transport costs.¹¹ Slaves came from all manner of places, but the chief sources were Phrygia and Thrace.¹² The second point is that the supply of slaves was relatively abundant, another contributing factor to the relatively low price of slaves at Athens. This low price, combined with the relative prosperity of the citizen population, enabled a striking pattern of slave ownership to develop, one that plumbed well into the lower reaches of the citizenry and was very far from an elite monopoly.¹³ Classical Athens provides us with one of the most extensive series of slave prices from the ancient world.¹⁴ When approaching this evidence, one must not expect uniformity: the value of slaves could vary widely, a fact of which the Athenians were well aware (Xen. Mem. 2.5.2).¹⁵ Apart from some exceptional

⁹ Braund and Tsetskhladze (1989): 120 claim that the story ‘is not beyond suspicion’, but surely this is too sceptical; cf. the incident reported by Patrick Leigh Fermor (1958): 101: a Corsican drummer boy descended from Maniots (to be specific, a Maniot group that had left the Mani and relocated to Corsica in the seventeenth century) exclaimed, upon overhearing native Maniots speaking during the Morea expedition in the 1830s, ‘Tiens! C’est le patois de mon pays!’ Hunt (2015): 128–9 rightly takes Xenophon’s story seriously. On child slaves and the slave trade, see Lewis (forthcoming d). ¹⁰ In a recent article Mayor, Colarusso, and Saunders (2014) argue that a number of ‘nonsense’ inscriptions associated with ‘Scythian’ images on Attic vases render words from ancient Caucasian languages in Greek script, evidence (if their argument is correct) for the familiarity of the Athenian ear with the cadences of Black Sea tongues. (This would bear out to some degree the claim of the Old Oligarch that the Athenians were familiar with every dialect, Greek and non-Greek: [Xen.] Ath.Pol. 2.8; cf. Plato Comicus fr. 61 K-A for Thracian; Pl. Cra. 410a for Athenian knowledge of the Phrygian language.) ¹¹ Lewis (2015a): 322–3 and Chapter 14 of this volume. ¹² See Lewis (2011) passim and (2015a): 317–18; (forthcoming d). ¹³ Lewis (2015a): 325. That is not to say that slaves were cheap in an absolute sense: the cost of a slave would normally have consumed much or most of an average Athenian’s yearly income. The point is rather that they were relatively cheap compared to other societies at other times and in other places: see Scheidel (2005). ¹⁴ Ruffing and Drexhage (2008): 321–3 provide a conspectus of the sources. For interpretation, see Scheidel (2005), with the modifications of Schmitz (2011) and Tordoff (2013): 31 n. 176. For the value of the slaves mentioned in Dem. 27.9, see Harris (2006): 179–80. ¹⁵ On methodological issues surrounding slave prices, see Harper (2010), who points out (at p. 212) that in relation to the US South some thirty-four variables have been identified that affected the price of slaves. Factors that the Athenians viewed as desirable and that would have affected price include inter alia age (Xen. Oec. 3.10), the ability to speak Greek (Xen. Cyn. 2.3; Pl. Meno 82a–b; cf. Dem. 19.209), literacy (schol. ad Dem. 2. 134), physical beauty (Hyp. 3 passim; [Dem.] 59.18–19), craft skills (Xen. Oec. 7.41), good behaviour (Xen. Mem. 2.5.5; cf. Poll. 3.125), and exotic origins (Theophr. Char. 21.5).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

171

cases, however, it seems that the standard price range for slaves in late fifth- and fourth-century Athens lay between 200–500 drachmas for an adult, somewhat less for a child.¹⁶ These figures can be compared to real wages and grain prices, and Scheidel’s 2005 study (which undertakes precisely such a task) indicates that, relatively speaking, slaves were more affordable in Athens than in Rome and considerably more so than in the American South: ‘In terms of their cost, their distribution, and their pervasive effect on society as a whole, they were rather like our own necessary luxury, the automobile.’¹⁷ Various texts aid us in establishing upper and lower limits of normal slaveholdings among the citizenry. At the top end we possess several large figures in Xenophon’s Poroi (4.14) relating to Nicias, Hipponicus, and Philemonides, who allegedly owned 1,000, 600, and 300 mine slaves respectively. The accuracy of these figures has, however, rightly been called into question.¹⁸ More credible figures for the top end of the spectrum can be found elsewhere. Plato, for instance, thought fifty slaves to be the holdings of a very wealthy man (Resp. 9.578e).¹⁹ Demosthenes’ father owned thirty-two or thirty-three slaves who worked in a knife manufactory, and held a further twenty kline makers as security for an antichretic loan (Dem. 27.9). The defendant in Dem. 37, Against Pantaenetus had thirty slaves employed in his mining concession (Dem. 37.4). The rich metics Lysias and Polemarchus inherited some 120 slaves from their father Cephalus, most of them employed in their shield manufactory (Lys. 12.19);²⁰ and the income of Pasion’s shield manufactory is of an order of magnitude that suggests some sixty to seventy slaves working there.²¹ These figures are fairly consistent and indicate that the wealthiest Athenians probably owned on average no more than fifty slaves, though there were some exceptions. The distribution of slave ownership correlates with the distribution of wealth across the citizen population. The naval catalogue IG I³ 1032 illustrates this trend well: it lists the crews of eight triremes, with each ship’s complement divided into categories such as trierarchs, archers, and epibatai (naval hoplites), as well as citizen rowers, mercenary rowers, and slaves (whose owners’ names are given in the genitive case).²² For two of these triremes the text is sufficiently complete to illustrate differences in slave ownership between the wealthier officers and the lowlier citizen rowers. On one trireme, both

¹⁶ Scheidel (2005). ¹⁷ Golden (1988): 457. ¹⁸ Westermann (1955): 8; de Ste. Croix (1981): 553 n. 26; Roy (1994): 190; Harris (2012): 358–9. ¹⁹ Cf. Menander, Kolax B 37, [ . . . . πεν]τήκοντα πα[ῖ]δες. ²⁰ There are no grounds for the common view (e.g. Kyrtatas 2011: 100–1; Acton 2014: 28, 139) that all 120 slaves laboured in the shield works. Thompson (1982): 82 makes the more cautious inference that most did, but that is as much as we can say. ²¹ Sargent (1924): 97–8; Davies (1971): 433–4. ²² On this document, see Bakewell (2008). For the slaves, see Robertson (2008).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

172

Greek Slave Systems

trierarchs owned slaves on board, Pytheas owning two, Chairedemos two or three. Slave owners are recorded among the officers and epibatai, but none of the slaves appears to have belonged to the twenty-five citizen rowers whose names are preserved. On the other trireme, slave ownership is attested among the trierarchs and epibatai, but not among the fifteen citizen rowers.²³ We should remember that even if the inscription were complete, these figures can only serve as proxies, and do not itemize the full holdings of the slave owners mentioned in the text, who may have had other slaves whom they left ashore. All the same, the trend wherein slave ownership dwindles as we descend farther down the wealth spectrum is tolerably clear, as is the implication that Athenians wealthy enough to serve as hoplites would normally own several slaves, from whom they might take one or two aboard ship to row or, if serving on land, bring one along as a batman (Thuc. 3.17.4; cf. 7.75.5).²⁴ Some litigants may have remarked in court that everyone owned slaves (Lys. 5.5; Dem. 45.86), but this sort of claim was probably meant to characterize the ‘ideal’ Athenian citizen rather than accurately describing the economic situation of the whole citizen body. At the very bottom of the wealth spectrum, at least, were many Athenians who did not own a single slave (Xen. Mem. 2.3.3; Lys. 24.6).²⁵ Not owning a slave at all was a sign of some indigence: in Aristophanes’ Wealth we meet a formerly rich man reduced to rags, but he still has a slave boy as an attendant (Ar. Plu. 843).

II. SLAV ERY AND THE L ITURGICAL C LASS The Athenian elite could draw on more diversified income sources than their Spartan or Cretan contemporaries. Unlike Crete, which was comparatively backward in terms of economic development, Attica was by ancient standards a highly developed commercial hub, and Piraeus (a city in its own right, vying

²³ For discussion and references, see Herzogenrath-Amelung (2017): 55–7. One should not suppose that rank-and-file oarsmen never owned slaves or brought them along with them on triremes; the Old Oligarch certainly thought this occurred: [Xen.] Ath.Pol. 1.19. HerzogenrathAmelung (2017): 56 points to a possible instance in IG I³ 1032. ²⁴ Osborne (1995): 29–30. ²⁵ Aristotle’s statement at Pol. 1252b12—that for the poor man the ox takes the place of a slave (ὁ γὰρ βοῦς ἀντ᾽ οἰκέτου τοῖς πένησίν ἐστιν)—is somewhat puzzling, given that the ownership of an ox would make no sense unless one owned a farm of c.5 hectares, i.e. one befitting hoplite rank. See Osborne (1995): 33–4 for useful observations. The solution, I think, lies in the fact that Aristotle has just quoted Hesiod (Op. 405) on getting a farm started. Rather than issuing a statement on the level of poverty where someone could not afford a slave at all, then, Aristotle’s throwaway remark is probably just an elaboration on Hesiod’s point by saying that a farmer needed an ox before acquiring anything else.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

173

with Athens itself in terms of size) the pre-eminent port of the eastern Mediterranean.²⁶ And unlike Spartan citizens, the Athenian elite was in no way barred from involvement in commerce. The scale and complexity of the Athenian economy afforded the upper classes a number of opportunities for enrichment, as several writers point out. In Xenophon’s Memorabilia (3.11.1–4), Socrates asks the coy hetaera Theodote about her sources of income, listing the typical sources of wealth: ownership of a farm, rental property, or a workshop full of craftsmen. Theodote replies that she has none of these, but depends on the generosity of her friends. The important point for our purposes is that the income sources that Socrates lists were the standard ones. When Aeschines (1.105) accused Timarchus of squandering his patrimony, he listed the various sources of revenue on which a respectable man (and thus the opposite of Timarchus) might draw: an apartment block, a farm, slaves, and loans (cf. Arist. Rhet. 1361a12–16). A survey of these investments (as well as other important sources of revenue such as mining concessions) will enable us to gauge the degree to which Athenian elite fortunes depended on the exploitation of slave labour. Farming represents a convenient place to start. After examining the full range of income sources for the elite, I will endeavour to come to a balanced view of the overall contribution of slave labour to elite fortunes, before turning to slaveholding among sub-elite elements of the Athenian citizenry.

Elite Agriculture The best-known source for this issue is unquestionably Xenophon’s Oeconomicus, but we must remember that this tract is not an agronomic handbook per se (although such works did exist at the time when Xenophon was writing).²⁷ Whilst it provides many important details on slavery and agriculture, it is above all concerned with the moral aspects of managing a household and its human and material resources, as well as the social obligations of a man of consequence. It is not a practical instruction book for running an oikos, so we should not be surprised if some details are left out, such as hired labour, which is mentioned in other sources.²⁸ (Furthermore, it has little to say on

²⁶ Hansen (2006c): 42: the intramural zone may have been larger in Piraeus, but more of that zone was occupied in Athens. ²⁷ Hodkinson (1988): 44–5; Kron (2014). ²⁸ Wintering sailors seem often to have earned their keep as hired farmhands: Xen. Hell. 2.1.1; 6.2.37. Cf. Theophr. Char. 4.6 (where the assumption is that the hired hands are non-citizens). See also Teleclides fr. 8 K-A, who mentions a song of the hired labourers going to work in the fields (cf. Phrynichus fr. 14 K-A for a winnowing song; Ar. fr. 352 K-A for a women’s song in the fields; for female agricultural labour, see Dem. 57.45). Other references to hired labour: Ar. Vesp. 712; Xen. Hiero 6.10; Dem. 18.51; [Dem.] 53.21. One should note the continuity in practices

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

174

Greek Slave Systems

animal husbandry; this is only problematic if we suppose that it is a comprehensive practical work; but if we realize that it is a selective moralistic work, the problem vanishes.) At any rate, Xenophon’s tract presupposes that the workforce on a rich man’s farm will be made up of slaves, including a slave overseer, a detail that matches up to the claim in the pseudo-Aristotelian Oeconomica that slaves can be divided into two categories, workers and overseers (1344a25–6).²⁹ We learn from Aristophanes’ Knights (947–8) that a slave in a managerial role (the poet uses the verb ταμιεύω) might wear his master’s signet ring, the seal of his authority in conducting transactions.³⁰ We should note, however, that it was generally the mistress of the household who kept track of accounts and issued instructions to slaves, and that slaves in managerial roles may have been as much beholden to their mistress as to the household’s kyrios (Xen. Oec. 9.15; cf. Ar. Lys. 18).³¹ We should not neglect the importance of leasing. This must have been an attractive option for men who owned a number of parcels of land that were widely scattered across Attica. It would be impractical to send one’s core workforce across significant stretches of the countryside on a regular basis to cultivate far-flung plots. Leasing, therefore, presented one possible solution to this problem.³² Our best literary evidence for leasing comes from Lysias’ speech On the Olive Stump (Lys. 7) in which we hear of one plot of land that was leased for short periods of time to a series of individuals: Callistratus (for two years); Demetrius (for one year); Alcias, freedman of Antisthenes (for an unknown time period); and Proteus (for three years) (Lys. 7.10).³³ The lessees ranked among Attica’s poorest residents, seeking to eke a living from the soil in conditions of insecure tenure. The freedman Alcias certainly cannot have been wealthy; by way of comparison, Menander (Heros 27–30) has a character named Tibeios, a farmer and former slave, who died in penury leaving unpaid debts to his ex-owner that his children had to work off from the archaic period, where the use of a core slave workforce supplemented by some hired labour was already in place c.700 BC. See Chapter 5, section III, p. 116 above. ²⁹ That the workforce is made up of slaves is clear from Xen. Oec. 5.15–16; 13.9–11. For Wood’s untenable claim that the word oiketes means ‘servant’ and thus cannot be used as evidence for slavery in Attic agriculture, see section III below and the Appendix to this volume. ³⁰ For slaves as business agents, see Harris (2013c). ³¹ See Harris (2014): 196–200. ³² Davies (1981): 54. There were others: slaves could have resided in the vicinity of the plot in question and cultivated it with little managerial intervention from their owners; or the owner could have sold the plot and used the money to buy something closer to home. On the market for land in classical Attica, see Harris (2015). ³³ On leasing, see Osborne (1988), who underscores the unevenness in our sources: epigraphic evidence for the leasing of land mostly relates to demes or associations; we are poorly informed about private leasing, but it may well have been quite significant. Foxhall (1992): 155 claims that private leasing was almost never practised in Attica, inexplicably citing Osborne (1988), who argues no such thing.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

175

(cf. Isoc. 7.32–3 for a more optimistic view). Likewise, among our epigraphic sources for metic occupations, georgos (‘farmer’) is a frequent entry: these men must have either rented land (like Alcias) or hired themselves out as day labourers.³⁴ Evidently, Athens’ classical efflorescence did not benefit everyone equally, and if we wish to locate Attica’s poorest farmers living on the breadline, it is here that we should look. At any rate, when it comes to considering elite farming, we must opt for a flexible picture: farms of a certain size will surely have had a core workforce of slaves, supplemented when needed by hired labour. Far-flung holdings are more likely to have been rented out on short-term leases. Slavery, therefore, played an important role in elite agriculture, but did not dominate it entirely. A few words should be written on the products of these farms. Xenophon’s tract largely focuses on cereal crops,³⁵ but a broader view of Athenian agriculture must include animal husbandry in its purview, which formed a symbiotic relationship with arable farming (Eupolis fr. 163 K-A: ‘fields and flocks and cattle’), representing continuity in agrarian practices from the archaic period discussed in Chapter 5.³⁶ Hodkinson has documented some of the ways in which animal husbandry was exploited by the Athenian elite to meet its needs for ready cash, including the rearing of sheep, whose wool was a notable product of the Athenian countryside (Antiphanes fr. 177 K-A).³⁷ In the same fragment, Antiphanes mentions honey, another famed product of Attica, especially the area around Hymettus.³⁸ Much of this produce, though, must have been destined for local markets, especially the large urban markets of Athens and Piraeus. At any rate, Osborne and Hodkinson are certainly correct to stress the fact that elite farming was a profit-oriented business.³⁹ Elite Athenians could not (in most cases) ignore their farms and be content with drawing on its income as indolent rentiers: the pressures put on them for ready cash via the liturgy system—and in particular for those who sought political advancement—meant that they had to continually think about how to

³⁴ Of the metics who assisted the democratic faction at Phyle, a number were farmers: IG II² 10 B 1, 9, 11, 13, 15; a keporos (‘market gardener’) is also listed at line 6. For the metics in the phialai inscriptions, individuals listed as georgos (‘farmer’) can be found at Meyer (2010) I 24; II-IXA 109, 248, 392; II-IXB 12, 60, 111, 208; XVI 22, 40; XX 69; ampelourgos (‘vine dresser’): II-IXA 486; II-IXB 52; zeugotrophos (‘keeper of a yoke of oxen’): XXVII 73–4. ³⁵ Xen. Oec. 16–18, with some attention to arboriculture in Oec. 19, but as Hodkinson (1988): 37 points out, the tract itself (Oec. 5.3) links animal husbandry to tillage. Cereals were certainly important in elite farming: one might note the relatively large farms in Atene studied by Lohmann (1992), one of which had two large threshing floors (Lohmann 1992: 49). ³⁶ On agro-pastoral farming, see Hodkinson (1988); on herders, see Roy (2012). ³⁷ Hodkinson (1988); see also Kron (2014). ³⁸ Attic honey appears among other products exported to Egypt in the third century: see Gabrielsen (2013): 72. On honey production, see also Moreno (2007): 66–8. ³⁹ Osborne (1991): Hodkinson (1992); cf. Bresson (2016): 135.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

176

Greek Slave Systems

maximize the income from their farms, which meant where possible specializing in lucrative cash crops and animal husbandry.⁴⁰ It is important to stress the small scale, in comparative terms, of elite farming in classical Attica. The largest known unitary plot of land in Attica, the estate of Phainippos, probably covered at least 110–120 ha, and elite farms were more usually made up of scattered parcels of land.⁴¹ (We know rather less, though, about the size of elite holdings on cleruchic land.⁴²) When we consider slaveholding and Athenian agriculture, then, we must not imagine a sharp step up from the slaveholdings of small farmers to those of the elite. Elite farmers had perhaps a handful more slaves involved in cultivation than the richest hoplite farmers, and it is highly likely that even male slaves on elite farms will have been used flexibly for a number of tasks, agricultural or otherwise. These elite Athenian farms were hardly plantations on either the Roman or the Atlantic model.

Textile Production Xenophon’s Oeconomicus provides good evidence for the domestic production of textiles by women, a major industry that was not confined to the elite but was practised in almost every household. Elite households, however, will have had a sufficient number of female slaves to produce not only enough fabric to clothe the immediate household (Xen. Oec. 7.36; 41), but also produce a surplus that could be stored and then sold at an opportune moment.⁴³ Xenophon’s image of the housewife as a queen bee (Oec. 7.32–4; cf. Democritus 68 F 227 D-K) is a very old one, and can be found in the work of the archaic poet Semonides of Amorgos (fr. 7 [West]). We can see here, therefore, continuity in practices from the early archaic period.

Income from Real Property In classical Athens, broadly speaking, only citizens could own real property. This meant that a certain proportion of the population (propertyless citizens, metics, and freedmen) had to rent housing from citizen landlords.⁴⁴ The metics alone may have constituted something like 10,000 adult males of ⁴⁰ A point made by E. M. Harris in Harris and Lewis (2015): 28. ⁴¹ Phainippos: Bresson (2016): 146–7. Elite landholdings: Davies (1981): 53–4; Foxhall (2007): 55–83. ⁴² Davies (1981): 55–60. ⁴³ On textile production for the market, see Tsakirgis (2015). For the iconography of textile production on Attic vases, see Vidale (2002): 325–474. ⁴⁴ Davies (1981): 49–52.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

177

military age in the late fourth century (reflecting a total metic population of perhaps c.33–46,000 individuals).⁴⁵ A good example of the rental properties in which most metics lived can be seen in Isaeus’ speech On the Estate of Philoctemon, where we find mention of a man named Euctemon who owned a tenement house (synoikia) which was managed by a former prostitute (Is. 6.19). The Euctemon example also highlights another kind of investment, though one that cannot have been all that common: the ownership of a bathhouse (Is. 6.33). These were far simpler affairs than their elaborate Roman successors, and several dating to the classical and Hellenistic periods have been excavated in Attica.⁴⁶ These generally followed a standard design: a tholos-style building (i.e. a kind of rotunda) containing about fifteen to twenty contiguously placed hip baths arranged in a circle; a bath attendant (balaneus) would ladle hot water over the customer from a furnace-heated tank (Theophr. Char. 9.8). The fact that the Old Oligarch complained that the poor were always to be found at the baths hints both that the fee must have been low and that demand must have been relatively widespread ([Xen.] Ath. Pol. 2.10).⁴⁷

Sweatshops (ergasteria) Athens’ burgeoning demand for a wide variety of commodities, paired with its commercial links to wider markets in the Aegean and beyond, created a profitable niche for the manufacture of all manner of goods. Our sources preserve evidence on various kinds of manufacturing enterprise.⁴⁸ We have already noted the knife and furniture workshops held by Demosthenes’ father and the shield manufactories of Cephalus and Pasion. One might also mention the shop of bronzesmiths (or carpenters) owned by Sophocles’ father (Vita. Soph. 1);⁴⁹ the aulos manufactory owned by Isocrates’ father (Dion. Hal. Isocr. 1 P. 534; [Plu.] Mor. 836e; Strattis fr. 3 K-A); Lysicles’ shop of shoemakers (Plu. Per. 24.6; schol. ad Ar. Eq. 739 and 740); Aeschines of Sphettus’ unguent-making shop (Lys. fr. 1 [Thalheim]); the tanneries of Cleon and Anthemion ⁴⁵ Hansen (1988): 10–11. ⁴⁶ See Lucore and Trümper (2013) for collected essays on the topic, up-to-date bibliography, and a catalogue by Trümper et al. of pre-Roman bathhouses in the Mediterranean. Euctemon’s bathhouse is possibly to be identified with the rock-cut bath in Piraeus listed at Trümper et al. (2013) no. 28. ⁴⁷ The fact that Diogenes the Cynic is reported to have visited the baths (Diog. Laert. 6.40; 6.47) hints that even the poorest could afford to attend them. The low, populist reputation of bathhouses is confirmed in Ar. Nub. 991; cf. Eq. 1060–2; 1397–401; Ar. fr. 59 K-A. We do possess entry prices for baths during the Hellenistic and Roman periods: Faucher and Redon (2014). ⁴⁸ Davies (1981): 41–5. ⁴⁹ Leocrates (Lyc. Leocr. 58) is said to have bought bronze workers; Davies (1981): 43 suggests that this means he owned a foundry.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

178

Greek Slave Systems

(Xen. Apol. 29); and the pottery of Cephalus of Collytus (schol. ad Ar. Eccl. 253). Xenophon mentions several well-known figures who made their money from similar enterprises: Nausycides of Cholargus, who owned a mill; Cyrebus, who owned a bakery; and Demeas of Collytus and Menon, both owners of cloak-making businesses. These shops were all manned by imported barbarian slave labour (Xen. Mem. 2.7.6: οὗτοι μὲν γὰρ ὠνούμενοι βαρβάρους ἀνθρώπους ἔχουσιν). In truth, though, these examples represent only the tip of the iceberg: a glance at the many commodities in circulation in classical Athens hints at a large and very diverse manufacturing sector.⁵⁰ The term ergasterion simply means ‘workplace’ or ‘workshop’, but we should distinguish between small workshops owned by poorer citizens or rented by metics, who might own a slave or two to assist them in their endeavours (see section III, pp. 180–1 below), and larger sweatshops staffed by slave labour.⁵¹ These specialized in the production of single commodities, and the managerial structure of these enterprises probably amounted to nothing more than a shop steward who oversaw production, the procurement of raw materials, and the sale of finished goods.⁵²

Mining and Quarrying To this day the landscape of the Lavreotiki in southern Attica remains pockmarked with the shallow shafts and spoil heaps associated with silver mining. Silver was Athens’ most famous and important export, and it has been estimated that at its apogee the mining district produced over 700 talents of silver per annum.⁵³ Since the Athenian state claimed ownership of everything below ground in this district, individual entrepreneurs would have to take out a lease sold by the poletai; a number of fourth-century inscriptions published ⁵⁰ For a preliminary list of commodities circulating in the Athenian economy, see Lewis (2015b): 381–98 (this only draws on the evidence of Old Comedy). For a very rich study of manufacturing drawing on the evidence of painted pottery, see Vidale (2002). For the economics of workshops in Athens, see Acton (2015). For the range of artisan occupations, see Harris (2002b). ⁵¹ We should also note a third category: slaves who dwelt on their own and ran small businesses or workshops, paying a regular apophora fee to their owners. From a wealthy Athenian’s perspective, these arrangements worked in a manner not unlike moneylending. If one had idle cash, one could lend it to an entrepreneur at interest, eventually getting back the principal, and regularly receiving interest payments. One could, alternatively, invest the money in buying a slave craftsman (or craftswoman): the slave would deliver regular apophora payments with minimal supervision and eventually pay for their freedom out of their savings, allowing the owner to recoup the initial capital outlay for the slave. See Kamen (2016). ⁵² Vertical specialization: see Xen. Cyr. 8.2.5, with Harris and Lewis (2015): 24–5. See also Kamen (2016): 415–16. ⁵³ Conophagos (1980): 138–52; 341–54.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

179

by this board of magistrates provide details on the various individuals involved in leasing mining concessions.⁵⁴ Osborne’s prosopographical study has shown that most of these lessees came from wealthy families of southern Attica, relatively close to the mining district itself.⁵⁵ Scholars widely agree that the mine labourers were slaves; estimates vary, but at the peak of production there could certainly not have been fewer than 11,000 slaves employed, and Siegfried Lauffer has suggested as many as 35,000.⁵⁶ The capital requirements for running a mine were high, for one would have to rent the mine, acquire the slaves, either rent or buy the processing plant and transport equipment, and pay for the slaves’ food and the charcoal needed for smelting; this could result in complex financing arrangements of the sort we see in Dem. 37 Against Pantaenetus.⁵⁷ In this case, the lessee was also the owner of the thirty slaves who worked the concession, though they were encumbered for two loans to raise capital for the enterprise. An alternative was to rent slave miners from an entrepreneur, and we know of businesses of this sort that grew on the back of the labour demands of mining enterprises: Xenophon (Poroi 4.14) writes of several such entrepreneurs who invested money in buying slaves, and then rented them to the lessees of mines.⁵⁸ Mining was potentially tremendously profitable, but risky; a concession could prove a dud, and the lessee could lose a great deal of money.⁵⁹

Tax Farming, Moneylending, and Banks In a society where liability to perform liturgies was based largely on visible assets, it made sense to convert a portion of one’s wealth to cash and then lend it out at interest: this kept part of one’s fortune out of the public eye, but at work generating revenue.⁶⁰ Some Athenians set up banks that made their money from investing the funds of depositors: these enterprises were generally run by talented and well-rewarded slaves such as Pasion and Phormion.⁶¹ Such enterprises were not, however, the usual way elite Athenians invested their money. Much more common were loans contracted on an individual basis and secured with real property; and it would appear that maritime trade ⁵⁴ Crosby (1950; 1957). ⁵⁵ Osborne (1985): 112–23. ⁵⁶ Lauffer (1979) tables 10 and 11; for an up-to-date view, see Rihll (2010). ⁵⁷ This is one of the most difficult speeches in the corpus of Attic oratory to understand. For an explication of the case, see Harris (2006): 163–206. For the economics of mining investments, Christesen (2003): 39–46 provides a superb overview. ⁵⁸ On the credibility of the figures for their slaveholdings, see n. 18 above. We have rather less information on the lessees of quarries: see Osborne (1985): 93–110. ⁵⁹ For choosing between investments (including mining), see Christesen (2003): 46–53. ⁶⁰ See Davies (1981): 64–6; Gabrielsen (1986); Cohen (1992): 191–206. ⁶¹ On banks, Cohen (1992) presents a modernizing account, whose several misreadings of the evidence are noted by Bogaert (1995). For a more accurate view, see Shipton (1997).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

180

Greek Slave Systems

was largely financed by high-risk, high-interest loans of this sort rather than by bank loans.⁶² When we find evidence for ‘portfolios’ of investments, money lent at interest is a standard feature.⁶³ Other occasional ventures in investing financial capital included tax farming, when private entrepreneurs would bid for the contract to collect a certain tax farmed out by the state, pocketing the difference between the amount paid for the contract and the amount collected in taxes. Since the initial investment might be beyond the means of a single individual, we find partnerships (koinoniai) in which groups of wealthy individuals pooled the capital for a given tax contract bid.⁶⁴

Slavery and Elite Wealth at Athens: An Overview Athens’ relatively well-developed and diverse economy enabled its elite to draw on a wider variety of revenue sources than their Spartan or Cretan counterparts. Our evidence shows that a typical rich man would not (so to speak) have all his eggs in one basket, but rather a portfolio of several kinds of investment. Various considerations influenced how he would structure his sources of income: profit versus risk was certainly a consideration;⁶⁵ another was ease of supervision, with the majority of those with investments in the Laurion mines living in close proximity to the mining district. Slavery was thoroughly enmeshed with most of these ventures, especially farming, mining, and the ownership of craft workshops. But the importance of money lent at interest as well as the rental of real property—interests in which slave labour played a less decisive role—should not be neglected. In sum, Athens certainly qualifies on the Finleyan definition as a ‘slave society’, but Athenian elites leant less completely on slave labour than their Spartan and Cretan contemporaries.

II I . SU B- E L I T E S LAV E O W N E R S H I P AT AT H E N S As we noted earlier, the relative prosperity of Attica’s citizenry, paired with low slave prices, opened up the option of slave ownership to a far broader swathe of the population than merely the elite. One cannot understand the importance of slavery to the Attic economy by focusing on the elite alone; we must, therefore, turn to average working Athenians to gauge the true scale and effect of slavery on Athenian society. To a certain degree, this is an uncontroversial point. There is little disagreement regarding the use of slaves in small ⁶² See Davies (1981): 60–2 for sources and discussion. ⁶³ See Davies (1981): 62–4 for sources and discussion. ⁶⁴ Davies (1981): 68; Harris (2006): 241–8. ⁶⁵ Christesen (2003): 46–53.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

181

workshops, and scholars have long agreed than in, for example, a workshop of about four to five workers, one would find an owner, perhaps the owner’s son learning his father’s trade, and a couple of slaves working side by side.⁶⁶ This pattern is most strikingly visible in the Erechtheum building inscriptions, where we can see petty craftsmen such as masons at work side by side with a couple of their own slaves.⁶⁷ (Of course, as Xen. Mem. 2.3.3 implies and Lys. 24.6 proves, not everyone could afford slave helpers for their workshops.) Far more controversy surrounds the use of slave labour by sub-elite farmers, a complex topic that now requires our consideration. The two most famous antagonists in the debate over the use of slaves in subelite farming are E. M. Wood and Michael Jameson, although the debate goes back further than the work of either of these scholars. Particularly noteworthy is A. H. M. Jones’ 1952 article on the economic basis of the Athenian democracy, in which he heavily downplayed the role of slavery in agriculture. It was Jones’ view against which Jameson reacted in his 1977–8 essay, which was in turn endorsed and strengthened in 1981 by G. E. M de Ste. Croix.⁶⁸ In Jameson’s opinion, Athenian small farmers were engaged in intensive cultivation of their plots and had high labour demands; they would probably have owned several slaves who could have helped them in a variety of tasks throughout the year as well as other, non-agricultural tasks where necessary. This view, termed the ‘maximalist’ position by Fisher,⁶⁹ does in fact fit rather well with our (somewhat meagre) sources, as we shall see. The so-called ‘maximalist’ views of Jameson and de Ste. Croix triggered a splenetic reaction from E. M. Wood, whose rejoinder, published in 1983 and later incorporated as part of her book Peasant-Citizen and Slave (1988), took these two scholars to task for suggesting that Athens’ citizen farmers— the backbone of its democracy—could have been involved in slaveholding.⁷⁰ Wood’s argument is complex and at several points convoluted, and it is not possible to reply here to every argument she marshals. However, it is worth expending some effort on this issue: despite a flurry of work in the late 1980s and 1990s,⁷¹ the debate has remained relatively dormant for the past twenty years, and has been left in an unsatisfactory state. Fisher’s useful summary of the controversy, published in 1993, identifies four rough positions on the matter: extreme minimalist (A. H. M. Jones); moderate minimalist (E. M. Wood; R. Sallares); moderate maximalist (R. Sinclair); and extreme maximalist (M. Jameson; Y. Garlan; G. E. M. de ⁶⁶ e.g. Mossé (1969): 90. ⁶⁷ See Randall (1953). ⁶⁸ De Ste. Croix (1981): 505–9. ⁶⁹ Fisher (1993): 41. ⁷⁰ Jameson (1992): 142 was politely understating Wood’s reaction when he described it as ‘little short of outrage’. ⁷¹ Sinclair (1988): 196–200; Sallares (1991): 55–7; Gallant (1991): 30–3; Jameson (1992); Fisher (1993): 37–47; Rosivach (1993); Osborne (1995); Ameling (1998). For a recent review, see Tordoff (2013): 16–33.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

182

Greek Slave Systems

Ste. Croix).⁷² Fisher’s summary of the debate is admirable and gentlemanly, but leaves the reader with a strong sense of agnosticism, as if the evidence is too inconclusive to rule out altogether any one of these positions. Yet this agnosticism is misplaced: Wood’s position largely rests on a combination of untenable assumptions and the dismissal or avoidance of much of the relevant evidence, and it seems rather odd to categorize it as ‘moderate’ and Jameson’s as ‘extreme’, for his position is advanced tentatively, assuming little, and paying close attention to all the relevant evidence. It is, therefore, worth revisiting this issue; for if it cannot be resolved with the degree of finality that one would like, we can at least eliminate a number of untenable arguments that have too long stood unchallenged.⁷³ Wood’s key arguments and assumptions can be summarized as follows. First, she discusses the Attic citizenry in terms of two strata: the rich and the peasantry. The latter, she insists, were subsistence farmers (and in general she presents them in a highly homogeneous manner, almost a uniform caste).⁷⁴ Second, she scoffs at the logic of intensification of production: for Wood, there is no point in investing in slave labour if one aimed only at subsistence (which is presumed, not demonstrated).⁷⁵ Third, she dismisses or ignores key pieces of evidence. For instance, she claims that the evidence of comedy is inadmissible because comic poets exaggerate, making their testimony untrustworthy.⁷⁶ Furthermore, she banishes from the discussion all passages that use the term oiketes, which she claims means ‘servant’ rather than ‘slave’. According to Wood, this term may include slaves but is not limited to them, and is at any rate too vague to constitute a diagnostic status marker, meaning that all passages that employ it must be left to one side.⁷⁷ As for the evidence of slave prices, she ignores them altogether.⁷⁸ Having swept most of the key ⁷² Fisher (1993): 40–2. Fisher (1993): 46 states his own preference for a moderate version of the maximalist position. ⁷³ Jameson did not deign to reply fully to Wood’s challenge (see Jameson 1992: 142–6 for a short reply), and this has fostered the mistaken idea among some (e.g. Scheidel 2008: 107) that the debate between the two is irresolvable. ⁷⁴ Wood (1988): 62–6. ⁷⁵ Wood (1988): 52–3. Cf. p. 63: ‘we should not underestimate how difficult it would have been for a peasant family, living always close to the margin . . . to solve its problems of subsistence by adding to the household yet more permanent and alien mouths to feed’. At p. 63 she dismisses the idea of farming for profit as an anachronistic application of a capitalist mentality to antiquity. The denial of profit-oriented behaviour is one of the more bizarre excesses of the primitivist approach to the ancient economy, and an idea that Wood pursued elsewhere in her work (especially in The Origin of Capitalism). How little it corresponds to the reality of Attic farming can be illustrated by a black-figure amphora from the sixth century, which shows oil being sampled prior to sale in an olive grove; the inscription reads ‘O father Zeus, I wish I could become rich!’ See Ehrenberg (1961): viii and plate Vc = Chatzidimitriou (2005) plate E 14. ⁷⁶ Wood (1988): 173–4. ⁷⁷ Wood (1988): 48–9. ⁷⁸ Rather ironically, indeed, since some of her more caustic denunciations of Jameson and de Ste. Croix (e.g. Wood (1988): 52–3; 65; 67; 72) accuse them of ignoring evidence. She claims (Wood (1988): 70) that de Ste. Croix holds a dogmatic, abstract view of the inherent profitability

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

183

evidence to one side, Wood is left with a blank canvas on which to paint a picture of the Attic countryside better attuned to her ideological agenda. Wood’s Attica is one populated by a mixture of rich farmers (who rely on some slaves, but mainly tenant farmers) and poor peasant farmers, almost none of whom are slave owners. Wood’s position is riddled with errors and false assumptions to the point that it must be categorically rejected. Some of her arguments can be easily countered: her insistence that the term oiketes means ‘servant’, which might include (without being restricted to) slaves, is an obvious error, and has been roundly rejected by specialists.⁷⁹ But three aspects of her argument require closer attention: (i) her wholesale dismissal of the comic evidence; (ii) her division of the citizenry into the categories of ‘rich landowner’ and ‘peasant farmer’, the latter of which is presented as a more-or-less homogeneous stratum operating at bare subsistence level; and (iii) the notion that slaveholding and the intensification of cultivation made no economic sense for Athenian small farmers. We will deal with each of these in turn.

Taking Comedy Seriously: Slavery and Agriculture in the Comic Poets It is revealing that proponents of the ‘minimalist’ view simply dismiss out of hand the evidence of comedy. Their reason for doing so is obvious: the picture it paints is at serious odds with their views. Since comedy is full of exaggeration, so the argument goes, we cannot trust it as evidence of reality in any sense and are best served by ignoring it altogether.⁸⁰ This is not entirely out of tune with trends in scholarship over the past few decades. As several recent studies have pointed out, modern scholars have been (on the whole) rather reticent when it comes to utilizing comedy as

of slavery, but, as he rightly points out (quoted, indeed, by Wood on p. 65), this profitability is a function of the price of slaves at Athens, ‘the lowness of which is astonishing, in comparison with what is known of slave prices in other societies’ (de Ste. Croix (1981): 40; my emphasis). All that Wood has to say on slave prices is that ‘the evidence concerning slave prices drawn from the Attic stelai is severely limited’ (Wood (1988): 193 n. 57). But the Attic stelai hardly constitute our only source of data on this topic: Pritchett and Pippin (1956): 277 provide a list of nonepigraphic references. Wood, remarkably, does not cite this key study in her book. ⁷⁹ Wood (1988): 49 also claims that the term therapon means ‘servant’, but as Osborne (1995): 40 rightly notes, ‘I do not find any clear evidence for thinking that in the classical period oiketes (or therapon) could be used of a free man.’ Cf. Jameson (1992): 142, (2002): 168 and Foxhall (2007): 74, who make the same point. In the Appendix to this book I show through a systematic study of all attestations of the term oiketes in Greek of the classical period that the term is never applied to a free servant. ⁸⁰ Jones (1952): 20; Wood (1988): 173–5; Ameling (1998): 286.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

184

Greek Slave Systems

evidence for social history.⁸¹ De Ste. Croix’s insistence that Aristophanes was an old ‘Cimonian’ conservative who peddled elite views has long troubled historians, who, quite naturally, gravitate towards less overtly partisan texts and more straightforwardly ‘historical’ genres as their first port of call when investigating this or that issue.⁸² To a large extent this view is misplaced, and scholars such as Kron and Canevaro have rightly emphasized the populist nature of comedy as a genre (for one thing, Aristophanes can hardly have enjoyed the success that he did if he were merely a spokesperson for elite snobbery).⁸³ But even if we accept this point, it does not justify throwing caution to the winds and treating comedy as a straightforward reflection of reality. Like other literary genres, it may contain important historical information; but this has to be teased out with care. For the economic historian, the greatest contribution of comedy to our understanding of realia lies in the mass of incidental background details, the canvas of real Athenian society onto which Aristophanes and his contemporaries paint their (often absurd and exaggerated) plots and characters. Nobody, of course, would suggest that the Athenians hurtled around Attica on the backs of oversized dung beetles; even so, discerning what counts as incidental detail and what counts as exaggerated embroidery is not a straightforward matter. However, it seems preferable to make the effort to grapple with these problems rather than simply hedging our bets and leaving comedy to the literary scholars. The potential rewards, indeed, far outweigh the risks involved. For now, I shall simply set out the picture portrayed by the comic sources.⁸⁴ In Aristophanes’ plays the protagonists are usually average citizens, often farmers; Table 8.1 (based on Lévy’s 1974 study) presents the evidence for slaveholding among these characters.⁸⁵ We should take special note of the farmers. As de Ste. Croix points out: ‘In the Plutus, Chremylus the farmer, who is specifically described as a πένης (line 29) and is one of the τοῦ πονεῖν ἐρασταί of line 254, owns several slaves (lines 26, 1105).’⁸⁶ To this we may add several other examples.⁸⁷ From the Acharnians we have Dicaeopolis, who had farmed in the deme of Cholleidae (Ach. 406; cf. 28–36) before being been forced to move behind the long walls because of the Spartan invasion (Ach. 33, 71–2), yet still has several slaves:

⁸¹ Moreno and Lape (2014): 337; Pritchard (2012): 14–18. An exception to this trend: Vlassopoulos (2007b). ⁸² See Pritchard (2012) for a critique. ⁸³ Kron (1996): 69–80; Canevaro (2016b). ⁸⁴ Due to matters of space I will restrict my attention to Aristophanes and the late fifth and early fourth centuries, and leave to one side for now the evidence of Menander. ⁸⁵ Lévy (1974): 33. I have left out his gradations of ‘riche’, ‘paysan moyen’, and ‘paysan’. Instead of categorizing these characters according to etic gradations of wealth, it is better to pay closer attention to the description of their individual situations in the play in question: see below. ⁸⁶ De Ste. Croix (1981): 505. ⁸⁷ See Lévy (1974): 32–3 for a full list of references.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

185

Table 8.1 Aristophanic Slaveholders Play

Date

Master

No. of slaves

Male

Female

Acharnians Knights Clouds Wasps Peace Peace Frogs Ekklesiazusai Wealth

427 426 425 422 421 421 405 392 388

Dicaeopolis Demos Strepsiades Bdelycleon Trygaeus The chorists Hades Chremes Chremylus

4 4 2  7 or 8 2 3  5, 6, 7 or 8 3 3

2 4 1 6 2 1  4, 5, 6 or 7 2 1

2 not mentioned not mentioned  1 or 2 not mentioned 2 1 1 not mentioned

Xanthias (Ach. 243) and several unnamed slaves (Ach. 259, 1003). Then there is Strepsiades in the Clouds, who is an autourgos (Nub. 43–50; 70–2; 104–6), though wealthy enough to marry into a distinguished family of town dwellers: he owns a number of slaves (Nub. 7), one of whom is named as Xanthias at line 1485. Chremes in the Ecclesiazusae cultivates a vineyard (Eccl. 817–22) and owns several slaves: Sicon (Eccl. 867), Parmenon (Eccl. 868), and one other (Eccl. 1113). Trygaeus in the Peace also owns a vineyard some miles out of town (at Athmonia: Pax 190, 919), and owns a couple of slaves whose discussion opens the play. And there is the chorus in the Peace: as Lévy points out, the chorus is made up of farmers who also seem to own a couple of slaves (Pax 1138, 1146).⁸⁸ Lévy concludes: ‘[a]insi le théâtre d’Aristophane suggère qu’un paysan moyen possédait en général au moins trois esclaves, ce qui confirme l’importance des esclaves à Athènes’.⁸⁹ Wood claims that we cannot trust this evidence; but is this picture of slavery in sub-elite agriculture exaggerated for comic effect or is it merely a (realistic) background detail? Prima facie, one might suppose the latter, for the simple reason that there is no obvious dramatic advantage to be gained by misrepresenting slaveless subsistence farmers as slave-owning smallholders.⁹⁰ Aristophanes, it is generally assumed, represents run-of-the mill Athenians to appeal to a broad support base; but he would have failed in this had the average citizen farmer been a slaveless peasant of the E. M. Wood variety, who would have had little in common with Aristophanes’ characters. The best we can do in defence of Aristophanes’ verisimilitude in this respect is to demonstrate that the picture he presents is both historically credible and economically sound. This takes us to the second of Wood’s main assumptions, the idea that ‘peasant’ farmers were a relatively uniform caste of subsistence farmers for whom slave labour was an unnecessary and illogical investment. ⁸⁸ Lévy (1974): 32. ⁸⁹ Lévy (1974): 34. ⁹⁰ For a useful approach to sifting reality from dramatic exaggeration in relation to presentations of agriculture in Greek plays, see Roy (1996).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

186

Greek Slave Systems

Sub-Elite Farmers: Subsistence Peasants or Prosperous Smallholders? Wood’s views on this point need to be assessed in light of a broader consideration of the distribution of landed wealth in Attica. Here, it may be useful to sketch the kind of view of Attic land distribution that could, potentially at least, support some version of Wood’s picture. A good illustration of the difficulty of this problem can be seen if we consider one influential reconstruction of land tenure in classical Attica, that of Foxhall.⁹¹ Foxhall’s task begins by estimating the amount of cultivable land in Attica, which, if we take into consideration pasturage and marginal land, she estimates at ‘something over 1000 km²’.⁹² Since the Attic countryside is still there for us to study and measure, this estimate cannot be too far off the mark. She then proceeds to consider the basic subsistence plot, which (with some judicious caveats) she considers to have been around 5.5 ha. With these figures in mind, she proceeds to divide up the Attic countryside from known population figures, based on a figure of 22,000 citizen households in the late fifth century.⁹³ All of these numbers are of the right order of magnitude. However, it is important to emphasize that Foxhall’s reconstruction of Attic land tenure is, nonetheless, largely speculative.⁹⁴ The key problem here is the ratio between those citizen households whose livelihood came from farming and those working in the non-agricultural sector. If we assume that Athens was a dynamic commercial centre with a sizeable non-agricultural population, we can assign a not insignificant portion of these 22,000 citizen households to the non-agricultural sector, leaving more land to divide up between fewer farming households. On the other hand, if we assume that the non-agricultural citizen population was relatively small, then we need to squeeze almost all of these 22,000 households into the Attic countryside. How do we choose between these approaches? How many citizen households should we apportion to the agrarian sector? Here, only one contentious piece of evidence appears to provide guidance: a fragment of Lysias preserved by Dionysius of Halicarnassus (Dion. Hal. Lys. 32). Dionysius informs us that in 403 BC there was an attempt to restrict citizenship to those who owned land (μὴ πᾶσιν ἀλλὰ τοῖς[τὴν] γῆν ἔχουσι), which, if implemented, would have led ⁹¹ Foxhall (1992); cf. (2002). However, one must be careful not to conflate Foxhall’s view of slavery in agriculture with that of Wood, of whom Foxhall is rightly critical: see Foxhall (1992): 155. ⁹² Foxhall (1992): 156. ⁹³ Note that this figure does not reflect the full number of male citizens, but rather households in which citizen sons had not yet inherited their fathers’ property. ⁹⁴ As Kron (2011): 129 n. 1 has rightly pointed out. Rose (2012): 211 n. 27 and 240–2 appears to miss the essentially speculative nature of Foxhall’s reconstruction, and leans a good deal of weight on it as a factual description of agrarian conditions in the late fifth century.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

187

to some 5,000 men losing their citizen status. Several scholars (including Foxhall) have supposed that these 5,000 men represent the non-agricultural component of the citizen body, and that ‘land’ (γῆ) here means farmland. In other words, nearly 80 per cent of Athenian citizens were farmers. But the issue is not so clear-cut. For one thing (and as Foxhall notices), we know of citizens such as Demosthenes who were wealthy but did not own any farmland; his fortune came from exploiting slave labour in workshops. If the move to disenfranchise landless Athenians were really a move to disenfranchise non-farmers, men like Demosthenes would have been disenfranchised as collateral damage. This is hard to accept, especially since recent work has exploded the idea that there was a kind of land-based aristocratic snobbery in Athens and all rich men who made money from non-agricultural pursuits were nouveaux riches.⁹⁵ As a corollary of her view that nearly all citizens were farmers, Foxhall tends to assume that all of the wealthiest Athenians had large farms and little else, but what we know of individual fortunes, as we have seen in section II of this chapter (pp. 172–80), suggests a diverse portfolio of investments (to be sure, one in which farmland normally occupied an important place; but this is hardly a straightforward picture of a landed aristocracy).⁹⁶ Thus, her basic assumptions tend to inflate the amount of land in the hands of the elite and to squeeze more citizen families onto the countryside than we can be reasonably sure were working there. The resulting picture is thus heavily conditioned by the initial primitivist assumptions of its author: that citizen farmers were relatively many and poor (with a large number of farmers owning under 5 ha of land); that the elite derived almost all of its wealth from farming; and that the non-agrarian sector must have been relatively tiny. Foxhall’s conclusions thus grow out of her premises, not the evidence. What if the 5,000 citizens who would have been disenfranchised were not non-farmers, but rather citizens who did not own any land (i.e. property, including houses and gardens) at all?⁹⁷ If this is the case, then Phormisius’ proposal, like most oligarchic approaches to citizenship, would have aimed to reserve the citizen franchise for those of sufficient wealth (assessed by the straightforward index of ownership of real property, rather than the cumbersome process of converting assets into cash equivalents) and was not a curious attempt to turn Athens into a society of farmer-citizens as Foxhall supposes, but merely an attempt to amputate the poorest (and most virulently democratic) part of the demos from the body politic. The balance of probability—and

⁹⁵ See the careful study of Kron (1996). ⁹⁶ Foxhall (2007), however, does bring out the diversity of elite income sources more fully than her 1992 study. ⁹⁷ Cf. Kron (2011): 131 on the term γῆ, ‘which I interpret to include garden or house plots, and certainly not just agricultural land.’ For poor artisans having to rent housing, see Xen. Symp. 4.4.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

188

Greek Slave Systems

the neutral term used for land—favours this interpretation. At any rate, we can see that the task of estimating the size of the agrarian sector, which seems a rather simple affair in Foxhall’s model, is in fact extremely complex once we realize that γῆ in Phormisius’ proposal probably does not mean ‘farmland’.⁹⁸ There are other reasons to doubt the legitimacy of any approach that seeks to squeeze a figure as high as Foxhall’s 80 per cent of citizens into the farming sector. Even if we assumed, pessimistically, that on average most of these could just about make ends meet (and thus feed themselves without producing any surplus for market), we would then be hard pressed to explain away the level of grain imports into Attica attested in our sources, which make clear both that there were many non-farmers in Attica and that not all farmers had grain production for domestic consumption as their chief priority.⁹⁹ Among the implications of this approach is the notion that the Attic countryside may have been largely divided up among elite farms and, below them, a relatively prosperous (though far from homogeneous) stratum of reasonably affluent, mainly hoplite-class, citizens.¹⁰⁰ As M. H. Hansen notes, ‘the typical Athenian citizen was not the subsistence farmer who tended to stay in his deme of origin, but was rather like Chremes whom Aristophanes describes as a citizen who attends the meetings of the ekklesia and, on other occasions, carries his wine to the market in order to buy flour’.¹⁰¹ To be sure, these farmers were heterogeneous in terms of their position on the wealth spectrum: some, like Strepsiades—whilst remaining an autourgos—were on the cusp of the liturgical class; many, if not most, were middling hoplite farmers; others were rather poorer.¹⁰² But they were certainly not the enormous homogeneous caste of subsistence farmers ‘living always close to the margin’ portrayed by Wood.

⁹⁸ Maeve McHugh (per. comm.) points out to me that if farmland were intended, there would have been other, more likely words to choose from, e.g. agros or khorion. See further Langdon (1991). ⁹⁹ According to Bresson (2016): 411 this amounted to some 75 per cent of all grain consumed in Attica during the fourth century, and he points out that the dependence on imports would have been more extreme in the fifth century (cf. Hansen 2006b: 44–5). This picture is difficult to reconcile with a view that holds most citizen households to have been self-sufficient farms, and the non-agricultural population comparatively small. Cf. Hansen (1988): 7–13. ¹⁰⁰ The question of agricultural productivity is highly complex; for the present purposes it will suffice to say that I find the position of Hodkinson (1988), who argues for a mixed-farming ‘agropastoral’ model, the most convincing. For a sensible overview, see Gallego (2007). On this approach, a farm of c.5 ha would require intensive cultivation but normally generate enough surplus to sell some produce on the market. Useful points on method and the debate on productivity also in Kron (2008; 2017), though focusing mainly on Roman Italy. ¹⁰¹ Hansen (1988): 13. ¹⁰² Cf. Jameson (1992): 145. My picture is in accord with the recent study of Guía and Gallego (2010) on the zeugitai, whom they treat as roughly commensurate with the hoplite class during the fifth century, farmers who owned at least 5 ha of land each. Pages 258–65 analyse ‘The zeugitai of the fifth century as the bulk of the Attic peasantry’.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

189

Markets, Intensive Farming, and the Acquisition of Slaves That leaves us with the matter of intensification and the economic rationality of slave ownership for small farmers. First, we should note that farmers all over Attica were familiar with market exchange, which emerges from a close reading of the comic sources, though it has been denied by some of a primitivist bent. For the latter approach, one might note the view of Finley, who pointed to a passage in the Acharnians (33–6) in which Dicaeopolis complains about the ongoing war and wishes he could return to his rural idyll where nobody knew the words ‘buy’ and ‘sell’ as evidence of the lack of market exchange in the countryside (cf. Ar. fr. 111 K-A; fr. 305 K-A).¹⁰³ But this idyllic picture draws on the Golden Age imagery of myth; as E. M. Harris points out, if one reads the rest of the Acharnians, we find that ‘Dikaiopolis does not want peace because he wants to retreat to the isolation of his autarkic homestead. Dikaiopolis makes his treaty with the enemy to reap the benefits of trade with other poleis.’¹⁰⁴ Indeed, we later find him meeting a Boeotian trader and revelling in the opportunity to buy foreign treats, especially Copaic eels (Ach. 874–90). One might compare another passage, here from Aristophanes’ Islands: You fool, you fool! All of it’s in this life of peace: to live in the country on his small plot of land, free of the rat-race of the market, owning his very own yoke of oxen, and hearing the bleating of his flocks and the sound of new wine being bottled up, snacking on little finches and thrushes, no hanging around the market waiting for smallfry days old, overpriced, weighed out for him by a crooked fishmonger with a thumb on the scales.¹⁰⁵

This nostalgia-infused portrait strikes a contrast between the small farmer¹⁰⁶ and the crooked wheeler-dealer of the marketplace (cf. Hyp. 3.26 for a similar contrast). But as with any exercise in portraying alterity, we should be wary of taking exaggerated contrasts—which tend to amplify dissonant features and play down similarities—too literally.¹⁰⁷ An important corrective is provided by those passages that do not explicitly deal with idealized town/country contrasts but merely convey incidental details. For instance, in the Wasps we find a farmer heading into town to sell his donkey and panniers at the market that took place on the first of the

¹⁰³ Finley (1999): 107. Cf. Gallant (1991): 100–1. ¹⁰⁴ Harris (2002b): 78. Moreno (2007): 74–5 makes the same point. ¹⁰⁵ Ar. fr. 402 K-A, tr. Henderson. ¹⁰⁶ Very much of the sort that Hodkinson (1988) has cogently argued to be common, practising a mixed-farming regime of cereals (hence the plough oxen), animal husbandry, and viticulture, as well as a little bit of trapping. ¹⁰⁷ Jones (2004) tends towards a view of division and hostility between town and country. For cogent criticisms, see Roy (2005).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

190

Greek Slave Systems

month (Vesp. 169–71);¹⁰⁸ in the Ecclesiazusae another farmer, Chremes, describes how he sold some of his grapes for cash and was aiming to use the money to buy flour (Eccl. 817–22). In the Peace a farmer talks about returning to the fields after buying some tarichos, salt fish (Pax 563); conversely, in a fragment of Antiphanes (fr. 69 K-A) we read of an itinerant fishmonger who visits the countryside selling sprats and red mullet.¹⁰⁹ One might compare Theophrastus’ caricature of the stereotypical country bumpkin (agroikos) from the late fourth century (Char. 4.15). What is remarkable is that even Theophrastus’ exaggerated caricature of the most unpolished rustic does not paint him as a figure isolated from the market: he goes into town to buy salt fish, get a haircut, mend his shoes, and visit the baths. The joke, rather, is that he views these visits to the town and its amenities as rare treats, whereas they are utterly humdrum to the town dweller. So we should not assume that the Attic countryside was an autarkic idyll free of market exchange based on a couple of wistful passages of Aristophanes where this or that character sought to get away from the town and seek the bucolic isolation of the chora. Aristophanes was no friend of the market huckster, and he sometimes exaggerates the peace and isolation of the countryside to underscore the vulgar lifestyle of the agoraios ochlos with its hawking, swindling, and interminable pursuit of profit.¹¹⁰ Other genres of evidence confirm and flesh out this picture. Recent archaeological finds, combined with epigraphic evidence, demonstrate the existence of over a dozen agorae scattered across Attica.¹¹¹ To this we may add the evidence of survey archaeology, which, though patchy in coverage of the Attic landscape, has revealed a mixture of scattered, reasonably sized farmsteads as well as nucleated settlements¹¹² connected by dense networks of

¹⁰⁸ On this monthly market, see Lewis (2015a): 324–5. Cf. Sobak (2015): 685–7. ¹⁰⁹ An image of this kind of activity can be found on an Attic red-figure pelike c.480–470 BC. On the front is an image of a fisherman catching fish with a line, while a boy stands by with a carrying pole and basket: see Villanueva-Puig (1992): 75. On the reverse the boy, his carrying pole loaded with two baskets of fish, trots off to sell them: see Villanueva-Puig (1992): 82. For a similar image, see Pipili (2000): 171. Cf. Arist. Rhet. 1365a25–6. ¹¹⁰ I think these contrasting pictures—the fanciful ‘market-free’ idyll and the more realistic picture one can construct based on incidental passages—demonstrate why we should not follow Ameling (1998): 286 and simply reject any information on the countryside in Aristophanes as utopian and unrealistic. ¹¹¹ Kakavogianni and Anetakis (2012); Jones (2004): 86; Harris and Lewis (2015): 13. ¹¹² There is a large body of work on settlement patterns across Attica. Osborne (1985) argues that the demes conformed to the nucleated settlement pattern model. Subsequently published survey data for certain parts of Attica has revealed, however, dispersed settlement patterns with numerous individual farmsteads, e.g. Lohmann (1992) for the deme of Atene. Fachard (2013): 94–5 suggests a predominantly nucleated settlement pattern combined with a scattering of dispersed farmsteads for the area between Eleutherae and Oenoe. Kellogg (2013) suggests a dispersed settlement pattern for Acharnae. Evidently, any overall answer must involve a mixture

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

191

roads,¹¹³ many accessible to carts, enabling easy communication between farmers and markets. But what about the transport of their goods to market? It was once common to emphasize the high costs of overland transport in antiquity, a point that remains relevant if we are to restrict ourselves to discussing the transport of marble blocks across the countryside in hired wagons.¹¹⁴ But as Colin Adams has shown in relation to Roman Egypt, numerous inexpensive options existed whereby farmers could transport their products to market, and many of these methods were also used in classical Attica.¹¹⁵ We might note the ownership of small carts, carrying poles, donkeys and mules with panniers, as well as the droving of livestock.¹¹⁶ In other words, Attic farmers had the resources, roads, and technology for transporting their surplus products to market with relative ease.¹¹⁷ Much of the preceding discussion has dealt with general problems that arise from the literary sources, but it is worth emphasizing that Attica represents a far from uniform landscape. We are dealing with a combination of a highly textured topography, uneven distribution of cultivation niches (e.g. good grain-growing land in the plains, areas better suited to pasturage in the various uplands, uneven forest cover), and irregular clusters of population exerting variable levels of demand for a variety of products.¹¹⁸ We might add to this the effect of demographic shifts (especially through war) during the classical era and their attendant effects on the exploitation of the countryside. This should remind us that although any assessment of the issue of slave labour in sub-elite agriculture deriving from the literary sources is bound to be fairly general in nature, in practice this must have mapped onto the classical Attic countryside in a far more complex manner, and shifted over time.

of both patterns of settlement type; but further surveys will enable firmer answers to be formulated and micro-regional variations to be plotted. McHugh (2017) contains useful discussion of the issue. ¹¹³ On roads in Attica, see the essays in Goette (2002) and Korres (2009); cf. Fachard and Pirisino (2015), who discuss the use of roads for commercial purposes and the interconnectivity of economic micro-niches in the Attic landscape. See also McHugh (2017): 143–8. ¹¹⁴ On which, see Goette (2010). ¹¹⁵ Adams (2007). ¹¹⁶ On carts in ancient Greece, see Crouwel (1992): 77–98; Raepsaet (2002). A cart was deemed essential to farmers by Hesiod, who offers advice on how to construct one (Op. 426), but by the classical period there were specialized cartwrights/wheelwrights (Ar. Pl. 513; Eq. 464) who could produce these vehicles. For cheap carts, see Theophr. Hist.Pl. 5.7.6. For an overview of land transport in ancient Greece, see Chatzidimitriou (2010). ¹¹⁷ Moreno (2007): 37–76 argues that the Attic demes were heavily involved in cash-crop cultivation for the Attic market. Although sympathetic to his argument, I would temper the degree to which Moreno pushes his model, as it depends on an overly pessimistic (essentially primitivist) view of grain cultivation, whilst his rejection of the agro-pastoral model (p. 18) is far too perfunctory. ¹¹⁸ For an overview of some of these issues, see Fachard and Pirisino (2015).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

192

Greek Slave Systems

Slavery in Sub-Elite Farming: An Overview Let us sum up and take stock of the evidence discussed above. First, and most obviously, when it comes to sub-elite farming in Attica we are not dealing with a uniform stratum of subsistence peasants living on the breadline; whatever our reconstruction, we are dealing with a sliding scale, with some farmers (like Strepsiades in the Clouds) owning enough resources to place them on the cusp of the liturgical class, and others with considerably less to their name. Second, as Jameson has pointed out, the very image of the near-starving subsistence farmer is a conceit of modern scholars and cannot be found in our Attic sources, which does not mean that they never existed, but rather that we should not suppose a priori that a large swathe of the Athenian citizenry were farmers of this type.¹¹⁹ Aristophanes’ farmers may gripe about poverty, but we must bear in mind the semantic differences between our own and the Athenians’ ideas of poverty: one cannot equate them aprioristically. Third, we have seen that there is no need to squeeze nearly 80 per cent of Athenian citizens into the agrarian sector and suppose that a large number of these lived below comfortable subsistence levels.¹²⁰ With a large non-agrarian population to feed, the production of surplus food destined for the market was an obvious and attractive goal for small farmers, even if production for their own subsistence was their chief objective.¹²¹ The fact that most Attic farmers were connected to markets (and thus plugged into the food demands of the urban sphere) has implications that entirely overturn Wood’s view of the ‘subsistence only’ objectives of small farmers. There was a clear rationale behind intensifying exploitation of the land, namely to produce products to sell for a profit. This required a modest amount of extra labour. As the passages from comedy just cited show, farmers had access to money and markets, which meant that they could potentially buy slaves and increase the productivity of their farms; and we have already noted the relatively prosperous position of many farmers and the relatively ¹¹⁹ Jameson (1992): 143–4. ¹²⁰ Cf. the admirable review of the problem by Tordoff (2013): 16–33. My one point of disagreement is that Tordoff supposes that subsistence farmers living on the breadline represented a large swathe of Attica’s farmers, but this is based on supposition, not on evidence. He is perfectly correct to state that the sort of farmer envisaged by Gallant (the owner of about 3 hectares of land, cultivated with primitive agricultural techniques) would probably not have owned a slave; but the crux of the problem lies in how typical such a farmer was. ¹²¹ Thorner (1971): 207 (cited in Gallego 2007: 9) rightly points out: we are sure to go astray if we try to conceive of peasant economies as exclusively ‘subsistence’ oriented and suspect capitalism whenever the peasants show evidence of being ‘market’ oriented. It is much sounder to take it for granted, as a starting point, that for ages peasant economies have had a double orientation towards both. Cf. Bresson (2016): 199–203; McHugh (2017): 138–43. On market gardening, see Amigues (2007); Bresson (2016): 130.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

193

low price of slaves in Attica. Seen in this light, the farmers of Old Comedy, who own a slave or two to help out on the farm and with a variety of other tasks, are perfectly credible. One further remark is worth making in terms of the implications of this picture for the workings of the democracy. It is clear that nobody believes (or should believe) in the demos as an idle mob that delegated all its manual tasks to slaves and spent its time loitering in the agora and Pnyx.¹²² The vast majority of Athenian slave owners had hard working lives themselves. That said, it is worth looking at the acquisition of slaves by small farmers not just in terms of needs, but of desires.¹²³ Ownership of a slave or two will not have made an Athenian small farmer a man of leisure; but it would have injected a high degree of flexibility into his daily life. He may not have attended every festival or every assembly session (and this, as is often pointed out, was especially true of those living in the more distant demes); but slave ownership would have enabled small farmers to participate in civic life to a far greater extent than could have been the case had they made do with just their own and their family’s labour. Indeed, Hansen has shown that the democratic institutions could not have functioned were the citizenry as a whole not highly engaged in its running, which leaves no room for a farming class that largely stayed out of politics due to a lack of time and leisure.¹²⁴ Slavery thus entailed significant lifestyle benefits—including political engagement—for average Athenians.¹²⁵ * * * Let us finish with this study of Athens with a few words by way of conclusion. Athens provides a useful counterpoint to Sparta, Crete, and the archaic worlds of Homer and Hesiod insofar as it shows how market exchange transformed the role of slavery in those Greek societies that were, by the classical period, particularly engaged in trade and commerce. The two points I have endeavoured to emphasize here—the greater access to slaves that markets provided for poorer citizens and the wider range of profitable enterprises beyond the direct exploitation of slave labour available to the elite—will have obtained in most outward-looking commerical poleis of the classical and Hellenistic

¹²² This is the thrust of Jones (1952) and, in this limited sense, Wood is right to follow him. ¹²³ In an important contribution to the debate, Osborne (1995): 33 views the problem from the vantage point of labour needs, concluding (convincingly) that a peasant farmer could just about get by at harvest time with the help of his family and neighbours. This provides a useful theoretical baseline. It is worth noting, though, that even if Attic small farmers could get by without slave labour, this does not necessarily mean they usually did so. A similar logic to Osborne’s could prove that most British families can get around the country by public transport; but, as we know, most households own a car, which in addition to its main function also brings benefits in terms of status and flexibility. On desire over need as an element of demand in Greek economies, see Davidson (1997); Foxhall (1998); van Alfen (2015). ¹²⁴ Hansen (1985); (1988); (2006b): 19–60. ¹²⁵ Cf. Fisher (1993): 46–7.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

194

Greek Slave Systems

Greek world. That is not to say that Athenian slavery should be seen as an ‘archetype’ that was reproduced in carbon copy across the Greek world: Corinth and Aegina, for instance, were renowned for their large slave populations, but had far less farmland than their Attic counterparts, and no silver mines. Conversely, Thucydides (8.40.2) noted Chios’ massive slave population, and J. K. Davies and Paul Cartledge are surely right to suppose that this was heavily involved in agriculture, especially the production of Chios’ famous wine.¹²⁶ Just as ‘helotic’ slavery stands as a loose category grouping together a diverse range of slave systems due to the shared feature of linguistic uniformity, esclavage marchandise too should not be viewed in an overly schematic fashion. Regional diversity was the rule. * * * To conclude Part II of this book, I offer one final note on the unity and diversity of Greek slavery. We have focused so far on regional case studies, an approach that is practicable only for those Greek slave systems whose details survive in relatively large quantities. We might, however, cast our net more broadly in relation to one issue, i.e. the offences of slaves; for here we have more evidence, widely scattered though it is. Earlier in this book we noted an Athenian inscription (IG II² 1362 = LSCG 37) that mentions the rules set out by the priest of Apollo Erithaseos on taking wood from the god’s shrine; recall that a slave caught so doing would be given fifty lashes of the whip, and his master reported to the archon basileus.¹²⁷ (A free offender, on the other hand, was also reported to the archon basileus, but had only to pay a fine of 50 drachmas.) To this one might compare the rules relating to another sanctuary of Apollo, that of Apollo Koropaios in Thessaly, which are set out in an inscription dating to the late third century BC (IG IX 2 no. 1109 = LSCG 84). This lex sacra forbids individuals from bringing animals into the sanctuary; the free person caught so doing faced a fifty-drachma fine, half of which went to the person who denounced the offender. The slave offender, though, faced a hundred lashes of the whip, and a fine of 1 obol per animal.¹²⁸ Another third-century sacred regulation deals with a similar issue, viz. fly-tipping in the sanctuaries of Dionysus and Leto on the island of Delos (SEG 23:498 = LSCG Suppl. 53). The dumping of kopros and ashes in the sanctuary was forbidden; offenders who were caught were reported to the boule, with slaves receiving fifty lashes and being placed in the stocks, and free offenders receiving a ten-drachma fine, half of which went to the informer. In the famous lex sacra from Andania in Messenia dating to the following century (LSCG 65) we find a similar scenario.¹²⁹ Here, the slave who committed ¹²⁶ Davies (1981): 46; Cartledge (1985): 35–6; see also Bresson (2016): 126–7. ¹²⁷ See Chapter 1, p. 45 n. 69. ¹²⁸ Chandezon (2003) no. 19 and p. 417, suggesting the slave paid out of his ‘peculium’. See also Roy (2012): 223. ¹²⁹ On this document, see Gawlinski (2012).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Classical Attica

195

theft would be brought before officials and, if found guilty, faced a beating; he was furthermore compelled to pay twice the value of the stolen item (ll. 75–7). Payment had to be made immediately, or else the owner would hand his slave over to the victim of the offence to work off the obligation as a debt bondsman (ll. 77–8). If the owner did not do this, he was liable for double the original fine. If someone were caught chopping wood in the sanctuary, he would be beaten if a slave, but fined if a free person; as in the Thessalian and Delian examples noted above, half the fine would go to the person who reported the offender to the sanctuary’s officials (ll. 79–80). The Attalid astynomoi inscription from Pergamum (SEG 13:521) also deals with the offences of slaves, in this case in relation to public fountains.¹³⁰ The law states that it is forbidden to water animals or wash clothing or vessels in the fountains: a free person who so acted had their animal, clothing, or vessel confiscated; moreover he faced a fine of 50 drachmas. For slaves, the penalty varied depending on whether he were acting on his owner’s instructions or not. If he were, then the item in question would be confiscated and the slave given fifty lashes of the whip; if he acted on his own initiative, though, confiscation would occur as before, but the slave would face a hundred lashes, find himself shackled for ten days, and given a further fifty lashes when released. These examples, in toto, neatly capture the diverse regional character of Greek slavery. In all of these places the same issue surfaced: what to do with a slave who had committed some manner of public offence. We can see some similarities from place to place regarding the legally prescribed response, for example the figure of fifty lashes (found at Athens, Delos, and Pergamum) or one hundred lashes (found at Thessaly and Pergamum; the same number is stipulated in a Ptolemaic law, P.Lille I 29 col. 2, lines 25–36 = C.Ptol.Sklav. 1). We might also note the common principle that the person who reported an offender would receive half the fine (as at Thessaly, Delos, and Andania), as well as seeing the same issue surfacing at Andania—i.e. a slave being given into debt bondage to pay off an obligation—as we noted in Gortyn in Chapter 7. (We also see in the Pergamum law another issue raised in Gortynian law, namely the distinction between acts committed by slaves on their owner’s instructions and those committed on the slave’s own initiative.) But all of these regions manifest singular, bespoke responses in their slave laws, the aforementioned similarities notwithstanding. An expansive range of poleis, scattered across a number of regions, fostered a diverse range of bespoke, local institutional responses to the problems posed by slavery. Some of these responses arose from issues common to all poleis; others reflected specific regional imperatives. This produced a vast patchwork of regional variants, a world of epichoric slave systems. * * *

¹³⁰ On this document, see Saba (2012).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

196

Greek Slave Systems

At this point, it is time to shift our geographical focus outwards, and attempt to gauge how these Greek systems of slavery fit into a wider Eastern Mediterranean world. As we shall see, reductive generalizations about ‘Near Eastern slavery’ should be avoided as much as those about ‘Greek slavery’. Part III begins with a study of slavery in Iron Age II Israel, requiring a shift back in time of several centuries from the classical era and a readjustment of our geographical focus nearly 800 miles to the south-east of Athens.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Part III Slave Systems of the Wider Eastern Mediterranean World

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

9 Iron Age II Israel Yawan, Tubal, and Meshech—they were your merchants; for your merchandise they traded slaves and vessels of copper. Ezekiel, early sixth century BC¹

During commercial dredging works at Haifa’s harbour in 2007 a bronze, Corinthian-style helmet dating to c.600 BC was found in the silt.² How it ended up there is impossible to know for sure; the most likely explanation is that it belonged to a Greek soldier of fortune serving in the wars that ravaged the Near East at this time.³ Yet Greeks were not only present in this region in a military capacity: the prophet Ezekiel, active in Judah around the same time, tells us that Tyre (about 35 miles north of Haifa) imported slaves from Yawan, Tubal, and Meshech, that is, from the Greek world, the kingdom of Tabal (which stretched from the Taurus mountains to north of the Halys), and Phrygia.⁴ At least some Greeks, then, were living in this region as slaves. These facts prompt an important question for the student of ancient Greek slavery: how different a world would this region have seemed to a Greek traversing it (or labouring in it under a Phoenician or Judahite owner) during the archaic period? Would they have been confronted with an economy

¹ Ezekiel 27:13: ‫ יון תבל ומשך המה רכליך בנפש אדם וכלי נחשת נתנו מערבך‬The use of ‫‘( נפש‬soul’) to denote slaves is paralleled by the use of the cognate Akkadian term napšāti for slaves in some Assyrian documents: Dandamaev (1984): 93. ² Jessica E. Saraceni, Archaeology Magazine, 29 February 2012, http://archive.archaeology. org/news/2012/wednesday-february-29/, accessed 5 February 2018. ³ On Greek mercenaries in the archaic eastern Mediterranean, see Luraghi (2006). Fantalkin and Lytle (2016) dispute the idea that many Greek mercenaries served in the Near East (though they accept widespread Greek service in Egypt); for discussion and some criticisms, see Iancu (2016). On trade between Greece and the Levant in this period, see Greene, Leidwanger, and Özdaş (2013). ⁴ On Yawan, see Rollinger (2009), esp. pp. 37–8. On Tabal, see Bryce (2012): 141–53; Melville (2010); Simon (2012); (2017). On Meshech, that is, Muški in the Assyrian annals, see van Dongen (2014): 699–700. On Ezekiel’s discussion of Tyrian trade, see Diakonoff (1992); on Tyrian trade more generally, see van der Brugge and Kleber (2016).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

200

Greek Slave Systems

that functioned in an entirely different manner from that of their Aegean homeland? This chapter aims to work towards the beginnings of an answer to that question. It focuses for the most part on the Iron Age II (or monarchical) period, an era for which we are fortunate to possess a variety of sources: abundant archaeological data relating to households, the agrarian economy, and settlement patterns (providing the materials with which to build a bottom-up picture of economic life), and a number of texts in the Hebrew Bible. Used in cautious combination, this evidence allows us to shed some light— if only a little—on the institution of slavery in this corner of the eastern Mediterranean during the first half of the first millennium BC. A few preliminary remarks are required before we can turn to this material in earnest. The region termed ‘ancient Israel’ by modern scholars encompassed two kingdoms that together occupied what is present-day Israel and portions of Jordan. The northern kingdom of Israel was larger, wealthier, and more densely populated than its southern neighbour, the kingdom of Judah. Biblical tradition holds that these two kingdoms were once united under the kings Saul, David, and Solomon, splitting into separate kingdoms during the reign of Solomon’s son Rehoboam in the tenth century BC.⁵ The notion of a united monarchy is now hotly debated (although this debate is for the most part peripheral to the subject of this chapter).⁶ The northern kingdom experienced a vibrant history during the ninth and eighth centuries, but was dismembered by the Assyrians in 722 BC. Many Israelite refugees fled to the kingdom of Judah; others were deported. Judah, on the other hand, experienced its heyday in the eighth and seventh centuries,⁷ but like its northern neighbour was eventually brought to heel by a Mesopotamian superpower, in this instance that of Nebuchadnezzar II, whose forces sacked Jerusalem in 586 BC and deported the Judahite elite to Babylon. Although for the most part politically insignificant on the international stage of their time, Israel and Judah have bequeathed a rich literary heritage in the writings that make up the Hebrew Bible. Despite the claims of several scholars that these texts were produced from scratch in the Persian and Hellenistic periods,⁸ the majority of historians and biblical critics believe in a more complex and lengthy process of textual evolution that began during the period of the monarchy with later additions and revisions made piecemeal ⁵ For a useful sketch of Israel’s political history based on the Hebrew Bible, see Kuhrt (1995): 417–72. ⁶ For a study critical of the notion of a united monarchy, see Finkelstein and Silberman (2001). Against this minimal position, see Halpern (1995): 32–3. Much more detail can be found in Halpern (2001), esp. pp. 427–78. Cf. Carr (2010): 356–75 for a balanced view. ⁷ See Na’aman (2007). ⁸ For the so-called ‘nihilist’ approach to ancient Israel, see Thompson (1992), Davies (1992), van Seters (1992), Ahlström (1992). Against this approach, see Dever (2001), and for a more succinct rebuttal, Halpern (1995).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

201

over the period between the onset of the Babylonian exile and the second century BC. Most of the materials referred to in this chapter date to the period of the monarchy, and the issue of dates will be tackled on a case-by-case basis.⁹ It goes without saying that the use of this material for historical reconstruction must proceed with the utmost caution. For that reason, section IV of this chapter (which deals with the economic role of slavery) starts from a solid grounding in the archaeological evidence for Israelite society. Few areas in the world have been as intensively surveyed and excavated as Israel, and this places a huge amount of data in the hands of historians with which to study regional settlement patterns, methods of agriculture and animal husbandry, domestic economies, and economic inequality.¹⁰ We will use this evidence to build a picture of the Israelite economy and society from the bottom up. Once this framework is securely established, we can turn to the textual material to see how and where it might fit into this larger picture, allowing us to address questions for which the archaeological data cannot provide answers.¹¹

I. THE S TUDY OF SLAVERY IN ANCIENT ISRAEL Almost all of the scholarly effort expended on studying slavery in ancient Israel has focused on a series of laws in the Torah.¹² As we shall see, these do not provide us with a straightforward picture of slave status in ancient Israel.¹³ Instead, they belong to a broader genre of social justice legislation that began ⁹ This is a vast subject. The recent work of David Carr (2010) provides an important critique of traditional methodologies; he advocates a ‘methodologically modest’ approach that is more theoretically robust and less (unduly) ambitious in epistemological terms than many of its predecessors. This chapter largely follows Carr’s approach. See also Richelle (2016). ¹⁰ The most important recent synthesis of the archaeology of Israelite society in our period: Faust (2012). ¹¹ This section, therefore, follows W. Dever’s ‘New Biblical Archaeology’ approach. See Dever (1993), Bunimovitz and Faust (2010). For a case study showing the potential of this approach: Faust (2008). ¹² The general studies of King (2009) and Snell (2011): 17–18 are typical in this respect. For scholarship focused on the laws in the Torah, see section III below. ¹³ These laws have often been misused in reconstructing slave status in ancient Israel in a positivistic sense. Indeed, many scholars once praised the Hebrew Scriptures for their humane stance on slavery. For example, Mielziner (1859): 7 stated that ‘Keine Religion und keine Gesetzgebung des Alterthums konnte ihrem innern Geiste nach so entschieden gegen die Sklaverei sein als die mosaische, und kein Volk konnte im Hinblick auf seine eigene Entstehung sich mehr zur Abstellung der Sklaverei berufen fühlen als das israelitische Volk.’ Cf. Wallon (1847): 11–17, Sulzberger (1923): 6–11. David Brion Davis made a similar mistake, as Finley (1969): 260 pointed out. On this general scholarly approach, see Hezser (2005): 3–5. As we shall see in section III, pp. 240–11 below, the spirit of these laws lies in protecting Israelites: they do not show any concern for the plight of foreigners, nor do they embody a proto-abolitionist agenda, and there are major questions as to how these laws were enforced ‘on the ground’, if indeed some or any of them were.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

202

Greek Slave Systems

to be produced early in the monarchical period, right until (and indeed beyond) the fall of the kingdom of Judah in 586 BC. This legislation was designed to protect members of the ethnic Israelite community from the full horrors of slavery, including permanent, heritable property status, sadistic mistreatment at the hands of owners, and sale to foreigners.¹⁴ Slave status was to be reserved for outsiders. The rich legal material has been thoroughly studied by biblical scholars, almost entirely outside the mainstream of slavery studies. Aside from these texts and the predictable array of legal questions associated with them, much less attention has been paid to questions of the economic role of slave labour in the economies of Israel and Judah.¹⁵ The bulk of this chapter is split into two sections: section III closely studies the ‘slave laws’ of the Torah and slave status in Israel more generally; section IV sketches Israelite economy and society in the Iron Age II period and attempts to determine the location of slavery therein. Before we proceed to the laws of the Torah, a few comments need to be made on terminology.

II. TERMINOLOGY FO R SLAV ERY The standard terms for slaves in Hebrew are ‫( עבד‬eved) for males and ‫אמה‬ (amah) for females.¹⁶ Leaving aside for now the so-called ‘slave laws’ of the Torah, there is no reason to translate these terms as anything other than ‘slaves’ in the true property sense.¹⁷ Hebrew texts show slaves being given as gifts (e.g. Gen. 12:16; 20:14; 24:35; 29:24); they mention sale to foreigners (e.g. Gen. 37:26–9) and individuals being enslaved in wars and raids (Num. 14:31; 21:29; Deut. 1:39; 20:12–14; II Kings 5:2). One law imposed death as the penalty for the kidnapping and sale of a fellow Israelite (Deut. 24:7, incidentally the same penalty for kidnapping and enslaving a free person in classical Attica, although we do not know if the Hebrew law was ever enforced). Just as in the Greek world, slaves could be home-born or bought (Gen. 17:12–13), and distinctions were made between purchased slaves and free hired workers (Ex. 12:43–4). ¹⁴ For pragmatic reasons I will not restrict the term ‘Israelite’ to the people of the northern kingdom in this chapter, but use it synonymously with ‘Judahite’ and ‘Hebrew’. The term ‘Jew’ is normally reserved for the post-exilic period and thereafter. ¹⁵ Some notable exceptions: de Vaux (1961): 80–90; Lemaire (2015). ¹⁶ On the word ‫אמה‬, see now Bridge (2012). ¹⁷ Most scholars treat the laws of the Torah as their first port of call when analysing slave status and the terminology for slavery in Israel. But other texts are a safer guide to quotidian realities, since the laws of the Torah aim to innovate rather than to codify existing practice. These laws do show an awareness of slave status sensu stricto: see Ex. 21:6; Deut. 15:17; Lev. 25:39; cf. Lev. 25:44–6; but they aim to protect Israelites from falling into this condition. This point is developed further in section III below.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

203

As with the Greek terminology for slavery, Hebrew has various terms, which can be used in a concrete legal manner or in a metaphorical sense. This is especially true with regard to the etiquette of deference in asymmetrical relationships. Just as an Englishman of a century or two ago might have referred to himself in a letter as ‘your servant’ as a point of politeness, so too in Hebrew texts do we find the use of the word eved enlisted by the junior party as a token of deference to social superiors.¹⁸ The king, too, might refer to any of his subjects as an eved, whether they were a slave sensu stricto or not; and all human beings might be referred to as avadim YHWH, slaves of god.¹⁹ This deferential employment of the terminology of slavery can be found all over the ancient Near East, and represents a major cultural contrast with the Greek world, where such language would be seen as the height of obsequiousness rather than mere good manners.²⁰ Some Hebrew scholars claim—as some Greek historians do in relation to Greek slave terminology—that the multiple uses of this terminology reveal an indeterminate or blurred concept of slavery.²¹ In both instances this position fails to take into account the context wherein these labels are applied. Context usually allows one to determine the difference between slaves in the legal sense and individuals simply being polite. (In some cases, though, the context is not sufficiently clear for us to make a decisive pronouncement.) The study of the ‘slave’ laws of the Torah, as has been mentioned, has proceeded outside the mainstream of slavery studies, for the most part engaging biblical scholars specializing in philology and law. These scholars usually use the term ‘Hebrew slave’ as a direct translation of ‫ עבד עברי‬in Ex 21:2 and believe that these laws provide rules on slaves sensu stricto. As section III will show, however, this terminology does not reflect the substantive condition that these laws regulate, which is better described as indenture.²² Likewise, the laws commanding release of the bondsman after a certain period of time has elapsed are usually termed ‘manumission’ laws. The analysis given below will not use these terms because they misrepresent the substantive

¹⁸ Most translations of the Bible soften the meaning by translating eved with the English term ‘servant’, but this word literally means ‘slave’ (most translations do much the same in the New Testament for doulos). As Yamauchi (1966): 40 notes, this goes back to the King James translation; he finds that ‘the deliberate substitution of the word “servant” for the word “slave” is highly puzzling’. For the use of eved as a term of deference, and other metaphorical uses, see van der Ploeg (1971): 84–7; Hezser (2005): 327–32; Bridge (2010). ¹⁹ See the evidence collected in Yamauchi (1966). ²⁰ See Chapter 3, section II, pp. 84–5 above. ²¹ e.g. Westermann (1955): 42–3; Wright (1997); Hezser (2016). ²² I avoid the term ‘debt bondage’ here because the laws do not imply that the master was ever the servant’s creditor; but the indentured person’s condition amounts to much the same condition as debt bondage.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

204

Greek Slave Systems

condition described in the laws, which is one of indenture, not slavery, and which is ended by release, not manumission.²³

I I I. TH E SO -C A L L E D ‘ SLAVE L AWS ’ O F T HE TORA H The Latter Prophets preserve a wealth of material on social injustices in Israel and Judah in the monarchical period.²⁴ One of the most lamented injustices— and one that can be documented from texts dating across several centuries of Israelite history—is the sale of impoverished Israelites to richer members of the same ethnic community. An early story in the Elisha cycle tells of a destitute widow pursued by rapacious creditors who threatened to enslave her children; the prophet miraculously rescued the widow from poverty with a never-ending jug of oil, providing her with a means to settle her debts (II Kings 4:1–7). Rich individuals who bought poor Israelites who had fallen into slavery as a result of poverty are twice rebuked in the sayings of the eighth-century prophet Amos (2:6; 8:4–6). Two detailed historical episodes shed light on this problem, and will be discussed at greater length in section IV below (Jer. 34:8–22; Neh. 5:1–13). These episodes show how vulnerable individuals were easily enslaved as a result of their poverty and might end up sold into foreign lands. Indeed, Nehemiah claimed to have redeemed poor Jews who had been sold to foreigners in the mid-fifth century (Neh. 5:8). Such was the danger of borrowing in this period that one proverb held that the borrower was the creditor’s slave (Prov. 22:7), a figurative maxim, but one based on a hard truth: borrowing was a slippery slope, and the debtor might lose everything, including his freedom. It is against this precarious background that three blocks of legislation in the Torah are best understood. The first block of ‘slave laws’ is found in the earliest stratum of the Torah’s legal material, the Covenant Code, at Ex. 21:1–11.²⁵ ²³ I therefore depart from the usage given in Levinson (2005; 2006a; 2006b). This is essentially a matter of appropriate descriptive (etic) terminology; in terms of the substance of the laws, I largely agree with Levinson’s extremely learned studies. Numerous other scholars refer to the Hebrew bondsman as a ‘slave’ and his release as ‘manumission’, e.g. Chirichigno (1993), Wright (1997; 2001), Lemche (1976), Lipiński (1976), and Dandamaev (1996). Several German scholars—possibly from a more strict application of the concepts of Roman law—hesitate to use the term ‘slave’ for these bondsmen. For example, Mielziner (1859) refers to the bondsman as a ‘hebräische Knecht’ in contrast to ‘nichthebräischen Sklaven’, whilst Cardellini (1981) refers to the ‘ “Sklaven” Gesetze’ in inverted commas. ²⁴ On the issue of dating the prophetic books, see Carr (2010): 318–38; Van der Toorn (2007): 173–204. For further discussion and comparison to archaic Greece, see Lewis (2017). ²⁵ It has been contended by some scholars that the laws in Exodus are in fact later than those of Deuteronomy, but this rather unorthodox position has little to recommend it. The work of Bernard Levinson and his methodology of inner-biblical exegesis strongly supports the traditional view that the rules in Exodus are the earliest, dating to perhaps the tenth–ninth century BC

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

205

These set out special rules to be followed if someone purchases a Hebrew slave. Let us begin with Ex. 21:1–6: If you purchase a Hebrew slave, he shall serve six years; in the seventh, he shall go free, without payment. If he came in single, he shall leave single; if he had a wife, his wife shall leave with him. If his master gave him a wife, and she has borne him children, the wife and her children shall belong to the master, and he shall leave alone. But if the slave declares ‘I love my master, and my wife and children: I do not wish to go free,’ his master shall take him before God. He shall be brought to the door or doorpost, and his master shall pierce his ear with an awl; and he shall remain his slave forever.

Let us unpack this passage piece by piece. The law describes the acquisition of a ‘Hebrew slave’. As Levinson points out, this phrase is proleptic, and does not presume that the Hebrew was already a slave before the point of acquisition.²⁶ We are thus dealing with an individual who, through straitened circumstances, either has sold himself or has been sold by the head of his household, into slavery. The term ‘Hebrew’ has generated a great deal of spilt ink, but the most convincing interpretation is the banal one: that the law refers to Israelites only, not to foreigners, and is thus a social justice measure designed to protect ‘insiders’.²⁷ The same concern for insiders and lack of concern for outsiders can be found in later laws (Deut 15:3; 23:20–1; Lev 25:44–6). What is remarkable about the law is that the buyer of the ‘slave’ is immediately told that he cannot treat his purchase as property: he can retain his services for six years only and then must release him. The text does not coin a special word for this individual, but retains the normal word for slave, eved, a fact that surely owes to there being no existing term in Hebrew to describe the special condition of the bondsman as envisaged in the law.²⁸ More important is the fact that, in (Carr 2010: 470–2), followed by the seventh-century laws in Deuteronomy, and finally the laws of the Holiness Code in Leviticus, dating to the exilic period (see n. 43 below). For an introduction to this approach, see Levinson (2010). For a detailed defence of the traditional order of relative chronology (followed here), see Levinson (2006b). Chirichigno (1993) presents an altogether idiosyncratic anaylsis of the material that sees all three blocks as coeval. For cogent criticisms, see Levinson (2006b): 288–90. ²⁶ Levinson (2006b): 294–5. ²⁷ Much discussion of the term ‘Hebrew’ followed the discovery of groups called ḫabiru in the Late Bronze Age Amarna letters. These groups were vagabonds who caused trouble up and down the Levant, and the similarity of the term to ‘Hebrew’ led some to conclude that the term ‘Hebrew’ in our laws is a social rather than an ethnic designation. Na’aman (1986) convincingly shows that whilst the term ‘Hebrew’ may be etymologically related to ḫabiru, its usage in our period is exclusively limited to an ethnic group. ²⁸ As we shall see in the laws of the Holiness Code, a later attempt is made to make the distinction more terminologically clear. As Levinson (2005): 619 points out, ‘The Holiness Code avoids applying the substantive ‫עבד‬, “slave,” to the situation of one Israelite serving another. Instead it employs the language of indenture, marked by the formula of exegetical analogy, introduced by the double –‫ כ‬clause (“like a hired employee” or “like a household worker”).’ To this point I would add one qualification: the Holiness Code does not use the language of

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

206

Greek Slave Systems

substantive terms, it is clear that the eved ivri is in fact no more than an indentured servant. The law goes on to clarify the rights and duties of the two parties: if the indentured servant has a wife at the time of his ‘purchase’, he retains her; but if his master allows him to marry one of his slave girls, their marriage does not trump the property rights of the master with regard to the girl, and any children born of the union are to belong to him.²⁹ The law then provides a choice for the indentured servant at the end of his term of service: he can depart or he can become the permanent slave of his master by undergoing a special ceremony, where he is taken to a shrine (‘before God’) and has his ear pierced. This should be seen as a form of ‘slave mark’ to provide a visible indicator of slave status, of the sort we find all over the ancient Near East. Raymond Westbrook thought that the permanent slave would wear an earring to which an ‘ownership tag’ would be attached.³⁰ If we leave to one side for now the rather elastic use of the Hebrew word for slave, we can see that in substantive terms, two statuses are described in this law: the first is one of indentured service, which entails various rights and legal protections; the second is that of a slave in the strict sense, since the master owns the slave permanently (‫)לעלם‬.a³¹ A separate set of rules governs the case of a Hebrew girl sold to a fellow countryman (Ex 21:7–11). She becomes a concubine, but one with rights, much like concubines in other Near Eastern laws:³² When a man sells his daughter as a slave, she shall not be released as male slaves are. If she proves to be displeasing to her master, who designated her for himself, he must let her be redeemed; he shall not have the right to sell her to outsiders, since he broke faith with her. And if he designated her for his son, he shall deal with her as is the practice with free maidens. If he marries another, he must not

indenture, but the language of hired labour, using comparatives, introducing the inseparable preposition –‫‘( כ‬like’) to approximate the condition of the bondsman to a free hired worker. The main point here is that there is no special Hebrew term for the sort of indentured servant envisaged by these laws, and thus their compilers are forced to fall back on using the standard language of slavery (Exodus 21:2) or to deal with the substance of the issue but leave the term ‫עבד‬ to one side (Deuteronomy 15:12–15) or to use a comparison with hired labour (Leviticus 25:40) to convey what they mean. Essentially, they have to make do with existing, imperfect terminology to describe something new and different. The essence of this problem was grasped by some early commentators: Philo (De spec. leg. 2.79–85) wrote that, although called slaves (douloi), the objects of these laws were really just labourers (thetes). Cf. WDSP 10 line 3, where the terminology of slavery is enlisted for the relations between a pledged slave and his master’s creditor. ²⁹ Levinson (2005): 618; Westbrook (1998): 223–4. Cf. Lewis (2013): 401–2. ³⁰ See Westbrook (1995): 1666–7. Cf. Isaiah 44:5 for the imagery of a slave mark or tattoo. ³¹ For this term in West Semitic slave sale documents, see Gropp et al. (2001): 31. ³² On female slaves and concubines in Near Eastern law, see Westbrook (1998). It is with a female slave (‫ )אמה‬that a high official of Judah claims to be buried in an inscription from Silwan dating to c.700 BC. See Dobbs-Allsopp et al. (2005): 507–10; Avigad (1953).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

207

withhold from this one her food, her clothing, or her conjugal rights. If he fails her in these three ways, she shall go free, without payment.

Once more, the fact that the girl is a Hebrew, an insider, limits the rights of the master. She cannot strictly be called a slave, even if the law falls back on the terminology of slavery for want of a better label. She herself has recognized rights: since the delicate subject of the girl’s virginity is involved here, she cannot serve the same six-year term that the male slave does, because her sexual services make her ‘damaged goods’ from the perspective of this patriarchal society. The law thus requires that she becomes a full and permanent member of the household. In the event that she is to leave the household, the law takes special pains to make sure she returns to her paternal household and is not sold to foreigners.³³ A further law later in the same chapter reads as follows: When a man strikes his slave, male or female, with a rod, and he dies there and then, he must be avenged. But if he survives a day or two, he is not to be avenged, since he is the other’s silver.³⁴

And: When a man strikes the eye of his slave, male or female, and destroys it, he shall let him go free on account of his eye. If he knocks out the tooth of his slave, male or female, he shall let him go free on account of his tooth (Ex. 21:26–7).³⁵

For quite some time commentators have praised the mercifulness and humanity of these laws. They were, so the argument goes, the first instances of humanitarian treatment extended to slaves, the meekest members of society.³⁶ However, this view is rather misleading. The most credible interpretation is that these rules are concerned with the eved ivri mentioned earlier in the chapter, and are not general rules regarding the treatment of all ‘chattel’ slaves. This interpretation is consonant with the thrust of all the blocks of ‘slave’ legislation in the Torah, which are designed for Israelites only. The phrase ‫הוא‬ ‫ כי כספו‬at v.21 is often inaccurately translated ‘since he is the other’s property’, but the Hebrew text does not say this: a more literal translation would be ‘since he is the other’s silver’. This need not indicate anything more than the notion that the master has paid for his control over the bondsman, and thus can exercise extensive levels of control over him.³⁷ ³³ See Westbrook (1998) passim. ³⁴ Ex. 21:20–1. ³⁵ Ex. 21:26–7. ³⁶ See n. 13 above. ³⁷ Pace Chirichigno (1993): 176. (One might compare WDSP 10, the pledge of a slave in return for fifteen silver shekels from fourth-century BC Samaria.) For good arguments in favour of seeing these rules as part and parcel of the rules expressed in Ex 21:1–11 on the ‘Hebrew slave’, see Cardellini (1981): 265–8; Jackson (1988): 95. The rules on beating also make good practical sense in this context. Unlike the Neo-Babylonian example given in Chapter 3 (pp. 87–8 above), where an indentured individual’s labour services over a period of time are calculated against the

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

208

Greek Slave Systems

The second block of legislation can be more precisely pinned down to an historical context: Deuteronomy 15:12–18 belongs to the reforms of Josiah promulgated in the late seventh century.³⁸ As Levinson has convincingly shown, these laws carefully retool the earlier laws of the Covenant Code, substantially repeating the core of the earlier law but couching the new one in the language and ideology of the Deuteronomical reforms, though representing them as an original Mosaic enactment: If a fellow Hebrew, man or woman, is sold to you, he shall serve you six years, and in the seventh year you shall release him. When you release him, do not let him go empty-handed: furnish him out of the flock, threshing floor, and vat, with which YHWH your god has blessed you. Bear in mind that you were slaves in the land of Egypt and YHWH your god redeemed you; therefore I enjoin this commandment upon you today. But should he say to you, ‘I do not want to leave you’—for he loves you and your household and is happy with you—you shall take an awl and put it through his ear into the door, and he shall become your slave in perpetuity. Even for your female slave must you do likewise.³⁹ When you do release him, do not feel aggrieved; for in the six years he has given you double the service of a hired man. Moreover, YHWH your god will bless you in all you do.

Several features of this reworking of the Covenant Code rules are worth pointing out. Let us begin with continuities. First, the six-year term is retained from the earlier laws, as is the option to become a permanent slave, which once more is a decision made by the bondsman, not the master. But there are several differences. First, the language of slavery has disappeared from the first clause: the term ‫ עבד‬only appears in relation to the bondsman in verse 17 if he decides to become a slave sensu stricto, i.e. in perpetuity.⁴⁰ This represents a more literal description of the condition of the bondsman than the rough-andready terminology of the Covenant Code. Second, the procedure for this status transformation has subtly changed. In the earlier laws, the bondsman was taken ‘before God’, i.e. to a shrine, to have his ear pierced, whereas now it takes place on the doorpost (presumably of the master’s house). The reforms of Josiah abolished all local shrines and centralized worship at the temple of Solomon in Jerusalem; the amendment is a necessary practical tweak to the

debt before he is released from the obligation, these laws, like the earlier cuneiform laws of Hammurabi (LH 117; 118), set a fixed term for the indenture, and make no calculation of productivity. If the master is not given the means to coerce the bondsman, the latter can potentially sit out this term in relative indolence, and the master will be out of pocket. But the law, whilst recognizing the right of the master to inflict corporal punishment and thus motivate the bondsman to work hard, provides limits to this in order to stop gross abuse of a fellow Israelite. ³⁸ On Josiah and Deuteronomy, see Otto (2013). ³⁹ For the translation of this line, see Levinson (2006b): 300. ⁴⁰ He becomes an ‫עבד עולם‬, a ‘permanent slave’. As Levinson (2006b): 304 points out, the word for master ‫ אדון‬is also removed in Deut. 15:12–18.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

209

earlier provision to avoid a potentially lengthy journey to Jerusalem to perform the procedure.⁴¹ Third, the concubine rules of the Covenant Code have been removed and the rules for females integrated into the rules for male bondsmen: ‘Deuteronomy folds both laws into a single law.’⁴² Fourth, the warm language of the Deuteronomical reforms, with its emphasis on social justice, frames the entire block of text: the law enjoins the master to remember the life of his ancestors as slaves in Egypt, and to provision the bondsman with wine, bread, and animals when he is released (Deut 15:13–15). Notably absent are the earlier rules licensing limited use of corporal punishment. And importantly, the bondsman’s labour is likened to that of a hired man (Deut 15:18), a point taken up in our next block of legislation. The equivalent laws found in the Holiness Code of Leviticus (25:39–46) differ in several ways from their predecessors in Exodus and Deuteronomy, which they adapt and rework into a set of utopian measures. They were most likely written at some point during the exilic period:⁴³ If your kinsman under you continues in straits and must give himself over to you, do not subject him to the treatment of a slave. He shall remain with you like a hired or household labourer; he shall serve with you only until the jubilee year. Then he and his children with him shall be released from your authority; he shall go back to his family and return to his ancestral holding. For they are my servants, whom I freed from the land of Egypt; they may not give themselves over into slavery. You shall not rule over him ruthlessly; you shall fear your god. Such male and female slaves as you may have—it is from the nations round about you that you may acquire male and female slaves. You may also buy them from among the children of aliens resident among you, or from their families that are among you, whom they begot in your land. These shall become your property: you may keep them as a possession for your children after you, for them to inherit as property. You may make permanent slaves of them.⁴⁴ But as for your Israelite kinsmen, no one shall rule ruthlessly over the other.

The most important difference is that the six-year term has been removed, and instead the release of bondsmen has been integrated into a larger cycle of debt abolition and land redemption called the jubilee. This subjects the whole land to a fifty-year cycle: in the fiftieth year, the jubilee, a release (deror) is proclaimed: debtors are freed from their obligations and lands that had been

⁴¹ Levinson (2006b): 303–4. ⁴² Levinson (2006b): 301. But cf. Deut. 21:10–14 on the concubinage law relating to captive women. ⁴³ See Carr (2010): 297–303. Carr’s date ties in well with the archaeological link between Iron Age II residential patterns and the assumptions behind the law at Lev 25:29–31 brilliantly pointed out by Faust (2008); Faust’s observation makes it likely that we are dealing with an early exilic author familiar with the pre-exilic settlement patterns of Judah. ⁴⁴ For the translation of this verse, see Levinson (2005): 623–30; I follow Meek’s translation given at p. 628 of Levinson’s essay.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

210

Greek Slave Systems

distrained upon are restored to their ancestral owners (Lev. 25:8–55). But these laws take special pains to point out the difference between the rules for indentured Israelites and enslaved foreigners, a subject already present in both earlier blocks of legislation, but here made even more explicit.⁴⁵ Missing in this text is any procedure for an Israelite to become a permanent slave; the bondsman’s condition is explicitly grouped with hired labourers and contrasted with slaves (Lev. 25:39–40). This fate is explicitly singled out for non-Israelites, and the legal powers of their masters emphasized, especially the permanence and heritability of the slave condition (Lev. 25:46). Whereas the laws of Exodus and Deuteronomy create a special position for Israelite bondsmen, yet permit the possibility that they may enter slavery sensu stricto through a transformative ceremony, the Holiness Code takes the idea of the earlier laws to its logical conclusion, forbidding any route whatsoever by which Israelites might become enslaved. Almost all scholars agree that the jubilee legislation in which these laws are contained is late and utopian, and was never put into practice.⁴⁶ It is more difficult to say that the rules of Exodus and Deuteronomy were never enforced ‘on the ground’. We do find an important episode at Jer. 34:8–22, where the people of Judah are castigated for not following the six-year bondage law; but there is little more that we can say beyond this. At any rate, several important lessons are worth drawing from these laws. First, they do not describe the standard historical form of slavery in ancient Israel. The fact that they all struggle with conventional terminology to describe a special form of indenture shows that they are introducing something new and unusual, not simply codifying existing practices. Second, as Westbrook and Wells have rightly pointed out, the differences in detail between one block of legislation and the next can sometimes blind scholars to the fact that all three blocks aim at substantively the same thing: the protection of Israelites from the full horrors of slavery.⁴⁷ This limiting of the rules to Israelites as opposed to foreigners is present right from the start, in the Covenant Code’s rules on the ‘Hebrew ⁴⁵ My sole disagreement with Levinson (2006b) is that he believes that these three blocks of legislation witness a change from enslavement of Israelites to mere indenture. At p. 304, Levinson argues that this change occurs in Deuteronomy, where the master no longer ‘buys’ a Hebrew slave, but a Hebrew ‘indentures’ himself to the master. But both cases use verbs of sale (‫ קנה‬in Ex 21:2 and ‫ מכר‬in Deut 15:12), showing that the substance of the transaction has not changed, only the way in which it is framed. As I have argued in section II, pp. 205–9 above, this protection of Israelites from slavery sensu stricto is implicit from the Covenant Code onwards. Much of this boils down to how one defines ‘slave’. Levinson takes the use of ‫ עבד‬in the Covenant Code literally to mean that we are dealing with slaves, calling their release ‘manumission’. But the substance of the condition is utterly different from slavery properly defined. It makes more sense to see the use of ‫ עבד‬as the nearest standard word the composer of the law can find to describe the condition that he means to regulate. It is not a good fit, which is why it is abandoned in Deuteronomy and Leviticus. ⁴⁶ See Westbrook (1991): 36–57 for further discussion. ⁴⁷ Westbrook and Wells (2009): 132–3.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

211

slave’. Deuteronomy reworks these rules in line with the broader themes of the Deuteronomical reforms, and the Holiness Code works in dialogue with both the Covenant Code and the Deuteronomical material, synthesizing and reworking the earlier rules to bring them in line with its new utopian vision. But all three blocks are aimed at social justice and the protection of insiders, and in this sense at least show a line of continuity even as they adapt and transform earlier rules. The laws should not be contrasted with the slave laws of other eastern Mediterranean societies in order to argue for a ‘special’ Israelite form of slavery. Rather, they should be likened to the social justice legislation both of Mesopotamian societies to the east which aim to protect the rights of free debtors (a tradition on which they draw) and of similar laws from Crete and Athens to the north-west, whose overall rationale runs along very similar lines, whether or not they were influenced by these Near Eastern antecedents.⁴⁸

I V . S L A V E R Y I N T H E E C O N O M Y OF IRON AGE I I ISRAEL Although there were a number of differences in the political structure of the kingdoms of Israel and Judah, there was a basic similarity in lifestyle at the micro-level of village and household.⁴⁹ Here, we will begin by sketching the basic structures of the agrarian economy and patterns of land tenure, before proceeding to examine inequalities of wealth and the estates of the elite. After taking these elements into consideration, we can then turn to issues of extrafamilial labour and the role of slavery in the generation of elite wealth. Israel and Judah display different geographical profiles that lent themselves to multiple methods of exploiting the landscape.⁵⁰ Broadly speaking, however, most of the good agricultural land in both regions admitted forms of exploitation that should be familiar to students of the Greek economy. A close symbiosis between animal husbandry and intensive arable farming provided the livelihood for the average small farmer, who also diversified his activities to include olive oil production and viticulture. The countryside was marked by walled village settlements enclosing clusters of houses, each house containing an extended family. On the village level, excavations have revealed cooperative efforts in processing oil and wine as well as the storage of foodstuffs.⁵¹ Land tenure was private; combined with a system of partible inheritance, this ⁴⁸ On Cretan laws of debt bondage, see Kristensen (2004a). For comparison between Israel and Greece, see Lewis (2017). ⁴⁹ Fleming (2012): 23. ⁵⁰ For a study of the economic regions within Judah, see Faust and Weiss (2005) passim. ⁵¹ On residential patterns in the countryside, see Faust (2012): 128–77.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

212

Greek Slave Systems

produced a landscape of fragmented smallholdings not dissimilar to that found in Greece. In a thorough collation and analysis of all available survey and excavation data, Avraham Faust has recently sketched the difference between the urban and rural sectors and between the rich and poor. Houses in village settlements are remarkably uniform in size, suggesting a relatively egalitarian society in the countryside. The urban remains display a somewhat different profile, with small numbers of large, well-built houses of wealthy families interspersed among a larger number of small houses that most likely contained nuclear families.⁵² Faust’s study thus contextualizes the many comments found in the prophetic books that speak of the grand houses of the elite as well as the abuse to which this sector of society subjected the poor.

Elite Households and Estates By the early years of the monarchy ample evidence confirms the existence of a wealthy urban elite. From where did this elite draw its income? Most scholars would admit that although the elite was far from a homogeneous group, the majority of wealthy individuals derived their affluence from estates in the countryside.⁵³ Indeed, the prophet Isaiah lamented the growth of such estates in the eighth century, rebuking those who joined house to house and field to field, pushing out smaller neighbours. Great houses, he warned, would become desolate (Is. 5:8). Similarly, Micah (2:1–2) rebuked powerful men who hatched plots to deprive their weaker neighbours of their dwellings and fields. This pattern must be seen in its archaeological context: cash crop production for foreign markets—wine and animal products, but above all wheat—had reached a remarkable scale in seventh-century Judah, a phenomenon that must be explained for the most part in terms of royal and elite landholdings.⁵⁴ We must now turn to the available evidence for the different forms of extrafamilial labour that were available to the elite for the cultivation of its lands. Here, archaeology provides no answers, and we must turn instead to the texts of the Hebrew Bible. Let us begin with a traditional view. Isaac Mendelsohn wrote: The Ancient Near Eastern peasants were a hard-working and thrifty people. The whole family worked on the land, which was the sole provider for its material existence. Since the landed property of the average farmer was small and his

⁵² See Faust (2012): 39–127. ⁵³ Faust (2012): 15–16. For pioneering archaeological work on large Iron Age II estates in the vicinity of Jerusalem, see Moyal and Faust (2015). ⁵⁴ See Faust and Weiss (2005): 74–5 on large grain surpluses from Judah (‘the “grain basket” of the entire region’, p. 80), pp. 75–6 for olive production in the Shephelah; pp. 80–2 on wine production in the vicinity of Jerusalem.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

213

family large, there was no great need for outside help in the form of hired labourers or slaves; the peasant household was self-sufficient. While this was the case of the average small-scale farmer, the situation in regard to outside help of the wealthy landowner and aristocracy was, of course, radically different. Their large estates had to be worked and supervised, if not also managed, by hired help. This help, however, was only to a very small degree drawn from the ranks of hired agricultural labourers and slaves. It came primarily and overwhelmingly from the ranks of the dispossessed peasantry in the form of agricultural tenancy.⁵⁵

Mendelsohn was, of course, generalizing across the whole of the ancient Near East. But a close examination of the Hebrew Bible not only fails to corroborate his assertion, but also provides strong grounds for rejecting it in relation to this region (at the very least).⁵⁶ We may leave aside for now the issues of both temple servants and the corvée: both existed in ancient Israel, but their benefits accrued to shrines, on the one hand, and the Crown, on the other.⁵⁷ We must look elsewhere to determine the labour resources of the wealthier classes. Perhaps the most remarkable fact in relation to this problem is that the Hebrew Bible contains not a single reference to agricultural tenancy.⁵⁸ One could argue that the theological and political focus of these texts might preclude attention to such matters. However, these writings are generically very diverse, many of them deeply rooted in the soil of farming and everyday life. Although an argument ex silentio, it is extremely unlikely on the basis of this striking omission that tenancy was as important in Israel as Mendelsohn believed. It was once popular to see in the admonitions of the prophets evidence of a wide-scale ‘latifundization’ of Israel and a dispossession of the ‘proletariat’, who worked as tenant farmers, but Faust has convincingly shown that such drastic extrapolations are not warranted: they do not fit the archaeological record, and although there are no good reasons to reject the information in the prophets out of hand, the information which they provide only attests the existence of social abuses, not radical social breakdown and the displacement of vast numbers of people.⁵⁹ The fact that not one of the ⁵⁵ Mendelsohn (1949): 109. The idea that slavery was unimportant in Israel can be found elsewhere. For example, Whybray (1990): 42 writes in relation to the term eved in the book of Proverbs that ‘it is noteworthy that it never occurs in the plural: there is no reference to multiple slave ownership on a grand scale’. But one could come to the same conclusion from reading Theophrastus’ Characters or Plato’s Protagoras in isolation; this hardly proves or disproves anything about the importance of slavery in Israel. Cf. Wright (2001): 83. ⁵⁶ Even the complete autarky of small farmers seems rather exaggerated: cf. Hardin (2010) for an in-depth analysis of the domestic economy of a single house excavated at Tell Halif. ⁵⁷ On temple servants, see de Vaux (1961): 89–90, but these oblates may have been in a similar legal position to the Babylonian širkus, and not slaves sensu stricto: see Levine (1963). On širku status see Chapter 11, p. 237 n. 11. On corvée labour, see Rainey (1970; 2003). ⁵⁸ As de Vaux (1961): 167 points out, ‘the system of rent-holding or métayage . . . was apparently never practiced in Israel in the early days, though it was known in Mesopotamia . . . the first mention of the renting of lands is found in the parable in Mt. 21:33–41’. ⁵⁹ See Lods (1932): 397–8 for the old-fashioned view; cf. Faust (2012): 14–15 and passim.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

214

Greek Slave Systems

prophets ever mentioned the abuse of tenant farmers is strong a priori grounds for assuming that tenancy cannot have been of central importance (though its contemporary existence in Greece and Mesopotamia suggests that denying its existence altogether may be rather extreme). The Hebrew texts do, however, provide us with much more information on two forms of labour that are familiar from the Greek world of the early archaic era: hired farmhands and slaves.⁶⁰ A good place to start is with an early text: the Decalogue at Ex. 20:2–17. On the Sabbath Day, the law enjoins, one must not work, nor must one’s son or daughter, one’s male or female slaves, animals, or any resident aliens staying in the household. A further commandment (Ex. 20:17) exhorts individuals not to covet a fellow countryman’s house, spouse, male or female slaves, animals, or other possessions. A similar formulation can be found three times in the laws of Deuteronomy (Deut. 5:12–15; 12:18; 16:11). As a basic description of the household, these texts do not amount to much. They do not tell us whether slave ownership was a common feature of Israelite society or was enjoyed by the elite alone; nor do they tell us the relative importance of familial, slave, and hired labour. All the same, these descriptions do provide a basic blueprint of the make-up of a household that must have been familiar to their immediate audience, and the fact that the formulation is repeated in Deuteronomy suggests that it was still relevant in the latter days of the Judahite monarchy. Various texts show that hired labour was a common phenomenon in Israel.⁶¹ Like the thes of Homeric poetry, hired workers in Israel were itinerant and might reside with a family for a period of time and act as a useful source of supplementary labour. As in Homer, Hebrew texts present the position of the hired worker as precarious: without a local network of kin from which to draw support, he might easily be abused or denied his wages (Deut. 24:14–15; Jer. 22:13; cf. Hom. Il. 21.441–7). A number of other texts, however, show that slavery merits serious consideration as the core labour force on elite estates. We may begin with some fictional stories: the stories of the patriarchs as related in Genesis. As Hezser has recently pointed out, the patriarchs (Abraham, Isaac, and Jacob) as well as their contemporaries are all presented as wealthy slave owners.⁶² Abraham is depicted as a wealthy nomad with a large household, including numerous slaves (Gen. 16:1–15; 17:12–13; 17:23; 24:1). When he moves to Egypt, Pharaoh lavishes many gifts upon him: flocks, oxen, donkeys, camels, and male and female slaves (Gen. 12:16). These items seem to be standard articles of wealth, since Abraham receives similar gifts from another potentate, Abimelech of Gerar: sheep, cattle, and male and female ⁶⁰ De Vaux (1961): 167 writes that large estates ‘were worked by slaves (2 S 9:10), or by paid workmen’, i.e. the same conclusion argued for in this chapter. Lemaire (2015) comes to a similar conclusion. ⁶¹ See de Vaux (1961): 76 for sources and discussion. ⁶² Hezser (2005): 286–8.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

215

slaves (Gen. 20:14). Abraham’s son Isaac is depicted as possessing similar goods: in Gen. 26:12 he settles to plant crops and raise livestock, acquiring riches and a ‘large household’, which, as Hezser points out, must mean a large number of slaves.⁶³ Elsewhere his wives are depicted as receiving slave girls from their father upon marriage (Gen. 29:24–9). Like his father and grandfather, Jacob is presented as a wealthy slave owner. After tricking his brother Esau out of his birthright, Jacob increases his fortune and prospers, becoming rich in flocks, male and female slaves, camels, and donkeys (Gen. 30:43). When he attempts to reconcile with his brother, he dispatches a messenger to tell Esau that he is staying with his father-in-law Laban, and owns oxen, draught animals, flocks, and slaves (Gen. 32:6). Esau himself is represented as having acquired a large household, presumably on the same model. It would be imprudent to use these stories as evidence for slaveholding among ‘historical’ patriarchs: few scholars today would be prepared to hold the sort of hopes for historical analysis of the ‘patriarchal age’ once ventured by historians such as Bright.⁶⁴ Besides, historians have long recognized various anachronistic features of these stories (such as the presence of camels) which show that although set in the distant days of the Late Bronze Age, they tell us no more about that era than Homer’s epics do about the Mycenaean age. What is much more interesting, however, is the notion of these texts as cultural artefacts that reveal something about the assumptions of their immediate audience.⁶⁵ Carr has recently argued that these stories are probably early texts deriving from the northern kingdom of Israel.⁶⁶ And one key assumption that runs throughout these tales is that a rich man will own many male and female slaves. A similar assumption can be seen in texts from the historical narratives of Judges, Samuel, and Kings. In Judges 6:27 we hear of one such character, Gideon, taking ten of his slaves to demolish a shrine of Baal. Again, this number of slaves must have seemed plausible to the text’s audience, and it is clear that they are presented only as a portion, and not the sum, of Gideon’s holdings.⁶⁷ Once more, though, this story belongs to a genre of semi-legendary ⁶³ Hezser (2005): 286–7. ⁶⁴ Bright (1981): 68–9. ⁶⁵ See King and Stager (2001): 7 in relation to biblical social history: it matters little whether the biblical accounts are “true” in the positivistic sense of some historians and biblical scholars. It is enough to know that the ancient Israelites believed them to be so. The stories must have passed some test of verisimilitude, that is, having the appearance of being true or real. (cf. their quotation of Oswyn Murray, p. 8). Cf. Bendor (1996): 46, who points out that the familial structure of these stories relates to the period of the monarchy. ⁶⁶ Carr (2010): 473–7, and pp. 484–5 on the Abraham stories, which he is less confident in dating. ⁶⁷ Carr (2010): 480 argues for much early northern material in Judges, but notes that ‘it is not easy to identify the contours of such originally Northern traditions with precision, since they have been reframed and revised in light of Judean and later semi-Deuteronomistic concerns’.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

216

Greek Slave Systems

folk tales which, though far closer to the historical realities of the Iron Age I period than the patriarchal narratives are to the Late Bronze Age, cannot be taken as ‘historical’ in a positivistic sense. One episode that may well, however, contain good historical information on slavery and the economic position of courtiers can be found at II Sam. 9:1–13.⁶⁸ Here, King David, upon assuming the throne and slaughtering the descendants of Saul, asks a man named Ziba (who is a slave or steward of Saul’s family) if any of Saul’s relatives are left alive. There is only one, a crippled man named Mephibosheth,⁶⁹ son of Jonathan, grandson of Saul (9:3). David spares the life of Mephibosheth and makes him a courtier, granting him his grandfather Saul’s ancestral lands. He throws into the bargain Ziba, who must return to these estates and cultivate them. David addresses Ziba thus: You and your sons and your slaves shall farm the land for him and shall bring in its yield to provide food for your master’s grandson to live on; but Mephibosheth, your master’s grandson, shall always eat at my table. Ziba had fifteen sons and twenty slaves. Ziba said to the king, ‘Your servant will do just as my lord the king has commanded.’ (II Sam 9:10–11)

The nature of the Hebrew vocabulary introduces an element of uncertainty as to the exact status of Ziba. He is called a slave (‫ )עבד‬of Saul’s house at II Sam 9:2, and owns twenty slaves himself; these latter could be interpreted as servi vicarii of a sort, not unlike what we find attested in Homer (Od. 14.449–52). One could thus interpret Ziba as a slave sensu stricto, occupying a high position in the household of King Saul and holding numerous slaves of his own. On the other hand, as noted in Chapter 3, section II, pp. 84–5 above, Hebrew utilizes the language of slavery figuratively in the context of asymmetrical relationships; when Ziba is described as an ‫ עבד‬of Saul’s house, it is possible that this is a generic term for a subordinate in Saul’s service and not literally a slave, just as Ziba addresses David by saying ‘Your (lit. ‘his’) slave (‫ )עבדו‬will do just as my lord (‫ )אדני‬the king has commanded’, i.e. using the language of slavery figuratively in the presence of the king, even though it is clear that David is not Ziba’s owner. This latter interpretation is probably the more likely.⁷⁰ Even The Gideon story, with its emphasis on monotheism, seems to be coloured with these late Judahite concerns. ⁶⁸ On the history of David in the books of Samuel, see Halpern (2001). For the date of II Samuel, see Halpern (2001): 57–72. ⁶⁹ Or in some versions of the text, Meribaal. See Halpern (2001): 34 with n. 20. ⁷⁰ This conclusion follows from later episodes in the Davidic narrative. In II Sam 16:1–4, David, abandoning Jerusalem during the revolt of Absalom, encounters Ziba once more, who provides provisions for him and his men. When David asks Ziba about Mephibosheth, Ziba states that he is staying in Jerusalem in the hope that he will be elevated to Saul’s throne; disgusted, David awards Ziba all of the property of Saul that had been previously restored to Mephibosheth. But later, in II Sam 19:25–31, after David has defeated Absalom and is returning

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

217

so, it is remarkable that the bulk of the workforce overseen by Ziba and which undertakes agricultural work to support Mephibosheth at David’s court is, in fact, made up of slaves.⁷¹ That said, it is probably unwise to venture ambitious generalizations about the standard size of slaveholdings among courtiers from this text alone. The story of this passage does not stop there. Its details have been used to explicate a group of over a hundred inscribed ostraca found in the northern capital of Samaria within a stratum dating to the reign of Jeroboam II (786–746 BC).⁷² The inscriptions have a common formula recording the delivery of commodities such as refined oil and vintage wine, mentioning the date, supplier, recipient, and commodity.⁷³ The texts were certainly produced in Samaria and thus were not dockets that accompanied the shipment; sometimes they even mention the place from which the supplies were sent. A couple of examples will illustrate the general appearance of these texts: In the ninth year. From Yaṣit, to Abinoam, a jar of old wine.⁷⁴ [In] the tenth year. From Sepher, to Gaddiyaw, a jar of refined oil.⁷⁵

Some debate has occurred relating to the grammar of these inscriptions; Yadin, for example, argued that the inseparable preposition ‫ ל‬should be translated not as ‘to’ but as ‘from’.⁷⁶ In the first example above, therefore, he would translate ‘from Abinoam’. Rainey has rightly objected that this translation is not admissible: ‫ ל‬must have its usual prepositional meaning of ‘to’, because elsewhere in the ostraca the scribes use the standard Hebrew term ‫מן‬ for ‘from’. The basic grammatical features of the documents are now securely understood (and the translations given above accurate).⁷⁷ As to what system of provisioning the ostraca denote, Lemaire has reviewed several competing hypotheses. He rightly points out that the documents cannot be evidence of a system of royal tithes or accounts for the Crown because (i) no biblical evidence supports such a hypothesis (and if they existed, we should expect them in I Kings 5:7–8, but they are not mentioned there), (ii) these hypotheses fail to explain the large number of personal names among the recipients, rather than just a couple of court officials (whereas the scribal hands are few, in palaeographic terms, so the recipients and scribes must be different to Jerusalem, Mephibosheth meets him in a dishevelled state, protesting his loyalty and claiming that Ziba had slandered him. David orders that Saul’s estate therefore be divided equally between Ziba and Mephibosheth. These two episodes make less sense if we suppose that Ziba had been Mephibosheth’s slave stricto sensu, and more sense if he were a free steward. Ziba is in fact called a ‫‘ נער‬retainer’ of Saul in II Sam 9:9. ⁷¹ II Sam 16:1 provides a sample of the products of their labour: bread, raisin cakes, fig cakes, and wine. ⁷² See Dobbs-Allsopp et al. (2005): 423–6 for an overview of the documents. ⁷³ For the nature of the refined oil in these documents, see Stager (1983). ⁷⁴ Dobbs-Allsopp et al. (2005) Samr 10. ⁷⁵ Dobbs-Allsopp et al. (2005) Samr 16a. ⁷⁶ Yadin (1959). ⁷⁷ See Rainey (1967): 32–4.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

218

Greek Slave Systems

persons), and (iii) the commodities trickle in jar by jar, not in large coordinated shipments.⁷⁸ He follows Rainey in interpreting the documents as records of deliveries of foodstuffs from estates scattered across the countryside to individual courtiers living at the palace of Jeroboam II. Both scholars argue that these estates were, like that of Mephibosheth, Crown property granted to courtiers to support them materially.⁷⁹ If the system attested in the Samaria ostraca was similar to that based at the court in Jerusalem (as seems likely), we must seriously consider the possibility that slave labour of the sort described at II Sam 9: 9–11 underpinned much of the productive process generating these prestige goods. A less specific but nonetheless revealing passage occurs at II Kings 5, a story in the Elisha cycle. Naaman, a leprous general of the king of Aram, visits Elisha on hearing of his reputation as a healer, and his leprosy is cured. Elisha refuses to take payment for the service, and Naaman departs; but a menial of Elisha’s household, Gehazi, sees an opportunity to make a profit, and follows Naaman out of town. Approaching the general, he pretends to deliver a message from Elisha asking for two silver talents and some clothing, and upon receiving these returns to town and hides them. Elisha, however, is not deceived: he confronts Gehazi and asks him ‘Is this a time to take money and to receive clothing and olive groves and vineyards, sheep and oxen, and male and female slaves?’ (II Kings 5: 26). Naaman’s leprosy then cleaves to Gehazi as a punishment for his treacherous act. Our interest in this story lies in the itemized list of riches: this incidental, off-hand remark is revealing, as it shows male and female slaves as a standard article of wealth alongside olive groves, vineyards, livestock, and clothing. Let us now turn to a very different text, Prov 31: 10–31.⁸⁰ This is an acrostic poem, and it provides a fascinating window into the female sphere of the domestic economy. Each verse extols the various virtues of the perfect wife, and each starts with a different letter, proceeding through the Hebrew alphabet from aleph to taw. From the content of the poem, it is clear that we are ⁷⁸ Lemaire (1977): 73–4. ⁷⁹ Rainey (1967): 39; Lemaire (1977): 75–7, whose interpretation is followed by Kuhrt (1995): 465. Rainey (1967): 39 points to the warning about monarchy at I Sam 8:14 as an apposite textual parallel, where monarchs will seize the lands of their subjects and distribute them to their courtiers. This is certainly possible, though it seems to me that to interpret all of these estates as land granted by the king is perhaps too mechanical an extrapolation of II Sam 9: 1–13. It is more probable that the estates were a mixture of lands granted by the king and ancestral estates of the courtiers: it makes little sense to suppose that courtiers would only draw upon estates granted by the king for their sustenance and not on their other lands as well. To reach the position of courtier, it must surely have been the case that individuals were already wealthy and prominent, which implies relatively extensive landholdings. ⁸⁰ For strong reasons to prefer an early date for Proverbs, see Carr (2010): 403–31. The claim of Wolters (1985) that the acrostic is of a later date is based solely on the notion that ‫‘( צופיה‬she watches over’) in v. 27 is a pun on the Greek term sophia (‘wisdom’), which seems (to me at least) rather a stretch.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

219

dealing with an elite household. Much like her Greek counterpart, the Israelite wife is in charge of the running of the household. She rises early in the morning, when it is still dark, and prepares the food for her household, including the rations for her female slaves (Prov 31:15: ‫ ;לנערתיה‬ταῖς θεραπαίναις in LXX). Several verses focus on textile production, including the provision of wool and flax, the spinning of yarn, and the weaving of fine purple garments (Prov 31:13, 19, 21, 22). These tasks must intensively occupy the aforementioned female slaves as well as the ideal wife, because not only are enough garments produced to clothe the household’s members in style, but a cloth surplus is also produced that is sold to merchants (Prov 31:24), which implies a great deal of labour invested in the production of textiles. The wealth of her house is such that new fields can be added to its existing holdings (cf. Hesiod, Op. 341), her husband has the leisure to sit with the elders at the town gate, and enough substance is left over to distribute to the poor (Prov 31:16, 23, 20). This poem provides only a partial, selective picture of the Israelite domestic economy, but it shows that just like elite households in the archaic Greek world, the mistress of the house would expend some effort in the production of textiles alongside her female slaves as well as taking charge of its day-to-day organization. One of the most detailed passages in the Hebrew Bible on the use of slaves by the elite can be found at Jer. 34:8–22. Its events date to 588 BC, in the final years of the kingdom of Judah, when the menace of the Babylonians was keenly felt and action was taken to ensure divine favour: The word which came to Jeremiah from YHWH after King Zedekiah had made a covenant with all the people in Jerusalem to proclaim a release (deror) among them—that everyone should set free his Hebrew slaves, both male and female, and that no one should keep his fellow Judahite enslaved. Everyone, officials and people, who had entered into the covenant agreed to set their male and female slaves free and not to keep them enslaved any longer; they complied and let them go. But afterwards they turned about and brought back the men and women they had set free, and forced them into slavery again. Then it was that the word of YHWH came to Jeremiah from YHWH: thus said YHWH, the god of Israel: I made a covenant with your forefathers when I brought them out of the land of Egypt, the house of bondage, saying: ‘In the seventh year each of you must let go any fellow Hebrew who may be sold to you; when he has served you six years, you must release him.’ But your fathers would not obey me or give me ear. Lately you turned about and did what is proper in my sight, and each of you proclaimed a release to his countrymen; and you have made a covenant accordingly before me in the house that bears my name. But now you have turned back and profaned my name; each of you has brought back the men and women whom you had released, and forced them to be your slaves again.

The release (deror) enacted by Zedekiah belongs in a long tradition of proclamations by Near Eastern kings to a similar effect, known as mīšarum edicts. These cancelled debts and otherwise provided relief to the impoverished, and

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

220

Greek Slave Systems

were sometimes proclaimed upon the accession of a king.⁸¹ Zedekiah’s deror is in fact a similar sort of blanket release to that outlined in the laws of Leviticus (Lev. 25: 10–12), although the latter represents a systemized approach rather than an ad hoc measure. The issue is relatively straightforward: various officials and people were holding their fellow Judahites as slaves. Jeremiah’s oracle specifically refers to a law on bondsmen, and the fact that it is transgressed must indicate that the time limit of six years of service was not being observed. The guilty parties had made a covenant with YHWH in the temple of Solomon to release the slaves, but reneged on the promise. For our purposes, this passage is revealing in two ways: first, it shows that better-off Judahites were exploiting slave labour, indeed enslaved Judahite labour; and second, it shows the limits of the laws studied earlier in this chapter. Without adequately strong institutions to enforce them, they were not followed. Zedekiah himself did not enjoy his throne for much longer. In 586 BC Nebuchadnezzar sacked Jerusalem, killed Zedekiah’s sons before him, blinded him, and bore him off to Babylon in chains. Judah was not emptied of people during the exile, but much of its elite was removed to Babylonia, and remained there for the next half-century. When we move to the post-exilic (Second Temple) period, one striking fact is the level of continuity with earlier practices. Judah was by this point the Persian province of Yehud, and the same problems plagued it in the mid-fifth century as in the early sixth.⁸² Nehemiah 5:1–13 provides a similar story to Jer. 34:8–22: There was a great outcry by the common folk and their wives against their brother Jews. Some said, ‘Our sons and daughters are numerous; we must get grain to eat in order that we may live!’ Others said, ‘We must pawn our fields, our vineyards, and our homes to stave off hunger.’ Yet others said, ‘We have borrowed silver against our fields and vineyards to pay the king’s tax. Now we are as good as our brothers, and our children as good as theirs; yet here we are subjecting our sons and daughters to slavery—some of our daughters are already subjected—and we are powerless, while our fields and vineyards belong to others.’ It angered me very much to hear their outcry and these complaints. After pondering the matter carefully, I censured the nobles and the officials, saying, ‘Are you pressing claims on loans made to your brothers?’ Then I raised a large crowd against them and said to them, ‘We have done our best to buy back our Jewish brothers who were sold to the nations; will you now sell your brothers so that they must be sold back to us?’ They kept silent, for they found nothing to answer. So I continued, ‘What you are doing is not right. You ought to act in a god-fearing way so as not to give our enemies, the nations, room to reproach us. I, my brothers, and my servants also have claims of money and grain against them; let us now abandon those claims! Give back at once their fields, their

⁸¹ See Lemche (1979).

⁸² For the date of the Nehemiah memoir, see Carr (2010): 208.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Iron Age II Israel

221

vineyards, their olive trees, and their homes, and abandon the claims for the hundred pieces of silver, the grain, the wine, and the oil that you have been pressing against them!’

Here once more we find an entrenched landowning elite enslaving their fellow Jews contrary to the spirit of the Torah (which by this point was a fixed canonical text).⁸³ Let us conclude our review of slavery in the Hebrew Bible with two further texts. The first is the Book of Job.⁸⁴ Job is presented as a man from Uz who possesses great wealth on the model of the patriarchs as discussed earlier in this section. He has seven sons, three daughters, seven thousand sheep, three thousand camels, five hundred yoke of oxen, five hundred she-donkeys, and many slaves (Job 1:1–3). Disaster befalls him when his livestock are carried off by raiders and his slaves butchered (Job 1:14–18). Reduced to dire straits, his remaining slaves no longer fear or obey him (Job 19:13–21); but by the end of the book his fortunes are restored to twice their original extent (Job 42:10–16). Again, we find the same assumption noted in other texts, that a wealthy man will own large numbers of slaves. Our second text is the Book of Qoheleth, or Ecclesiastes.⁸⁵ The author of this text claims to be no less than Solomon himself, though few would endorse this claim as historical, and it is generally regarded as a pseudepigraph. In his search for the meaning of life he makes various experiments in ways of living, one of which is the hedonistic life (Qoh. 2:1–12). He plants gardens and vineyards, builds palaces, and acquires estates with male and female slaves (including home-born slaves) to work them (Qoh. 2:4–7). Once more, the assumption is that the estates of this rich man will be worked by slave labour.⁸⁶

V . CONCLUSION S It is important that we acknowledge the many difficulties in utilizing the evidence of the Hebrew Bible when addressing the issue of slavery in ancient Israel. Controversy regarding the dates of different parts of this literary anthology is inevitable, and even if it were not, the slim pickings from Hebrew texts discussed in this chapter provide far less detail on labour on elite landholdings

⁸³ Van der Toorn (2007): 249. ⁸⁴ Joosten (2013) dates Job to the exilic period on linguistic grounds; cf. Coogan (2009), who presents a complementary case on the grounds of incidental details in the narrative. ⁸⁵ On the date and composition of Qoheleth, see Carr (2010): 448–55. ⁸⁶ Another text worth noting, though somewhat later than our period (early second century BC), is Sirach. As de Ste. Croix (1981): 142 points out, the statement ‘bread, punishment and work for a slave’ at Sir. 33:25 closely mirrors [Arist.] Oec. 1344a35.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

222

Greek Slave Systems

than we should like. That said, several points are worth emphasizing. First, there is simply no sign of agricultural tenancy anywhere in these texts, which is most unlikely to be a chance omission. Second, numerous references exist to slaves and hired labourers, the same basic labour force whose importance we have already noted in the contemporary archaic Greek world. Third, a consistent theme running throughout these texts associates rich men with large holdings of slaves, which is either an eccentric literary convention or (far more likely) reflects assumptions about real economic practices in ancient Israel. It is perhaps worth noting that regardless of ongoing debates among scholars on the date of this or that text, this theme remains consistent across a variety of texts dating to a variety of periods. Finally, though the so-called ‘slave laws’ of the Torah have been seen as a distinctive feature that sets Israelite practices in contrast with those of other places, a close look at their provisions shows that slavery sensu stricto (i.e. ‘chattel’ slavery) did exist in ancient Israel, and it was precisely this condition (entailing permanent ownership and the potential of sale to foreigners) that Israelite scribes laboured to outlaw as a potential fate for members of their ethnic group. Like Solon’s reform outlawing enslavement of free debtors and the debt bondage legislation of Gortyn, these rules aim to provide a measure of protection for ‘insiders’, whilst retaining the institution of slavery itself, so long as slave status was reserved for ‘outsiders’. * * * In Chapter 10, we will turn to the north-east: that is, to the heartland of the Assyrian Empire in the eighth–seventh centuries BC. The evidence for slavery there presents challenges almost the exact opposite of those posed by the Hebrew material: whereas the latter provides us with a tolerably clear picture of the basic contours of slavery in the elite economy but no fine-grained detail or precisely dateable texts, the Assyrian material provides us with some extremely detailed evidence which can often be accurately dated. The challenge is to reconstruct the contours of Assyria’s elite economy itself.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

10 Assyria The Eighth–Seventh Centuries BC

In his book Slavery in the Ancient Near East, Isaac Mendelsohn played down the economic contribution of slave labour to overall production in Near Eastern societies, though, as we noted in Chapter 4, this is something quite different from what classical historians have focused on over the past fifty years, viz. the contribution of slave labour merely to the wealth and position of elites. Mendelsohn, however, drew attention to an apparent exception: agricultural slavery in Assyria during the eighth–seventh centuries BC. Although (as we shall see) there are cogent reasons to dispute his view of the role of slavery in Assyrian agriculture, he was right to pinpoint this society as one of special interest to historians of slavery.¹ Of particular note is the large number of slave sale documents. A brief survey of the contents of two recent collections of Assyrian legal documents from the royal court at Nineveh² will serve to illustrate something of the quantity of material at our disposal: 91 out of a total of 350 of the documents in SAA VI (which covers the period 744–669 BC) concern slave sales, whilst the figure stands at 73 out of the 479 documents in SAA XIV (covering the period 668–612 BC).³ This sale data represents vast

¹ Mendelsohn (1949): 110–11. For slavery in earlier periods of Assyrian history, see Wiggermann (2000); de Ridder (2017); Larsen (2017). On Neo-Assyrian slavery, see most recently Baker (2016a), who begins with a fascinating letter from an exorcist to the Assyrian king Esarhaddon, urging him to test out drugs on a slave before administering them to the crown prince (SAA X 191). Some centuries later we find an Athenian proverb to the effect that risk should be on the Carian’s head (Eur. Cyc. 654; Pl. Lach. 187b; Euthyd. 285c; Cratin. fr. 18 K-A), expressing much the same sentiment about the expendability of slaves. ² These are private archives belonging to members of the court, possibly resident at the palace in Nineveh. Cf. Ahmad and Postgate (2007): xv on the archives from the palace in Kalhu: Although one must always stop to ask whether individuals are acting in a private or an official capacity, there is nothing in the formulation of these texts which suggests that the officials were making purchases either on behalf of the Queen herself, or as the holder of an office to which the property would then be attached. ³ This is not counting the ‘sales of land and people’ documents: see section III below.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

224

Greek Slave Systems

riches in comparison with the rather slender pickings afforded by our classical Greek texts; and that is far from the sum of our Assyrian source material for slavery in general, or slave sales in particular.⁴ The evidence of legal transactions largely derives from archives of various members of the elite; the lower classes are visible mainly through the lens of these documents.⁵ But as is so common with cuneiform archival documents, reconstruction is bedevilled by major blank spots in the record, and these challenge us to think hard about what can really be known about the sources of elite wealth in Assyria during the eighth–seventh centuries BC.⁶ Archives relating to individuals may or may not be complete. Furthermore, it is not known why some courtiers kept their legal records (all of them?) in the royal palace rather than in other properties they might have held elsewhere.⁷ Regarding slave sales, the clay contracts were originally kept in the archives of purchasers, meaning that we know less about vendors and their practices. And the palace-oriented focus of many archaeologists during the past century and a half has skewed the available documentary material towards certain elements of Assyrian society over others.⁸ These various factors warrant a cautious approach to the published texts. This chapter will provide an overview of some of the main evidence, before asking what role slavery may have played in generating the fortunes of elite Assyrians. As with previous chapters, we begin with a consideration of slave status.

⁴ One might note, for instance, the number of slave sale documents in several recently published archives: from the private archives in Assur published in SAAB V: 10 out of 66 documents; from those published in SAAB IX: 14 out of 140 documents; from Tall Šēḫ Ḥ amad (Radner 2002): 74 out of 205 texts; from the north-west palace at Kalhu (Ahmad and Postgate 2007): 21 out of a total of 55 texts. For other recently published slave sale documents, see Jiménez, Adalı, and Radner (2015); Radner (2015a); (2017a) nos. 48, 50. It is worth pointing out that these citations are a sample, not exhaustive, of the slave sale data; even twenty years ago Radner had collected some 363 slave sale texts: Radner (1997): 232–47. Cf. the comparatively modest 33 entries for classical Greece in the conspectus of slave prices given by Ruffing and Drexhage (2008): 321–3. Our Assyrian sources go beyond sale documents, however: see Baker (2016a) for an overview of the range of evidence that can be brought to bear on the problem. ⁵ For useful overviews of the publication history of these archives, see Fales (2001): 81–92; Galil (2007): 23–8. ⁶ A useful methodological primer in interpreting cuneiform documentation: Van de Mieroop (1997). ⁷ Heather Baker points out to me per litteras that ‘This may be a secondary phenomenon, i.e., what were originally private archival documents came to be collected and deposited centrally for some reason.’ ⁸ Cf. Postgate (1979): 194–6 = (2007): 72–4 for some cautionary notes on what kinds of reconstruction this material allows. As Radner (2017b): 228 points out, ‘Postgate 1979 remains the classic survey of the economic underpinnings of the Assyrian empire; most of what was unknown then is still unclear.’ See also Fales (2001): 92–115.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Assyria: Eighth–Seventh Centuries BC

225

I . S L A V E ST A TU S IN AS S Y R I A The basic Assyrian terms for ‘slave’ are urdu and amtu (the latter for female slaves).⁹ As with other Near Eastern slave terms, though, these words are polysemous and can be used in extralegal contexts to articulate hierarchical social relations; one must, therefore, pay close attention to context and genre when translating them.¹⁰ Nevertheless, the legal documents leave no doubt that we are dealing with the ownership and sale of persons, that is, slavery sensu stricto, entailing the same standard features of ownership that pertained in the other regions studied in this book.¹¹ As a fairly typical example of an Assyrian slave sale contract we might consider SAA VI 7, dating from the late eighth century BC and recording the sale of a man named Akbar by his owner Balat ụ -ereš to Mušallim-Issar, village manager of the chief eunuch: Instead of his seal he impressed his fingernail. Fingernail of Balat ụ -ereš, owner of the man. Akbar, his slave—Mušallim-Issar, [village] manager of the chief eunuch, has contracted and bought him from Bala[t ụ -ereš] for 100 minas of copper. The money is paid completely. That man is purchased and acquired. Any revocation, lawsuit, or litigation is void. Whoever in the future, at any time, lodges a complaint, whether Balat ụ -ereš or his sons or his grandsons, and seeks a lawsuit or litigation against Mušallim-Issar, shall pla[ce] 10 minas of silver (and) one mina of gold [in] the lap of Ištar residing in Arbela, [and shall return the mon]ey tenfold to its owners. He shall cont[est] in his lawsuit and not succeed.¹²

We can see here the same features of property law that we have observed elsewhere: the transfer of ownership through sale (right to capital); the indefinite duration of the interest acquired by the new owner (absence of term), and the buttressing of the new owner’s title through a warranty clause exacting a crippling sum on the seller and his forbears should they wish to contest the legitimacy of the sale (right to security).¹³ Other documents from this era, such ⁹ CAD s.v. ardu; amtu. Galil (2007): 188–9; Baker (2016a): 24–5. Amtu is cognate with the Hebrew word amah discussed in Chapter 9: see BDB s.v. ‫אמה‬. ¹⁰ Ahmad and Postgate (2007): ix; Baker (2016a): 24–5. ¹¹ Cf. Baker (2016a): 17: The notion of the ‘chattel slave’ has played a central role in discussions of Ancient Near Eastern slavery, because at various periods slave-sale documents feature prominently among the relevant sources. In spite of Patterson’s argument against defining slavery in terms of property, this approach has persisted (and is likely to continue to do so) because it presents us with a neatly circumscribed group of people whom we can confidently judge to have really been slaves. ¹² SAA VI 7, tr. T. Kwasman and S. Parpola. ¹³ The warranty clause in this document follows a standard format found in most other slave sales in SAA VI and SAA XIV. Some documents deviate from this norm, though: several contain warranty clauses against the slave having epilepsy (SAA VI 83; 103; 134; SAA XIV 8, 195, 243, 244), a concern in some later Greek texts concerning slave sales (Pl. Leg. 11.916a; SEG 47:1026). SAA VI 96 contains a highly unusual warranty clause that requires any disputants of the

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

226

Greek Slave Systems

as records of wills, demonstrate the transmission of slaves to heirs (e.g. SAAB V no. 52; SAAB IX no. 71). Assyrian legal documents also deal with debt bondage and enslavement for debt. Documents of the former sort contain redemption clauses that enable the pledge to regain their former legally unencumbered status upon repayment of the debt (e.g. SAA XIV 93, 101; SAAB V nos. 5, 13, 44, 46, 55). SAA XIV 108 contains a stipulation that the borrower’s son will serve as a pledge for three years in lieu of interest; at the end of the term the debt is to be repaid and the son restored to his former status. SAA XIV 97 provides an instance comparable to the Gortynian rules on debt bondage discussed in Chapter 7 above, for in this case the pledge is not a free person but the debtor’s slave. This document recognizes the original owner’s continuing title to the pledged slave, and stipulates that he shall be redeemed by his owner once the debt has been repaid; the owner bears responsibility if the pledged slave dies or flees. On the other hand, we also find instances of genuine enslavement for debt. A document from 652 BC records the conveyance of a girl named Ahat-abiša to a man named Zabdî; this transfer was made by her father Ubru-Aššur in restitution for a debt of 30 shekels of silver (SAA XIV 85). There is no redemption clause in this document; instead we find the warranty clause standard in slave sale documents.¹⁴ As with the Neo-Babylonian documentation for slave status discussed in Chapter 1, then, there is no reason to posit any fundamental difference in the legal character of slavery between Assyria and the Greek world.

II. ELITE SLAVEHOLDINGS AND THE LOCATION OF S LAVE OWNERSHIP The Neo-Assyrian documentary record, as noted in section I, contains large numbers of slave sale contracts. Although most of these are single purchases, the acquisition of multiple slaves in single transactions is relatively common; and some documents record the purchase of remarkably large numbers of slaves in single, agglomerated lots. A review of the slave purchases of several of the best-documented Assyrian slaveholders will help to illustrate the upper end of the scale of this phenomenon in elite circles.¹⁵

contract’s validity to swallow a mina of plucked wool and a vessel of tanning liquor. On physical penalties, see Postgate (1976): 20. On right to security, see also Radner (2017a) no. 55, mentioning a fine levied for unlawfully seizing the slave woman of a man named Šulmu-Adad. ¹⁴ On pledges and enslavement for debt, see Radner (1997): 368–83; (2001): 280–4; Galil (2007): 199–205. ¹⁵ On Neo-Assyrian elites and administrative hierarchies, see Mattila (2000); Postgate (2007): 331–60; Parpola (2007); Groß and Pirngruber (2014); Richardson (2016); Fales (2017).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Assyria: Eighth–Seventh Centuries BC

227

We have already encountered Mušallim-Issar, village manager of the chief eunuch to Tiglath-pileser III, in the sale of Akbar noted in section I (SAA VI 7). He purchased slaves on at least nine occasions (SAA VI 1–9). One transaction included five slaves (SAA VI 2), another a family of seven slaves (SAA VI 6); in another, the sizeable sum of sixteen silver minas was spent on a number of slaves (the exact number is not preserved: SAA VI 4). The records relating to Šumma-ilani, a chariot driver in the royal chariot corps (SAA VI 34–56), span several decades and are particularly rich in detail on slave purchases. In our earliest text relating to him (from 709 BC) we find a purchase of three slaves (SAA VI 34), and over the following decades his purchases of slaves were to grow in scale. For instance, he bought fifteen slaves in one transaction in 693 BC (SAA VI 40) and a further seven from the same vendor at another point in the same year (SAA VI 41). In another notably large transaction (for which the date is not preserved), he bought a batch of thirteen slaves (SAA VI 52). His purchases also included houses, vineyards, and farmland (SAA VI 37, 50). Overall, ‘he bought at least 50 slaves, including seven families, attested in 12 legal transactions, in the course of about 30 years (709–680 B.C.)’.¹⁶ Another individual, Nabu-šumu-iškun, chariot driver to Sennacherib, expended ten silver minas on one occasion to buy twenty slaves (SAA VI 57) and on another occasion spent thirteen silver minas on an unknown (but probably larger) number of slaves (SAA VI 58). One might also note the palace scribe Nabu-tuklatua, a prominent figure in the palace archives from Kalhu. These archives document his purchase of at least forty-five slaves (mostly male), including, as Ahmad and Postgate have noted, eight slaves in three transactions in a single year, 788 BC.¹⁷ A particularly interesting figure is Šulmu-šarri, whose detailed archive has been unearthed in his mansion (the so-called ‘Red House’) in Dur-Katlimmu, a provincial settlement on the Khabur River.¹⁸ His slave purchases show a different pattern from those of Nabu-tuklatua, for he bought rather more female slaves than males (viz. thirty-five females, fifteen males).¹⁹ Perhaps many of these were involved in textile work, for the remains of a number of looms have been uncovered in the ‘Red House’.²⁰ Now, these archives document purchases of slaves, and thus do not show the extent of prior slaveholding, e.g. slaves inherited from a deceased forebear; they therefore provide minima. The slaveholdings of these elite Assyrians would, then, have at the very least placed them in the same league as the wealthiest fourthcentury Athenians (cf. Dem. 27.9; 37.4; Lys. 12.19; Pl. Resp. 9.578d–e).

¹⁶ Galil (2007): 50. ¹⁷ Ahmad and Postgate (2007): xi. See p. x for a conspectus of the slave sales in these archives. ¹⁸ For Šulmu-šarri’s archive, see Radner (2002): 70–146; Zadok (2009–10). For a useful vignette of Šulmu-šarri, see Radner (2017b): 220–1. ¹⁹ Summary in Radner (2002): 71–2. ²⁰ Radner (2017b): 220.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

228

Greek Slave Systems

What about slave ownership among sub-elite elements of the Assyrian population? Here, direct evidence of purchases of the sort we possess for elites is harder to come by.²¹ However, the relatively extensive evidence of slave sales can aid us here: the fact that we have a great deal of evidence for slave prices (compared to other areas of the ancient world) provides reliable data on how much slaves might cost in the Neo-Assyrian period; and this can be compared to what is known about wages in the lower echelons of Assyrian society, enabling us to gauge very roughly how accessible slave ownership was for individuals at that level. Our evidence for wages is comparatively meagre, rather worse indeed than what we possess for classical Athens.²² But even on a wage of two shekels per month—not the lowest wage attested—a male slave costing a mid-range price of 1 mina (see later in this section) would take 30 months’ wages to afford, i.e. in the region of 900 days’ worth of work. Although much remains unclear about the finer details of Assyrian demography and social structure in this period, if these figures are even roughly representative, slave ownership was well beyond the means of the average wage labourer. Ownership of slaves would, therefore, seem to have been concentrated more densely at the higher end of the wealth spectrum than was the case in classical Athens, which, as we noted earlier, enjoyed relatively high wages, relatively low social inequality (among the citizenry at least), and low slave prices. One should not, however, over-schematize the picture that this evidence presents. The Neo-Assyrian period is punctuated by large-scale campaigns and, concomitantly, gluts of war booty periodically thrown onto the market. In one of his royal inscriptions Aššurbanipal, following a campaign against the Arabs, boasted as follows: [Pe]ople of both sexes, donkeys, camels, [cattle and small cattle] without number I brought to Assyria. [The area] of my whole land, [in] its [entir]ety, they filled as far as it stretches.[Ca]me[l]s I shared out like sheep and goats [t]o the people of Assyria. Within my land one bought a camel at the market gate for a shekel. The ale-wife obtained for one portion, the brewer for one jar, the gardener for a bundle of cress (?), camels and people.²³

Propagandistic exaggeration notwithstanding, major shifts in slave prices associated with gluts of war booty are not difficult to believe. They can be paralleled in the classical and medieval worlds; and perhaps in such ²¹ There are, of course, archives of commercial establishments whose members, though not of the same social and economic standing as the elite, were hardly poor. See, e.g., the division of inheritance in SAAB V 52, which includes eight slaves. However, once we descend further down the wealth spectrum, the position of average or poor Assyrians is far less visible. For urban craftsmen, see Baker (2016b). ²² On wages and hired labour in the Neo-Assyrian period, see Radner (2007), now published in an updated form as Radner (2015b), building on Postgate (1987) = (2007): 167–80. ²³ Weippert (1973–4) Ep. 2, tr. Kuhrt (1995): 519, with one modification proposed to me by Heather Baker.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Assyria: Eighth–Seventh Centuries BC

229

circumstances the option of slave ownership was indeed briefly opened up to a less illustrious stratum of Assyrian society.²⁴ (We need not, however, believe the extent of the boast; and perhaps co-ownership of slaves represented the closest a poorer Assyrian could expect to come to owning a slave.²⁵) Slave ownership in Assyria, then, seems to have been a largely elite concern. However, another point is worth making about Assyrian slave prices, and that relates to the uses to which these slaves may have been put. The nature of our documentation leaves us with a blind spot regarding the use of slaves; for an owner generally did not need to make a legal record on clay detailing the work he assigned to this or that slave.²⁶ But an important pattern in the price records suggests that these slaves were not merely bought to adorn elite households, but that demand for labour—hard, physical labour—was an important motivation in the purchase of slaves. Kyle Harper has pointed out that in most historical slave systems (that is, systems where slavery did not play an important economic role), female slaves cost more on the market than male slaves. The reason for this trend is that, by and large, elite wealth was based on other forms of male labour exploitation, and female slaves were therefore more valuable (albeit in relation to fairly limited labour roles such as domestic service and concubinage). One of the features of slave systems in which slave labour plays a more prominent role in the economy is that one tends to find either parity between the cost of male and female slaves, or that male slaves are more expensive than females. In such systems the profitability of using male slave labour in physically demanding tasks places it at a premium.²⁷ The evidence for slave prices in Assyria bears out the fact that we are not dealing with a society whose elite bought slaves merely for show or for domestic service alone. Two studies during the 1990s collected and analysed the available evidence for Neo-Assyrian slave prices: F. M. Fales discussed the issue in a general article on prices in Neo-Assyrian documents, whilst K. Radner treated it in her Die neuassyrischen Privatrechsturkunden.²⁸ More recently still, a study by G. G. W. Müller has built on the work of Fales and Radner, tracking fluctuations in Assyrian slave prices in the longer term.²⁹ More evidence has emerged since their studies were published,³⁰ but the relatively large samples studied by Fales, Radner, and Müller allow some robust basic conclusions to be set out. On the one hand, there is a fair degree ²⁴ For classical Greece, see, e.g., Xen. Ages. 1.18–21: Agesilaus took so much booty in Asia Minor that it drove prices down considerably; slave children were even abandoned by dealers, so little did they fetch at market. For the medieval world, Runciman (1954): 271–2; 350–1 provides some striking parallels. ²⁵ e.g. Ahmad and Postgate (2007) no. 1; Faist (2009). ²⁶ Cf. Jursa (2010): 234 on the invisibility of slaves in Babylonian agriculture. ²⁷ See Harper (2010): 220; 233–4. ²⁸ Fales (1996); Radner (1997): 230–48. ²⁹ Müller (2004): 189–97. ³⁰ See n. 4 above.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

230

Greek Slave Systems

of variation in the prices, which is not surprising for a commodity as heterogeneous as the human being. On the other hand, a clear pattern emerges. Fales points out that for male slaves in his dossier, 15 per cent cost 2 silver minas or more; 21 per cent cost between 1.5–1 minas; 25 per cent cost 1 mina; and 39 per cent cost less than 1 mina. The higher prices applied exclusively to adults, whilst the lower end of the range contained a number of boys. As for females, 24 per cent cost between 1.5 and 0.75 minas; 45.5 per cent cost between 0.75–0.5 minas; and 30.5 per cent cost less than half a mina.³¹ The clear trend is towards a significantly higher average price for males than females. Radner and Müller came to a similar conclusion in their studies.³² These results show quite firmly that male slave labour was at a premium in Assyria at this time.

I I I . L A B O U R IN TH E AS S YRI AN C OU NTRYS I D E ‘In various parts of the empire large private estates were built up by members of the palace sector, in addition to large tracts appropriated to members of the royal family and high officers of the state.’³³ If we are to view these estates as the main source of elite income, our next task must then be to gauge the kinds of labour that were available to work on them. A glance at the evidence for these labourers—and recent debates among Assyriologists³⁴ concerning the interpretation thereof—suggests that despite the sometimes large numbers of slaves owned by members of this echelon of society, slavery probably did not provide the bulk of their income. To place this problem in context, we must look at the broader agrarian economy of the Assryian heartland. (This region occupies a Mediterranean climate zone; unlike Babylonia to the south, Assyrian agriculture was based primarily on rainfall, irrigation by canal being less important than in the south, though a number of these have been found in recent surveys; and there is a degree of climatic variation within this zone.³⁵ Crop types, too, were of a similar profile to the Mediterranean: wheat, barley, and vines,

³¹ Fales (1996): 30. ³² Radner (1997): 248, ‘Zwischen dem durchschnittlichen Kaufpreis für Männer und Frauen gibt es Unterschiede. Der durchschnittliche Kaufpreis für einen Mann is besonders dann, wenn er in Silber berechnet wird, um einiges höher als der für eine Frau.’ Müller (2004): 191–2 accepts Radner’s conclusions in these regards. ³³ Postgate (1979): 214 = (2007): 92. For an overview of Neo-Assyrian land tenure, see Postgate (1989) = (2007): 181–92. ³⁴ Van Driel (1970); Fales (1973); Postgate (1974) = (2007): 27–45; Postgate (1989) = (2007): 181–92; Fales (1990); (2001): 170–8; (2009–10). ³⁵ See Baker (2016b) for further discussion.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Assyria: Eighth–Seventh Centuries BC

231

incorporated alongside animal husbandry.³⁶) With regard to the issue of labour in this landscape, three key issues must be taken into consideration: (i) First, the question of deportees to Assyria and their status (or statuses) subsequent to their arrival. If the figures cited by Assyrian kings in their royal inscriptions are to be believed, during the three centuries of the Neo-Assyrian period over four million people were uprooted from their homelands and relocated to other parts of the empire, either as a result of war or planned population relocation.³⁷ Even if we trim down these figures, suspecting boastful exaggeration on the part of the kings, it is clear that we are dealing with enormous movements of people. B. Oded has pointed out that in 85 per cent of the cases where we know the final destination of the deported population, Assyria proper was the terminus, thus swelling the population in the vicinity of the principal cities of Assur, Kalhu, Nineveh, and Dur-Šarrukin.³⁸ The question that naturally follows from this observation concerns the legal condition of these deportees once they arrived in Assyria: did they acquire a uniform status, or were there multiple possibilities? How likely a fate was genuine enslavement for the victims of deportation? In many cases it seems that these populations relocated to the Assyrian heartland were not treated as a servile class: family and community ties were preserved, and in a number of inscriptions these groups were explicitly ‘counted as Assyrians’ by the kings who resettled them.³⁹ The rationale here seems to have been to increase the density of the local population; left to their own devices, these smallholders would hopefully flourish, greatly increasing the tax and corvée resources at the disposal of the state.⁴⁰ A hint of this kind of outcome is preserved in the Hebrew Bible at II Kings 18:19–35 (cf. Isaiah 36:4–20), concerning Sennacherib’s siege of Jerusalem in 701 BC. The Assyrian king’s message to the Jerusalemites allegedly spoke of deportation to Assyria as a positive prospect: ‘Make your peace with me and come out to me, so that you may all eat from your vines and your fig trees and drink water from your cisterns, until I come and take you away to a land like your own, a land of grain and vineyards, of bread and wine, of olive oil and honey, so that you may live and not die’ (II Kings 18:31–2).⁴¹ To be sure, not all deportees were so fortunate. We noted in section II in relation to Aššurbanipal’s campaign against the Arabs the large number of war captives brought into Assyria; these were certainly enslaved, and apparently drove down the price of slaves on the market.⁴² Deportees might also be divided up and assigned to various persons and institutions: in an inscription recording the fate of Šubrian captives taken in a campaign to the north, ³⁶ ³⁷ ⁴⁰ ⁴²

Postgate (1989): 141–2 = (2007): 181–2; Radner (2017b): 212. Oded (1979): 19–22. ³⁸ Oded (1979): 28. ³⁹ Oded (1979): 24–5; 77; 81–91. Oded (1979): 59–74. ⁴¹ See Gallagher (1994). See Oded (1979): 101 for this and other examples.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

232

Greek Slave Systems

Esarhaddon notes how ‘I distributed the re[st of them] like sheep and goats among my palaces, my nobles, the entourage of my palace, and [the citizens of Ninev]eh, Kalhu, Kalzu and Arbela.’⁴³ It is far from clear whether these captives became the alienable property of the various nobles to whom they were granted; some may have been, but the existence of documents recording ‘sales of land and people’ (see ii below) allows the alternative possibility that these deportees were attached to the land as bound tenants.⁴⁴ Oded’s conclusions in his study of deportees in the Neo-Assyrian Empire are worth repeating: the socio-economic and legal status of the deportees was not uniform and their conditions were not identical. There were masters and dependants, full freemen and chattel slaves, soldiers and civilians, labouring freemen and labouring dependent persons, townsmen and villagers, tenants and glebae adscripti. The rights and duties of the individual deportee were determined by a wide range of circumstances.⁴⁵

(ii) Second, a number of sale documents scattered among the purchase records of elite Assyrians detail acquisition of land and people in single, agglomerated transactions.⁴⁶ Prima facie, one might assume these sales to have been of plots of cultivable land and the slaves that worked them. However, there are problems with this interpretation. Oded argued that in these documents the people in question were bound to the land being sold, and were not separately alienable.⁴⁷ Oded’s student G. Galil has, furthermore, recently drawn attention to the prices associated with these transactions: if they were indeed sales of land and slaves, one would expect much higher prices (for we have many examples of separate land sales and slave sales with which to compare these documents).⁴⁸ Like Oded, Galil concludes that the individuals mentioned in these transactions were not slaves, but bound tenants. Since they were bound to the plot of land in question, they could be considered in a sense as part of the inventory of the plot when it was sold; but since they were not slaves sensu stricto, that is, not separately alienable, they did not contribute as greatly to the capital value of the sale as would have been the case were they slaves. (iii) Third, the status of workers in the so-called ‘Harran census’ (SAA XI 201–20). This is the term given to a series of tablets discovered at the royal palace in Nineveh and dated to the reign of Sargon II.⁴⁹ They list landed properties in the vicinity of Harran (well to the west of the Assyrian heartland

⁴³ Leichty (2011) no. 33 iii lines 16–22, cited in Baker (2016a): 23. ⁴⁴ Oded (1979): 114. Baker (2016a): 23 comments in relation to this document, ‘these people had no say in their fate, but it remains completely open as to whether any of them actually became enslaved. Personally I doubt that many of them were.’ ⁴⁵ Oded (1979): 115. ⁴⁶ e.g. SAA VI 50; 65, 90, 91, 112, 123, 129, 153, 163, 169, 269, 315; SAA XIV 1, 2, 22, 36, 198, 207, 229, 254, 265, 345, 355, 401. ⁴⁷ Oded (1979): 95–6. ⁴⁸ Galil (2007): 207–10. ⁴⁹ Fales (1990): 88.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Assyria: Eighth–Seventh Centuries BC

233

proper) and the personnel dwelling and working thereon. The purpose of these records has been much debated; Postgate convincingly argues that since (a) their findspot indicates that they were composed by state bureaucrats and (b) they do not describe contiguous estates but a scattering of properties in different areas, they cannot be simple cadastral records for the purposes of tax collection: such records would have been organized on the provincial, not state, level; and would have dealt with contiguous properties. They are more likely, therefore, to list tax-exempt properties of various wealthy Assyrians.⁵⁰ Naturally, the question arises whether the workers described in the Harran census were of a uniform legal status, or comprised a mixture of those statuses we know from other sources, i.e. slaves, free tenants, deportees, debt bondsmen, or bound tenants of the sort probably described in the ‘sales of land and people’, perhaps corresponding to those deportees (or their descendants) noted in (i) above that were assigned to Assyrian nobles as gifts of the king.⁵¹ The question is far from decisively resolved. Galil, for instance, has recently made a case that rural workers in the Harran census were voluntary tenants.⁵² Fales’ response to Galil’s proposal perhaps best captures the current state of uncertainty: Galil’s reconstruction is certainly timely, in an age which is bringing to light the many implications of patrimonialism and clientship in the Ancient Near East, but this writer does not believe it may be applied here. While having long renounced the totally opposite, and equally sweeping, view that the Assyrian countryside was manned in its entirety by glebae adscripti, he however still believes that the specific selection of landed holdings owned by, or assigned in prebend to, members of the ‘master class’ which formed the so-called ‘Harran census’ was manned by menial labour forces which had no way of choosing, and little possibility of renouncing, their condition of long-term bondage to the land— whether because they had been enslaved through debt, had been pledged together with the properties, or were deportees installed on the farms.⁵³

⁵⁰ Postgate (2007): 30–2; followed by Fales (2001): 174. ⁵¹ Oded (1979): 72–3; 91 notes the presence of Gambulean captives in one of these documents (i.e. SAA XI 207). ⁵² Galil (2007): 224. ⁵³ Fales (2009–10): 167. This represents a modification of Fales’ earlier position (set out in Fales 1973) in line with the criticisms of Postgate (1974): 230–33 = (2007): 32–5. cf. Fales (2001): 174: ‘In primo luogo, il personale registrato doveva essere legato in maniera inscindibile e permanente ai lotti; si trattava cioè di veri e propri servi della gleba, pervenuti a tal condizione per debiti . . . o per deportazione.’ Fales goes on to note (pp. 177–8): Se si nota, infatti, che le donne in genere non sono abbondanti in questa popolazione, e si tiene presente una serie di contratti coevi, che coinvolgono la cessione di una madre con figlia in tenera età . . . , si potrebbe sospettare addirittura che i proprietari abbiano programmaticamente estratto elementi femminili da questi gruppi di servi agricoli, per dotarsi di schiave domestiche o di mogli per altri schiavi in loro possesso.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

234

Greek Slave Systems

The debate will doubtless continue. When it comes to the question of the relative importance of different kinds of labour on the estates of the elite, the cautious, tentative position of a distinguished Assyriologist such as Fales is telling, for it reflects the nature of our evidence. Despite the ostensible riches that the Assyrian documentary record affords, it is devilishly difficult to turn this data into a generalized picture; and reactions to Galil’s recent attempt to produce statistics from these records underscore the challenges involved.⁵⁴ Whilst the evidence for slavery and the economy in classical Attica rarely shows the particular in as startling detail as the Assyrian evidence, its varied genres at least provide the materials for formulating generalizations. The downside of the Assyrian evidence is a near total opacity regarding those parts of the overall jigsaw that are missing, and a lack of generalizing statements about labour conditions that allow us to gauge how typical this or that form of labour was on elite landholdings. What we can conclude at this stage is that elite Assyrians seem to have drawn their fortunes primarily from large estates scattered throughout the countryside, often in very disparate locales.⁵⁵ These grandees, as we have seen, sometimes owned large numbers of slaves, apparently with holdings at least on the scale of the richest Athenians of two or three centuries later. However, these slaveholdings were certainly augmented with other forms of labour; and these latter may well have eclipsed the significance of slavery in terms of their economic contribution to the wealth of the Assyrian elite.

⁵⁴ Radner (2008); Baker (2009); Fales (2009–10).

⁵⁵ Fales (1990): 114.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

11 Babylonia The Seventh–Fifth Centuries BC

Urban slavery in classical Athens has a reputation for being particularly innovative and advanced in its manifold methods of exploitation.¹ Yet almost all of these apparently innovative Greek techniques of exploitation were prefigured over a century earlier in the urban centres of Babylonia. Athenian apophora practices, for example—an arrangement whereby slaves worked independently of their owners and paid them a fee from their wages, retaining the remainder—have their earlier counterpart in Babylonian mandattu arrangements, which functioned in a similar manner.² Other incentives, such as allowing slaveholding by slaves, are also attested in Babylonia, as are manumissions with attached obligations for the freedperson to remain with and serve their ex-owner, like the sort that we can see in the wills of the philosophers and in the paramone clauses of numerous Delphic manumissions.³ Wealthy and successful slaves at Athens such as Pasion and Phormion were foreshadowed by successful Babylonian slave businessmen like Nergal-reṣua and Madānu-bel-uṣur.⁴ The latter, indeed, became entangled in a dispute whose motivations anticipate by a century the complaint of the Old Oligarch ([Xen.] Ath.Pol. 1.10) about Athenian slaves who dressed as well as citizens; for in October 524 BC Madānu-bel-uṣur was confronted by a jealous, less successful business associate, Nabû-ēt ị r, a free man irked by his rival’s flashy clothing and superior airs. Taking a key, he ripped Madānu-bel-uṣur’s fine woollen garment apart and beat him up, shouting ‘This is the word of ¹ e.g. Ober (2015): 232: ‘historically innovative forms of chattel slavery’. ² Dandamaev (1984): 379–83. Jursa (2010): 237 discusses an unpublished document from the British Museum dating to the reign of Nabopolassar which records the division of an inheritance, mentioning slaves ‘together with their property’. This is very similar to what we find in IG I³ 426.14; 24–39 in relation to the slave shoemaker Aristarchus. ³ Slaveholding by slaves: Dandamaev (1984): 372–8; manumission: ibid: 438–55; Wunsch and Magdalene (2014). ⁴ Biographies of these slaves in Dandamaev (1984): 345–71; Wunsch (1993): 43–8. See also Abraham (2004): 72–7; Head (2010): 67–78; Wunsch and Magdalene (2012).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

236

Greek Slave Systems

the king: “A slave does not wear a loincloth of first-class wool around his waist!”’ (Camb. 321).⁵ As at Athens, the uses of slaves in the Babylonian cities reflect the needs of a complex economy organized around the provision of urban markets and the enrichment of elite families with elaborate and varied commercial interests. Yet, although the institutional basis of slavery and the methods of exploiting slaves in sixth-century Babylonia were similar in many ways to that of Athens a century later, slavery never developed economically there to the same degree. Why did Babylonia never become a ‘slave society’? In this chapter we will look at some of the factors that might account for the differential development of slavery in these two locales.

I. SO URCES AND SCHOLARSHIP The Neo-Babylonian period begins with the accession of Nabopolassar to the Babylonian throne in 626 BC; he played an instrumental role (alongside his Median allies) in the downfall of the Assyrian Empire, most graphically marked by the sack of Nineveh in 612. The so-called ‘long sixth century’ that followed Nabopolassar’s accession is illuminated by an unusually rich documentary record, deriving in the main from a number of private archives and the archives of two major temples (Ebabbar in Sippar; Eanna in Uruk⁶); the palace sector, by comparison, is poorly attested.⁷ The coming of the Persians in 539 BC did not fundamentally change socio-economic conditions; but there is a real break following 484 BC, when the Babylonians revolted unsuccessfully against Xerxes. The repercussions that followed included a significant shift in the make-up of the Babylonian elite, and this is reflected in the documentary sources.⁸ The famous Murašû archive, focused on the fifth century BC, displays some features that differ from sixth-century conditions, including prominent interactions with wealthy Persians and the management of their Babylonian estates, as well as the business opportunities afforded by the Persian policy of exacting tax in silver on a population whose rural

⁵ See Wunsch and Magdalene (2012). For the kind of dealings vis-à-vis free persons during the same year that may have earned Madānu-bel-uṣur a reputation for bullish behaviour, see the discussion of Camb. 329 in Dandamaev (1984): 345. ⁶ Ebabbar: see Bongenaar (1997); Jursa (2010): 323–48; 510–40; Eanna: see Beaulieu (2005); Janković (2010); Kleber (2008; 2010). The prebendary archives from Borsippa shed much light on the functioning of the Ezida temple there: Waerzeggers (2010). ⁷ Detailed conspectus of Neo-Babylonian archival sources in Jursa (2005). ⁸ Waerzeggers (2003–4); Wunsch (2010): 40–1.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Babylonia: Seventh–Fifth Centuries BC

237

elements had limited access to precious metals.⁹ Neo-Babylonian archival sources provide a much richer vein of material for the reconstruction of economic life than what is available for the Neo-Assyrian Empire; what is more, the topic of economic history in Babylonia during this period has been worked on more intensively and for a longer period of time.¹⁰ Central to modern interpretation of the role of slavery is M. A. Dandamaev’s Slavery in Babylonia, from Nabopolassar to Alexander the Great (626–331 BC), which has been supplemented (and in certain areas superseded) by a surge of work in the last two decades on the economy of Babylonia in the first millennium BC, as well as several specialist studies on aspects of slavery.¹¹ As in previous chapters, we will focus mainly on slaveholding in the ‘private sector’, particularly among its elite elements. The urban elites of the Babylonian cities, visible through their private archives, have been generally treated in two categories, viz. prebendaries and entrepreneurs. Prebendaries are so called because they held prebends, that is, benefices from local temple institutions: they performed specific services for the temple and received an income in kind in return; these prebends were often heritable and, alongside land ownership (especially of date groves) in the vicinity of the city, generally allowed their holders to enjoy a comfortable rentier lifestyle.¹² Entrepreneurs, on the other hand, were businessmen involved in various commercial activities, often working on credit. These activities included commercial agriculture, organizing the transportation and sale of goods from the countryside to the city, and tax farming.¹³ (The categories of prebendary and entrepreneur are Idealtypen to a certain degree: some prebendaries had business interests of the entrepreneurial sort; but the distinction is useful nonetheless.¹⁴) In what follows (section II) we shall begin by looking at the ⁹ A classic study of the Murašu firm is Stolper (1985). We have already encountered the problems that the levy of taxes in silver caused for the inhabitants of fifth-century Yehud: see Nehemiah 5:1–13 with Chapter 9, section IV, pp. 220–1 above. ¹⁰ Jursa (2010) provides a detailed synthesis of the Babylonian economy during the long sixth century. ¹¹ Dandamaev (1984). The most significant change of outlook since Dandamaev’s book is the reclassification of širkus, temple oblates. Dandamaev classifies these as ‘temple slaves’ (see Dandamaev 1984: 469–557), but recent work has questioned this approach; for although these individuals in some sense ‘belonged’ to the temple institution, they could not be sold, enjoyed certain legal rights, and were paid salaries. See Ragen (2006); Kleber (2011); Wunsch and Magdalene (2014). Recent specialist studies of Neo-Babylonian slavery: Baker (2001); Head (2010); Magdalene, Wells, and Wunsch (2008); Magdalene and Wunsch (2011); Wunsch and Magdalene (2012; 2014); Oelsner (2014); Tolini (n.d.). ¹² Description of numerous such archives in Jursa (2005); overview of the economic position of prebendaries in Jursa (2010): 155–65; cf. 282–6. For in-depth discussion of the prebendary economy, see van Driel (2002): 33–151. For a lucid introduction to the daily life of temples and role of prebends, see Waerzeggers (2011). ¹³ For an excellent overview of Neo-Babylonian entrepreneurs, see Wunsch (2010). See also Jursa (2010): 286–94. ¹⁴ Jursa (2010): 285–6.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

238

Greek Slave Systems

distribution of slave ownership in Babylonian society. Section III sets out three thumbnail sketches of Babylonian propertied families: one prebendary family, the Nappāḫus, and our two best-documented entrepreneurial families, the Egibi family (whose affairs are mainly attested for the sixth century) and the Murašû family (whose affairs are mainly attested for the second half of the fifth century). These case studies will provide a flavour of the kinds of incomegenerating activities in which elite Babylonians were involved. After sketching the activities of these families, we will turn to look at the role of slavery and endeavour to understand why slavery did not develop to as significant a degree in the Babylonian economy as it did in Athens.

II. THE LOC ATION OF SLAVE OWNERSHIP IN NEO-BABYLONIAN SOCIETY The evidence for the location of slave ownership in Babylonian society displays points of similarity and difference with the Neo-Assyrian sources discussed in Chapter 10. In terms of similarity, we possess a relatively rich record of slave prices that can be contextualized in relation to the wages of lower-class workers. As for the archives of individual Babylonian families, the two most forthcoming genres of document are those concerning dowries and those detailing the division of inheritance among the children of deceased individuals. Documents concerning dowries are less useful than inheritance records, for they obviously do not provide us with complete records of the family in question’s slaveholdings; and the number of slaves granted to daughters is not necessarily a reliable indication of a family’s total slaveholdings.¹⁵ Documents recording the division of property among heirs are particularly valuable because they provide a snapshot of the family’s property at a specific moment in time.¹⁶ Slaveholding seems to have been standard among the Babylonian propertied classes, and in some cases the slaveholdings of wealthy families were of a remarkable size. At the very top end of the scale, Dandamaev notes Dar. 379, relating to the division in 508 BC of the estate of Itti-Mardukbalatụ of the Egibi family among his three sons: the eldest received inter alia ¹⁵ The Itti-Marduk-balat ụ branch of the Egibi family is a case in point: Roth (1991) notes eight recorded dowries for women of this family, five for women born into the family and three for women who married into it. No dowry contains fewer than two or more than ten slaves. However, Dar. 379, recording the division of Itti-Marduk-balat ụ ’s property, shows that he owned around hundred slaves: Dandamaev (1984): 212–13. ¹⁶ As was usual in the Near East, the eldest son in the Neo-Babylonian period was granted a double share of the inheritance. Wunsch (2003): 130–1; 144–5, however, shows an additional rule to have been in force: if there were four or more brothers, the eldest son gained half of the estate, the remainder being shared among the other sons. (Daughters received nothing.)

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Babylonia: Seventh–Fifth Centuries BC

239

forty-four slaves, the younger sons between them fifty-one slaves; and numerous other slaves belonged to these brothers before the inheritance was divided. This family, then, owned at least a hundred slaves.¹⁷ Jursa has also drawn attention to TCL 13, 223, mentioning 118 slaves inherited by a younger son of the Bābāya family; this family’s total holdings must have amounted to several hundred slaves.¹⁸ At a rung further down the wealth ladder, Dandamaev has noted patrimonies including thirty (TCL 12, 43), twenty-seven (Cyr. 161), and twenty-five (Dar. 429) slaves.¹⁹ The majority of cases, though, are found at a further remove: ‘apparently it was not uncommon for persons of moderate means to own three to five slaves’.²⁰ In terms of the size of elite slaveholdings, the pattern is not dissimilar to that of the propertied classes in classical Athens, with two qualifications: (i) the wealthiest Babylonian families appear to have owned rather more slaves than their later Athenian counterparts; and (ii) the demography and distribution of wealth in sixth-century Babylonian cities is far harder to quantify than for fourth-century Athens.²¹ (One presumes that Babylonia was rather less egalitarian than Athens.) As we have noted before, however, the impact of slavery must be assessed in relation to the labour market as a whole. The prices of slaves in Babylonia in relation to wages provide a point of contrast with Athens. Whereas in classical Athens the average price of a slave amounted to around 150–200 days’ wages, a male slave in Babylonia during the reign of Nabonidus cost on average around 50 shekels, i.e. 375–750 days’ wages (assuming a monthly wage of 2–4 shekels).²² This does not mean that exploiting slave labour was necessarily unprofitable, but that the capital outlay for investing in a slave placed slave ownership beyond the reach of most families, like the situation in Assyria in the preceding two centuries.²³ (To this one might add Bresson’s observation that the demos in Greek cities such as Athens had little to no tax burden, unlike their counterparts in Babylonia.²⁴) Slave ownership in Babylonia, then, was largely an elite phenomenon and did not filter far down into the ranks of the working free population, as was the case at Athens. We will return to this issue in section IV of this chapter, for, when contextualized in terms of the broader labour market, it is of key importance to understanding the overall importance of slavery in Babylonia.

¹⁷ Dandamaev (1984): 212–13. ¹⁸ Jursa (2010): 233. ¹⁹ Dandamaev (1984): 215–16. ²⁰ Dandamaev (1984): 216; cf. Jursa (2010): 233. ²¹ Cf. Chapter 8, section I, pp. 168–9 above. On the distribution of wealth in Babylonia, Baker (2014) provides an introduction to the important proxy data provided by house sizes. ²² Jursa (2010): 745 n. 3780. See also Dandamaev (1984): 200–6, and Oelsner (2014) for the late Achaemenid and early Seleucid periods. ²³ The point on profitability is crucial: see Dandamaev (1984): 246–8. Despite the comparably unfavourable relationship between slave prices and wages, it is worth noting that the disparity was even more pronounced in the US South: see Fogel and Engerman (1974): 72–5. ²⁴ Bresson (2016): xxv.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

240

Greek Slave Systems

III. THREE THUMBNAIL SKETCHES How did these elite slaveholders make their money? Here, I set out three thumbnail sketches in relation to some of the best-attested propertied families of the period in order to give a flavour of the various income-generating options. We begin with the Egibi family, perhaps the most famous family of Babylonian entrepreneurs. Our second sketch concerns the Nappāḫu family, coeval with the Egibis: this family, however, drew its income in a different manner, viz. from temple prebends (though, like the Egibis, they did also own real estate), and was far less wealthy. Our third sketch is of the Murašû family and provides a slightly different perspective, for it relates to the post-Xerxes period, when, as we noted in section I, p. 236 above, the composition of the Babylonian elite seems to have changed; and it takes us out of Babylon itself, to the city of Nippur in the south. a. The economic dealings of the Egibi family are known to a remarkable degree of granulation due to the survival of a substantial part of its large archive, spanning five generations from the reign of Nebuchadnezzar II to that of Xerxes.²⁵ The family’s head in the first generation, Šulaja, was a parvenu, and laid the foundations of the family’s wealth. He engaged in ḫarrānu partnerships in which his business partners provided the capital whilst he undertook the work and organization, which in his case involved the purchase, transportation, and sale of barley, dates, onions, and wool: these were bought up from the countryside, moved via canal, and sold in city markets.²⁶ Šulaja’s son Nabu-aḫḫē-iddin continued his father’s business as well as working as a scribe for influential Babylonian clients, eventually serving as a judge under Nabonidus; the profits of his business were invested in land, houses, and slaves. Most remarkable is Nabu-aḫhē̮ -iddin’s purchase (alongside his business partner) in 559 BC of a large estate running for some 2.3 km along both sides of the New Canal, on the outskirts of Babylon near the Enlil Gate.²⁷ This estate was sublet to tenant farmers on a sharecropping basis, providing the family with a secure income stream. This was, however, just the largest of a number of landholdings the Egibis had around the city and from which they drew rental income.²⁸ The affairs of the family were handled in the third generation of the archive by Nabu-aḫḫē-iddin’s son Itti-Marduk-balat ụ , who appears to have invested considerable effort in currying favour with Babylonia’s new Persian rulers in order to maintain business relations with state institutions such as the ²⁵ For an overview of the family and its activities, see Wunsch (2007). For an overview of the archive, see Jursa (2005): 65–6. Detailed studies of aspects of the family’s dealings in Wunsch (1993; 2000); Abraham (2004). ²⁶ On ḫarrānu partnerships, see Jursa (2005): 42–3; (2010): 206–13; Wunsch (2010): 51–3. ²⁷ See Wunsch (2000): 65–82; for a briefer overview, see Wunsch (1999): 401–7. ²⁸ For a detailed study of the Egibis’ property, see Wunsch (2000).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Babylonia: Seventh–Fifth Centuries BC

241

palace and army. After his death, his son Marduk-nāṣir-apli took over the family business; the archive comes to an end in the next generation, during the time of his son Nidinti-Bēl in the early years of Xerxes’ reign. In sum, the family drew its income mainly from commodity trading, rental payments on agricultural land and urban property, tax farming, and slave dues (mandattu).²⁹ b. The Nappāḫu family provides a useful case study of elite Babylonians whose income derived from prebends (as well as several other interests); Heather Baker has recently treated its archive in a detailed monograph.³⁰ The archive itself mainly relates to the second half of the sixth century and the first two decades of the fifth century, and principally concerns two generations of the family, that of Iddin-Nabû and that of his son Šellebi. Like the Egibis, the Nappāḫus were based in Babylon itself. Their fortunes were rather more modest, though, than those of the Egibis; and their prebends were not held in Babylon’s foremost temple, the Esagila temple of Marduk, but in the cults of several minor deities (especially Išḫara).³¹ In return for performing the specified duties required of prebend holders, they received emoluments in kind. The Nappāḫus also invested their money in property in and around Babylon: they owned at least four houses, several of which were leased to tenants; and through Iddin-Nabû’s marriage to Ina-Esagil-ramât (of a branch of the Egibi family) they acquired orchards in three locations which were leased out on a sharecropping basis.³² The family owned a modest number of slaves; most of them appear to have been employed in maintaining their household.³³ c. After the Egibis, the next most important and well-known private entrepreneurial archive is that of the Murašû family. Some of the archive’s earlier texts concern Enlil-ḫatin, son of Murašû; most of the texts concern the dealings of his brother Enlil-šum-iddin and Enlil-ḫatin’s son, Rimut-Ninurta.³⁴ The archive charts business activities under the Persians during the second half of the fifth century BC and relates to activities focused on the city of Nippur in southern Babylonia, where the archive was unearthed. Key to the system of Persian control of Babylonia was a system of land tenure— already attested in Neo-Assyrian and Neo-Babylonian times—in which the Crown granted estates to their holders in return for military service; older scholarship often referred to these as ‘fiefs’. The sources most commonly name them using terms such as ‘bow land’ (bīt qašti), ‘horse land’ (bīt sīsî), and ‘chariot land’ (bīt narkabti).³⁵ Over time, monetary payments often came to be offered instead of military service; and this provided a business niche that was exploited by the Murašûs. The main orientation of their activities was the management of such lands; for they leased them from their holders in return for paying rent and taxes, and in ²⁹ ³¹ ³² ³⁴

Abraham (2004): 11–13; Wunsch (2010): 47–50. On the family’s prebends, see Baker (2004): 34–46. Houses: Baker (2004): 47–62; orchards: ibid: 63–9. Stolper (1985): 18–20. ³⁵ Stolper (1985): 24–7.

³⁰ Baker (2004). ³³ Baker (2004): 70–3.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

242

Greek Slave Systems

turn sublet the land to tenant farmers.³⁶ A second strand of their business involved predatory moneylending on an antichretic basis to farmers in distress.³⁷ Third, and like the Egibis, the Murašûs seem to have dealt with the marketing of produce in the city, which was bought for silver; they then used this silver to exploit another business niche, i.e. providing a tax-paying service (since the Persian king levied tax in silver) for landholders without easy recourse to specie.³⁸

IV. THE USES OF SLAVES To what uses did families such as these put their slaves? As is the case for NeoAssyrian documentation, the ownership of slaves by this or that individual might be clear to see through purchase records or (more rarely) records of inheritance divisions, but the nature of their employment is often opaque. The basic problem is that the majority of the evidence available consists of legal records impressed on clay, which in terms of labour relations tend to capture arrangements in which the parties had rights and duties vis-à-vis one other that needed to be written down (e.g. contracts of hire, tenancy, promissory notes, etc.). Since slavery as an institution is based on the fact that the slave has no rights, only duties, we face an epistemological problem, a missing jigsaw piece that is a by-product of the Babylonian epigraphic habit and the institutional features of slavery, compounded by the lack of literature on slavery parallel to what we find in Athens in, e.g., comedy, Xenophon’s various tracts, and the Attic orators.³⁹ The mention of slave occupations generally occurs obliquely, when the slave’s work interconnects with the sort of legal matters that are routinely recorded on cuneiform documents. The hiring out of slaves to third parties could create a tablet trail, as could the apprenticing of slaves to master craftsmen to learn this or that skill.⁴⁰ When slaves were employed as business agents of their owners, they acted as parties to various transactions, and therefore slaves in this occupation are more visible.⁴¹ When slave

³⁶ Stolper (1985): 27; 53–103. ³⁷ Stolper (1985): 28; 104–24. ³⁸ Stolper (1985): 28; 149–50. ³⁹ Cf. Van de Mieroop (1997). On this problem in relation to Neo-Babylonian archives, see the comments of Jursa (2005): 57–9 and the excellent diagram in Baker (2004): 6. Regarding slaves, Jursa (2010): 233 notes that ‘most female slaves were apparently employed in the house. Neither they nor their activities tend to leave many traces in the written record; they appear only when they are alienated or handed over as part of dowries.’ Cf. Dandamaev (1984): 249 and Baker (2001): 23. ⁴⁰ Hiring out slaves: Dandamaev (1984): 112–36; slave apprentices: ibid: 279–307; Hackl (2010): 5.6.4.5. ⁴¹ Dandamaev (1984): 308–19; Jursa (2010): 237–40; Head (2010) passim.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Babylonia: Seventh–Fifth Centuries BC

243

businessmen became embroiled in legal disputes, records would be produced to document the trial’s outcome.⁴² When slaves became involved in contractual obligations to free citizens, the same might occur.⁴³ But it may be methodologically prudent to suppose that much of the work of slaves went undocumented; and naturally, one wonders what the hundred or so slaves of the Egibis, or the several hundred slaves of the Bābāyas, were up to. The higher average price of male slaves than females suggests that male labour roles were more valued overall, which cuts against any view that these slaves were merely bought for, for example, domestic duties or conspicuous display.⁴⁴ There was an economic motive, then, for buying slaves. As we noted earlier in this section, entrepreneurial families such as the Egibi and Murašû families gained much of their wealth through rents from tenant farmers or the management of land in and around Babylonian cities; and prebendaries frequently drew income from the produce of estates in similar locations. The question of agricultural labour is, therefore, crucial to our understanding of why slavery did not develop as much in Babylonia as it later did at Athens. Dandamaev argued persuasively that slavery played a limited role in agricultural production; the evidence for sharecropping tenancy arrangements is certainly very extensive and shows the ubiquity of the practice.⁴⁵ Jursa has recently come to a similar position, but with an epistemological caveat: slaves working in agriculture on a non-contractual basis:⁴⁶ are necessarily invisible in our documentation—lease contracts and so forth almost never mention the lessee’s labour force, for instance. In other words, Dandamaev’s argument is to some extent an argumentum e silentio. There must have been a role for privately owned slaves, either slaves of the landowners or slaves of the lessees, in the cultivation of private estates.⁴⁷

⁴² See, e.g., Dar. 509 with Dandamaev (1984): 390–1. ⁴³ See, e.g., Camb. 330 with Joannès (1992): a man named Marduk-iqišanni was required to deliver certain items of inventory to a female slave belonging to the Egibi family who was setting up a tavern in Kiš. See further Tolini (n.d.). ⁴⁴ Dandamaev (1984): 204; on the significance of the relationship between the prices of male and female slaves, see Harper (2010): 220; 233–4. At 234 n. 89 Harper regards the NeoBabylonian pattern as an exception to the general trend that in ‘societies with slaves’ as opposed to ‘slave societies’, female slaves are on average more expensive; but perhaps the problem here lies in the oversimplification of these Idealtypen. Babylonia in this period was not a ‘slave society’ in the Finleyan sense; but it did display many elements in common with fully developed slave societies such as classical Athens. If one were to use the ‘intensification’ model of Lenski (2018a), then growth in the market price of male slaves relative to female slaves might be seen as an indication of the development of a slave system in the direction of the ‘slave society’ archetype, without having fully achieved that status. ⁴⁵ Dandamaev (1984): 272–4. ⁴⁶ There are some cases where slaves appear in the documents as tenants: Dandamaev (1984): 252–70. ⁴⁷ Jursa (2010): 234–5.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

244

Greek Slave Systems

Even so, given the positive evidence for the frequency of legally free tenant cultivators, Jursa does not view slaves as a significant labour force on these lands, and he accepts Dandamaev’s argument as to why this was: the input costs of slave labour (i.e. purchase and supervision) in comparison to hired labour or tenancy were very high, and landholders (or those leasing plots of land from their owners) found it expedient to fall back on the large body of tenant farmers as a more flexible and less costly option.⁴⁸ As at Athens, slaves might be trained as artisans and set to work bringing in a steady income stream for their owners (Xen. Oec. 7.41; Eq. 2.2; Aeschin. 1.97). Apprenticeship contracts capture this investment in human capital and enable us to see how wealthy Babylonians used slaves as a form of incomebearing investment.⁴⁹ Take, for instance, the following contract from 495 BC: Marduk-nāṣir-apli has voluntarily given his slave Itti-Uraš-pāniya to Gūzānu, son of Ḫ ambaqu, descendant of Mandidi, in order to learn the baker’s profession until further notice but for (at least) three months. He shall teach him all the skills of the baker’s profession. If he has taught him, Marduk-nāṣir-apli shall give to Gūzānu an uzāru-garment; if he has not taught him, Gūzānu shall give to Marduk-nāṣir-apli three qû of barley for each day as compensation for Itti-Urašpāniya. They have taken one (copy of the) document each. [A list of witnesses follows.]⁵⁰

In this respect, there is far more similarity with the Athenian situation than in the realm of agriculture, though with two caveats: first, Babylonian craft slaves would have taken longer to amortize their initial purchase price and generate clear profit than their later Athenian counterparts; second, we do not hear of slave-staffed sweatshops of the ergasterion variety in Babylonian sources.⁵¹ Perhaps the best attested use of slaves, though, was as business agents.⁵² As noted earlier in this section, two slaves of the Egibi family, Nergal-reṣua and Madānu-bel-uṣur, provide well-documented cases of how entrepreneurs might delegate many tasks to competent and trustworthy slave managers, who besides might be allowed a kind of peculium and run business interests of their own on the side.⁵³ Babylonian entrepreneurial families—like their later Athenian counterparts—tended to spread their interests in several different ⁴⁸ Dandamaev (1984): 270–8; Jursa (2010): 234–5. ⁴⁹ Dandamaev (1984): 246–8; Wunsch (2010): 50; Jursa (2010): 235–7. ⁵⁰ Abraham (2004): no. 84. ⁵¹ Dandamaev (1984): 279–307; Jursa (2010): 235–7. ⁵² Dandamaev (1984): 308–19. Unlike Athens, where no concept of legal agency existed and slaves filled the role of commercial agents on behalf of their owners (Harris 2013c), forms of agency (allowing free persons to contract as representatives of other free persons in, e.g., sales) are documented in Babylonia: see Jursa (2008): 610–12. That need not have made slaves any less useful as business agents, but simply meant that they did not have a monopoly on the role. ⁵³ See the bibliography on these slaves cited in n. 4 above. Ronan Head (2010) has recently studied slave agents and the problem of peculium in the archive of Iddin-Marduk of the Nūr-Sîn family and the Murašû archive. (Iddin-Marduk’s archive came to be stored with the Egibi archive when his daughter, Nūptāya, married into the family. For a detailed study, see Wunsch 1993.)

OUP CORRECTED PROOF – FINAL, 28/6/2018, SPi

Babylonia: Seventh–Fifth Centuries BC

245

ventures; in the words of Beaulieu, they had ‘a finger in every pie’.⁵⁴ This created a pressing need for managerial assistance. Trusted slaves were able to fill this occupational niche, constituting reliable personnel on the ground to deal with a wide variety of transactions and negotiations. When we come to the fifth century, the role of these managerial slaves seems, if anything, to have increased in significance; Jursa states that: the scale of the Murašus’ activities definitely surpasses quantitatively all other private businesses known from the sixth century . . . the need for these families to delegate managerial tasks, as well as their ability to do so (given their large number of dependants), were far greater than that felt by any of the entrepreneurial families of the sixth century’.⁵⁵

It is, therefore, at this secondary remove, rather than in primary production, where the contribution of slave labour to the fortunes of entrepreneurial families in Babylonia was perhaps most significant.

V . CONCLUSION S The more limited degree of development of slavery in Babylonia (when compared to Athens of a century or two later) can be better understood once slave labour is contextualized in terms of the broader labour market. Such a viewpoint brings out some similarities to and some differences from the later Athenian situation. As at Athens, Babylonian slaves who were hired out by their owners were paid the same amount for the same work as free wage labourers.⁵⁶ Unlike Athens, however, it would seem that the urban working classes in Babylonia had no compunction about working as the employees of other citizens; and the size of the working population that might labour for wages or as tenants was extremely large. This placed slave and free labour in direct competition; and the comparatively high purchasing costs of slaves limited the extent to which slave labour infiltrated a labour market that was dominated by large numbers of free wage labourers and free individuals seeking work as agricultural tenants.⁵⁷

⁵⁴ Beaulieu (2000). ⁵⁵ Jursa (2010): 240. On the status of ardu managers in the Murašû archive, see now Head (2010): 130–66. In essence, the semantic range of the term—like that of other Semitic slave words (on which, see Chapter 3, section II, pp. 84–5 above) encompassing slaves, but also used to articulate relations of social subordination—has engendered some debate over the legal status of Murašû managers. ⁵⁶ Dandamaev (1984): 130. ⁵⁷ On the labour of free persons in Babylonia, see Jursa (2010): 660–728.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

12 The Persian Empire In the case studies undertaken so far in this book, we have (Chapter 5 aside) found ourselves in the fortunate position of dealing with relatively discrete regional units, which can be treated as ‘slave systems’ in the manner discussed in the Introduction.¹ When we move on to discuss the Persian Empire, though, we run into difficulties, for it cannot be treated as a single regional unit: as J. K. Davies rightly notes, it ‘has to be seen economically as a partially interlocking set of separate local–regional economies, overladen by the two mildly unifying factors of a tax system and of estate holdings in provincial/satrapal areas on the part of Persian grandees’.² To this we can add a further problem: the extreme unevenness of our knowledge of the economies of these different regions. For some regions, such as Babylonia, we possess relatively rich documentation.³ For others, though, information is far harder to come by. In this chapter, I do not propose a comprehensive study of slavery in the Persian Empire, but rather a continuation of the ‘test pit’ method set out in the earlier chapters. We will focus on three particular regions of the Persian Empire: Anatolia, Egypt, and Fars (i.e. the Persian heartland). These regions offer the most promising material for the student of ancient slavery, and they allow us to gauge to what degree a Greek travelling through these regions during the fifth century BC would have encountered substantial systems of slave labour in action.

I. ANATOLIA Even if, for heuristic purposes, we split up the Persian Empire into discrete regional units, we should still not overemphasize the factor of regional unity, ¹ See pp. 7–8 above. ² Davies (1998): 251. On Persian grandees throughout the empire, see Briant (2002): 472–514. The empire is now seen as rather more bureaucratically complex than was thought twenty years ago: see the remarks of Henkelman (2013): 531–2. ³ See e.g. Stolper (1985) and Chapter 11 above.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

248

Greek Slave Systems

especially not in Anatolia.⁴ Here we find a patchwork of subregions: some areas, such as Cilicia and Paphlagonia, remained in the hands of local dynasts operating under the suzerainty of the Great King; other areas were under the direct control of Persian satraps. In those satrapal centres and their hinterlands, one might find considerable numbers of Persians ‘on the ground’, so to speak. These Persians were a mixed bunch, ranging from vastly wealthy satraps (e.g. Xen. Hell. 4.1.15–16) to relatively humble foot soldiers, with various intermediate ranks in between.⁵ In other areas, such as the Phrygian highlands, the fact of Persian control was probably far less noticeable for the inhabitants.⁶ Yet other regions, such as Mysia and Pisidia, were noted for their constant wars and rebellions against the Great King.⁷ It is against this backdrop that we must discuss several passages from classical authors that illustrate the use of slave labour among non-Greeks. By far the most detailed is an account related by Xenophon (Anab. 7.8.7–22), which describes an attack on an estate located in the plain of the River Caicus in Mysia that belonged to a wealthy Persian named Asidates. Xenophon tells us that he and a detachment of the Greek mercenary army were received in Mysia by a woman named Hellas, who apprised them of Asidates’ wealth and encouraged them to attack his property. The Greeks executed their raid at night: at the centre of Asidates’ estate was a strongly built tower (τύρσις), eight bricks thick, manned by numerous soldiers. When the Greeks arrived, Xenophon tells us that most of the slaves and animals that were around the tower ran away, the attackers letting them go so that they could concentrate of seizing Asidates himself (Anab. 7.8.12: ἐπεὶ δὲ ἀφίκοντο περὶ μέσας νύκτας, τὰ μὲν πέριξ ὄντα ἀνδράποδα τῆς τύρσιος καὶ χρήματα τὰ πλεῖστα ἀπέδρα αὐτοὺς παραμελοῦντας, ὡς τὸν Ἀσιδάτην αὐτὸν λάβοιεν καὶ τὰ ἐκείνου). They failed to capture Asidates, and had to beat a hasty retreat. What is interesting, though, is that the Greeks still managed to capture some 200 slaves (ἀνδράποδα), even though Xenophon states that many of Asidates’ slaves ran off (Anab. 7.8.19). The 200 captive slaves, then, can only have represented a fraction of Asidates’ overall holdings. Classical historians have been troubled by this passage, and have attempted to evade its plain meaning in two ways. First, N. Sekunda and V. Rosivach have claimed that the term ἀνδράποδα must have been intended proleptically: on this view, Asidates’ workers were not slaves, but are referred to as such by Xenophon in anticipation of their capture and reduction to slave status.⁸ It is true that Xenophon sometimes writes of ‘taking andrapoda’ in the sense of ⁴ In general, see Debord (1999); Dusinberre (2013). ⁵ Sekunda (1985; 1988; 1991); Dusinberre (2013): 32–113. ⁶ Thonemann (2013): 8–15. ⁷ Strattis fr. 36 K-A; Xen. Hell. 3.1.13; Anab. 1.2.1; 2.5.13; 3.2.23; Mem. 3.5.26; Hell.Oxy. 21.1; Arr. Anab. 1.25. ⁸ Sekunda (1985): 13; Rosivach (1999): 137.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Persian Empire

249

capturing free persons and reducing them to slavery, by using the term in combination with verbs of taking or capturing such as λαμβάνω or ἁλίσκομαι: note, for example, Xen. Anab. 6.3.3: ἀνδράποδά τε πολλὰ ἔλαβον, ‘they took many captives’.⁹ However, the usage of the term at Anab. 7.8.12 does not fall into this pattern: if we had only the reference to ἀνδράποδα at 7.8.19, we might suppose that the term refers to individuals of an unknown status who were captured by Xenophon and his fellow soldiers and thus reduced to slavery. But having it preceded at 7.8.12 in reference to individuals called ἀνδράποδα who ran away and were not captured makes it certain that the term is used for slaves in the normal sense, a usage that Xenophon often adopts elsewhere in his works.¹⁰ There is no good reason, then, to doubt that Asidates’ estate was manned by slaves, indeed a great number of slaves; and it is further worth noting that Asidates, though very wealthy, was not among the top tier of Persians living in the region.¹¹ On the other hand, G. E. M. de Ste. Croix accepted that Asidates owned many slaves,¹² but put this down to him imitating his Greek neighbours, a view required by de Ste. Croix’s belief that wealthy landowners in the Near East gained their riches from exploiting serfs and debt bondsmen, not slaves.¹³ Such an explanation of Asidates’ slaveholdings may, of course, be possible; but it is a solution whose logic stems from the prior belief that Near Eastern grandees never made much use of slave labour. A story in Herodotus (7.27–8), however, suggests that we need not explain away Asidates’ use of slaves in this manner. Herodotus relates that a Lydian named Atys, the wealthiest man in Asia, offered to give all his wealth to Xerxes to help fund his invasion of Greece, retaining only his farms, to which he would retire, living off the revenue of his slaves (andrapoda) and flocks. Whether the episode is historical or not, the key point is that Herodotus found nothing odd about a non-Greek with slave-manned farms; and it is far simpler to believe that this method of exploiting slave labour was common in the region and did not stop at the margins of Greek settlement only to be adopted by the occasional Hellenizing easterner. Likewise, if we skip forward in time to the late fourth century, we can find Eumenes of Cardia paying his mercenaries by capturing a number of fortified country houses in Phrygia that were full of slaves and flocks (Plu. Eum. 8.5). ⁹ Similar examples: Anab. 6.6.38; 7.6.26; Hell. 1.2.4; 3.2.2; 3.2.26; 4.6.6; 6.2.6. ¹⁰ Especially in his De Vectigalibus: 4.4; 4.15; 4.17; 4.21; 4.23; 4.24; 4.25; 4.26; 4.30; 4.35; 4.36; 4.39; 4.49; see also Mem. 2.4.5. This simple meaning of ‘slave’ and not ‘captive’ gives rise to the adjective ἀνδραποδώδης, ‘slavish’ (e.g. Xen. Mem. 4.2.22). ¹¹ See Sekunda (1991): 83–4. ¹² As he notes (1981: 507–8), ‘they must surely have been slaves in the Greek sense, rather than dependent peasants’. ¹³ ‘Barbarian grandees were often only too ready to adopt Greek practices.’ See de Ste. Croix (1981): 569 n. 42a.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

250

Greek Slave Systems

There is a further reason to believe that this form of slave-staffed estate was no oddity among the non-Greeks of Asia Minor. This region—especially Phrygia—was a major source of slave labour for the Aegean Greeks; and as we noted in Chapter 9, even a century before the classical period, Ezekiel speaks of slaves being imported into Tyre in the Levant from Yawan, Tubal, and Meshech, that is, from the ‘Ionians’, (i.e. Greeks), the Cappadocian kingdom of Tabal, and Phrygia.¹⁴ By the late fifth century, one comic poet would single out Phrygia as the Athenians’ source of slaves par excellence (Hermippus fr. 63 K-A). Our epigraphic sources paint a similar picture: although Phrygians are rare in the so-called ‘Attic Stelai’ (IG I³ 421–30) in which Thracians predominate, they are far more prominent in other epigraphic dossiers, for example the naval inscription IG I³ 1032, in which Phrygians outnumber Thracian slaves; the inscriptions relating to the mine slaves at Laurion, where Phrygians again appear as the single largest ethnic group; and the gravestones of non-Greeks from Attica, in which we find significantly more Phrygians than Thracians.¹⁵ Anatolia in general and Phrygia in particular would continue to export slaves around the Mediterranean for centuries to come.¹⁶ We have already noted the surprisingly low prices of slaves in classical Athens,¹⁷ and as David Braund has rightly pointed out in a recent study, that fact implies another: that the original purchase price at the point of origin of these slaves must have been even lower, since between the act of initial enslavement and the final purchase of the slave in Attica we must figure in the profits of traders, the cost of feeding, transport and supervision, and sales taxes along the route.¹⁸ Since the Anatolian estates just noted were rather closer to the point of origin of these slaves than Attica, their owners would have been able to secure an even cheaper source of forced labour than their mainland Greek counterparts.¹⁹ We must not, of course, oversimplify the problems of labour and land tenure in Anatolia; for one thing, we know of other forms of tenure through which prominent individuals obtained considerable wealth, such as grants by the Great King of revenues from certain cities or regions, or of lands.²⁰ Sources from the Hellenistic period furthermore depict landholdings with bound ¹⁴ See p. 199 above. ¹⁵ Lewis (2011), with pp. 102–4 and 111–13 for epigraphic evidence. On grave stelai, see Bäbler (1998). ¹⁶ See Lewis (forthcoming d); Harris (1980). ¹⁷ See Chapter 8, section I, pp. 170–2 above. ¹⁸ Braund (2011): 115; on transport costs on the route from Phrygia to Attica, see Lewis (2015a): 321–3 and Chapter 14 below. ¹⁹ As noted by de Ste. Croix (1981): 508. ²⁰ Themistocles, for example, was allegedly granted revenues of Magnesia, Lampsacus, Myus, Percote, and Scepsis: Thuc. 1.138.5 and Plu. Them. 29, with Dandamaev and Lukonin (1989): 138–9. Cf. Pl. Alc. 123b–c with Briant (2002): 460–3.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Persian Empire

251

peasants, the so-called basilikoi laoi, whose status is reminiscent of those bound peasants of the Neo-Assyrian period noted in Chapter 10.²¹ Rather, it should be admitted that within this complex mosaic of economic practices, some non-Greeks exploited slave labour—almost certainly local Anatolian slave labour—on a substantial scale.

I I. E G Y P T The export of slaves from Anatolia, as we noted above, had reached Tyre at least as early as the days of Solon. An intriguing dossier of Aramaic letters dating to the late fifth century BC shows that, by that point, enslaved labour from Anatolia was also playing a role farther south, on the estates of Persian grandees in Egypt. These documents, written on leather and kept today at the Bodleian Library in Oxford, comprise the correspondence of the satrap Aršama—who also owned lands in Babylonia, Assyria, and Syria²²—and furnish us with a valuable window into Persian-controlled Egypt. A letter from Aršama to his subordinate Nakhtḥor is particularly revealing: From Aršama to Nakhtḥor. And now: previously, when the Egyptians rebelled, then Samšek the previous official, our personnel and goods [which] are in Egypt he guarded with force, so that there was not any loss from my estate. Also, from elsewhere (‘another place’), personnel of artisans of every kind and other goods, sufficiently he sought and made over to my estate. And now: it is thus heard by me here, that the officials who are [in Lo]wer (Egypt) are being diligent in the disturbances (?), and are forcefully guarding the personnel and goods of their lords. They are also seeking others from elsewhere, and are add[ing t]o the estate of their lords. But you (and your colleagues) are not so doing. Now, I have also previously sent (word) to you concerning this: ‘You are to be diligent. Guard [my] personnel and goods forcefully, so that there shall not be a[n]y loss from my estate. Also, from elsewhere, personnel of artisans of every kind, seek (in) sufficient (numbers), and bring (them) into my courtyard, and mark (them) with my brand, and make (them) over to my estate, just as the [pre]vious officials were doing.’ Thus let it be known to you, if from the personnel or from my other goods there should be any loss, or from elsewhere you should not seek and should not add to my estate, you will (all) be questioned forcefully, and a severe sentence will be produced for you (Nakhtḥor). [Ar]taḥaya knows this order. Rašta is the scribe.²³

²¹ See p. 232; for the basilikoi laoi, see Papazoglu (1997); Bussi (2001); useful overview of the issue in Thompson (2011): 195–200. ²² See Tuplin (2013a): 30–8. ²³ TAD A6.10, tr. D. G. K. Taylor.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

252

Greek Slave Systems

It is clear that we are dealing with a request to seek out slave labour here, since Aršama demands that these newly acquired labourers be brought to his courtyard and branded (or tattooed) with his mark, evidently an example of the time-honoured Near Eastern tradition of marking slaves out as the property of a certain individual.²⁴ Equally interesting is the mention of other grandees in lines 3–5 whose subordinates—unlike Nakhtḥor—were successful in obtaining labourers for the estates of their lords, which shows that Aršama’s economic situation was not unique in Egypt.²⁵ Other letters provide us with further details—including onomastics—for Aršama’s slaves, many of whom are referred to as Cilicians. TAD A6.9, for instance, mentions Cilician slaves assigned to Nakhtḥor for a journey to Egypt; in TAD A6.15 Aršama’s steward Virafša complains to Nakhtḥor that five Cilicians promised to him had not been delivered, and in TAD A6.7 Aršama writes requesting the return of thirteen Cilician slaves who had been captured during a period of local unrest. Some of these individuals appear to bear Anatolian names.²⁶ Slavery was hardly new to Egypt,²⁷ but it is interesting to find foreign slaves who were probably sourced through commerce occupying such a prominent role in the affairs of a Persian satrap in Egypt (and in this respect, Aršama’s approach was not so different from the Athenian model of slaving). That Egypt was enmeshed in the wider currents of Eastern Mediterranean trade is not to be underestimated: the early fifth-century Elephantine customs scroll, recording incoming Greek and Phoenician merchant ships and their return cargoes of natron, brings this forcefully home.²⁸ This trading network provides a compelling context for the arrival of Anatolian slaves in Egypt, and as good a warning as one can give against detaching Greek slavery from its Eastern Mediterranean context: from the aforementioned Greek, Hebrew, and Aramaic sources we can see Anatolian slaves being traded in at least three directions (towards the Levant, at least during the archaic period; towards Greece; and towards Egypt); we have seen in Chapter 10 that the Assyrians of the eighth–seventh centuries BC also imported Anatolian slaves, whilst Ezekiel’s testimony shows that exchange also flowed eastwards from Greece,

²⁴ On the translation ‘brand’, see the characteristically detailed commentary of Tuplin (2013b): 82–4. ²⁵ As pointed out by Tuplin (2013b): 77–8. ²⁶ See Tuplin (2013b): 33. For the problems with deriving ethnicity from slave names, see Lewis (2011): 93–8; (forthcoming d). ²⁷ For a useful survey of the changes in agricultural labour in Egypt during the preceding centuries, see Moreno Garcia (2015). A new, up-to-date book on Egyptian slavery to replace Bakir (1952) is a desideratum; the illuminating article of Morris (2014) showcases much of the New Kingdom material. See also Muhs (2016): 85; 129–30; 168–9; 204–5. ²⁸ Yardeni (1994); translation in Kuhrt (2007): 681–703. For the wide range of commodities in circulation in Eastern Mediterranean trade during this period, see van Alfen (2015).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Persian Empire

253

with Greek slaves arriving in Tyre.²⁹ A similar situation emerges in the fourth century, for Justin reports a major slave revolt in Tyre shortly before Alexander’s invasion, showing that Tyre’s imports of slaves were still significant several centuries after Ezekiel’s time;³⁰ whilst Theopompus of Chios (FGrHist 115 F 114) reports that King Straton of Sidon acquired many prostitutes from the Peloponnese; similarly, Plutarch notes a slave trader named Theodorus of Tarentum peddling young boys in the Levant during Alexander’s invasion of Asia (Plu. Alex. 22). The slave-trading networks of the Eastern Mediterranean linked these disparate regions into a larger system of demand and supply, of merchants and markets, which criss-crossed these regional divides. The legal aspects of slavery in Egypt during this period also mirror the fundamental shape of slave law in Greece and the Near East set out in Chapter 1, from the demotic documents studied by Eugene Cruz-Uribe³¹ to the Aramaic documents from the Jewish colony at Elephantine studied by Bezalel Porten.³² One might note, for example, a document of 410 BC from Elephantine concerning the division of property between two heirs, Jedaniah and Mahseiah; the former inherited a slave named Petosiri (a common Egyptian name), whilst the latter inherited a slave named Besa.³³ Both were children of the same mother, Tabi; and like Aršama’s slaves, they too were branded or tattooed with a mark, here indicating that they belonged to the testator, Mibtahiah.³⁴ Slave ownership, then, was not just restricted to the fabulously rich, but could also be found among rather humbler denizens of the satrapy.³⁵ One might also compare the legal features of the Elephantine papyri to the remarkable fourth-century Samaria papyri from Wadi Daliyeh, whose clauses setting out the permanent, transmissible interest of the buyer, warranty, and so forth contain various similarities with the Babylonian sale contracts discussed earlier in this book.³⁶ They further demonstrate that

²⁹ For slaves being traded away from Greece, see Lewis (2011): 109 and the sources cited there. ³⁰ See Justin 18.3.1–19 with Elayi (1981). ³¹ Cruz-Uribe (1982); see in particular pp. 65–7, arguing against Orlando Patterson’s definition of slavery in favour of the traditional approach. ³² Porten (1996). ³³ Porten (1996): 199–201. ³⁴ Cf. Stolper (1984), providing an edition of a Babylonian slave sale document from the Persepolis Fortification Archive, in which the slave who is the object of the transaction bears an ownership mark on his right hand. ³⁵ Cf. Hezser (2005): 289–91, who discusses slave ownership among the Jews of Elephantine and concludes (at 290) that ‘it is most likely that they owned a few slaves each who worked within the household economy’. One cannot but agree, though it is worth noting that this would put the slaveholdings of these Elephantine Jews in the same range as those of the small farmers portrayed by Aristophanes, whose real-life models lived at precisely the same time, hundreds of miles to the north-west. ³⁶ This archive, discovered in the Abu Shinjeh cave in 1962, mainly dates to the reign of Artaxerxes III (358–338 BC); of those eighteen documents published by Gropp et al. (2001), ten concern slave sales. For further analysis, see Dušek (2007), with detailed discussion and debate

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

254

Greek Slave Systems

none of these legal features were distinctive to Greek slavery, but were shared far more broadly.

I I I. F A R S A rather more complex set of problems emerges from an examination of the royal economy of Fars, that is, the Persian homeland, in particular the district surrounding Persepolis. By far the bulk of our evidence for this economy derives from archives discovered at Persepolis: the so-called Persepolis Treasury Tablets (PTT) and the Persepolis Fortification Tablets (PF; PFa).³⁷ These documents, written mainly in Elamite, were unearthed during digs in the 1930s. They provide a highly detailed snapshot of the inner workings of the royal economy, covering various transactions and interactions, but particularly the disbursement of rations to work gangs—Matthew Stolper has called the picture they offer ‘the view from the lunchbox’³⁸—with PF and PFa covering the years 509–493 BC and PTT the years 492–458 BC. These gangs were active in the construction of the palace, royal workshops, and other activities in the surrounding countryside. Rich evidence they may be, but reconstructing the royal economy from them is no straightforward matter. First, they document short windows of time, and unevenly at that: Wouter Henkelman and Kristin Kleber point out that: one of the most distorting characteristics of the Persepolis Fortification texts is the fact that the texts come from a very limited time span . . . and as much as 46% of the dated texts are from two years . . . this makes it very hard to establish medium- and long-term patterns in the administration of the Achaemenid heartland economy.³⁹

Second, those tablets that we do possess are not all published; and even if they were, they do not represent the original sum of the documentation.⁴⁰ Perhaps over how much the formularies owe to Mesopotamian practices. Several documents attest to the sale of multiple slaves in single agglomerated lots: Gropp et al. (2001) nos. 2, 6, and 8 all record the sale of two slaves in single transactions; whilst in no. 5 we find three slaves sold, and perhaps four in no. 7. Most remarkable is no. 9, a contract for the sale of seven slaves. We also find joint ownership in several documents: joint buyers in nos. 3 and 9; and joint sellers in nos. 4, 5, 7, 8, and 9. One individual mentioned in the archive, Netira bar Yehopadani, can be seen to acquire at least twelve slaves (Gropp et al. 2001: 95). An Athenian of the same period making purchases on this scale would surely rank among the liturgical class. ³⁷ PTT: Cameron (1948); PF: Hallock (1969); PFa: Hallock (1978). ³⁸ Stolper (2014): 21. ³⁹ Henkelman and Kleber (2007): 169. ⁴⁰ Briant (2002): 423. For continued work on the archive, see Stolper (2014): 23–7, who publishes an annual bulletin on the progress of the Persepolis Fortification Archive Project in the University of Chicago’s Oriental Institute Annual Report.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Persian Empire

255

the most pressing question for our purposes, though, is the status of the workforce, labelled kurtaš in the documents. On this issue, various opinions have been offered: Cameron, for instance, thought the kurtaš in the PTT were most likely low-status Persian workmen, but reviews of his work by Hinz and Falkenstein dissented on this point, considering it more likely that they were enslaved war captives.⁴¹ Dandamaev, although considering the bulk of the kurtaš to have been slaves,⁴² showed that some, at least, were corvée labourers.⁴³ We are likely to get off on the wrong foot if we ask ‘What was the status of the kurtaš?’,⁴⁴ for that question presupposes that they held a uniform legal status. Nor does the observation that the term kurtaš is a calque of the Old Persian *grda and can also be found in Aramaic as garda (used for Aršama’s workforce in his correspondence) get us very far either; Dandamaev’s claim that ‘this seems to be conclusive evidence that also the kurtaš in Persia and Elam were slaves’⁴⁵ leans a great deal of weight on the historical implications of lexical similarity. It is safer to adopt Briant’s position, i.e. to translate the term as ‘worker’, since ‘garda is more a label than a technical term whose juridical status can immediately be defined’.⁴⁶ Without assuming a uniform status, then, or inferring status from a cognate term in another language, we must, therefore, look at the picture that emerges from the documents themselves and, in that way, attempt to gauge the legal status (or statuses) of the workforce. G. G. Aperghis has estimated that the size of the kurtaš workforce c.500 BC lay in the region of 10–15,000 individuals.⁴⁷ The workforce appears often to have been subdivided into gangs based on ethnic origins: we find references, for instance, to gangs of 303 Lycians (PF 1368), 150 Thracians (PF 1363, 2055; PFa 18), 547 Egyptians (PF 1557), and 980 Cappadocians (PFa 30).⁴⁸ What brought these gangs from such various and far-flung regions? We may rule out the notion that these workers were economic migrants seeking wages, ⁴¹ Hinz (1949); Falkenstein (1952). ⁴² Dandamaev (1963): 150; (1975); cf. Dandamaev and Lukonin (1989): 152–77. This view is similar to that espoused in Zaccagnini (1983): 262–4 and Aperghis (2000). Cf. Briant (2002): 433–9, who (at 439) writes that their condition was ‘much closer to slavery than the “helot” type of rural dependency’ (which Briant views as similar to the laoi in certain Hellenistic inscriptions). ⁴³ Dandamaev (1975): 77–8; Dandamaev and Lukonin (1989): 173; Henkelman and Kleber (2007); Henkelman (2009), with a summary of earlier views at 282 n. 33. ⁴⁴ As does Aperghis in his otherwise very valuable study (2000). ⁴⁵ Dandamaev (1975): 76. ⁴⁶ Briant (2002): 458. For the translation ‘worker’, see Henkelman and Kleber (2007): 168. ⁴⁷ Aperghis (2000): 138–9. ⁴⁸ Not to mention Greeks: as Lewis (1985): 107 comments: There are men in our texts who are simply called Yaunā. No one that I know of has spoken against the obvious view that this is not a true proper name, that the persons concerned are Greeks, known by their ethnics instead of their strange and no doubt unpronounceable names, just as the Greeks habitually called slaves Skythes or Kar.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

256

Greek Slave Systems

since the rations disbursed to the kurtaš were sufficient only for survival; Dandamaev has shown that they were given a mere sixth of what labourers were paid in Babylonia during the same period:⁴⁹ It is clear that the Babylonians, Egyptians and representatives of some other peoples mentioned in the Fortification texts were not in Iran of their own free will to earn money, but were led off there by force and exploited by direct coercion, regardless of whether they were there only temporarily or for their entire lives.⁵⁰

Several other possibilities remain open. On the one hand, it is clear that at least some of these labourers were recruited through the time-honoured Near Eastern institution of corvée: several texts refer to workers drafted for temporary service in Iran.⁵¹ That, however, can only account for a part of the workforce: it is clear that for many of these workers their condition was permanent, not temporary.⁵² We also find fixed incentives to encourage reproduction among the kurtaš population, i.e. extra rations for women who fell pregnant.⁵³ One might suppose this population resulted from Assyrianstyle deportations and relocations; but, if so, the condition of the kurtaš was very different from that of deported populations in Assyria, which, once relocated, were largely left alone and not micro-managed.⁵⁴ The enormous bureaucratic effort that emerges from the tablets shows, rather, a closely managed workforce whose subsistence was provided through ration disbursements rather than the domestic consumption of food grown on their own smallholdings. Some of the most important clues to the condition of this workforce can be found in the writings of classical authors, who refer to both the enslavement of war captives and the forcible removal and relocation of communities.⁵⁵ Herodotus notes several instances of mass enslavement: at 1.156 he tells us that Cyrus the Great sold into slavery certain people who had attacked Sardis; and Darius I allegedly enslaved the inhabitants of Barca, relocating them to a village in Bactria (Hdt. 4.203–4). Darius also was alleged to have enslaved several Paeonian tribes and relocated them to Asia (Hdt. 5.15, 17); and a ⁴⁹ Dandamaev (1975): 74; more detail in Dandamaev and Lukonin (1989): 161–74; cf. Aperghis (2000): 131–5. ⁵⁰ Dandamaev and Lukonin (1989): 174. ⁵¹ Dandamaev and Lukonin (1989): 173–4 discuss, inter alia, Mich. 26, mentioning a man who had to provide a year’s labour service in Elam; PTT 22, mentioning a disbursement of money to Syrian kurtaš ‘who were conscripted and now released’. For corvée labourers in Iran in the time of Cambyses, see Henkelman and Kleber (2007). ⁵² Briant (2002): 430. ⁵³ Briant (2002): 435 views these as reward rations and notes that ‘these documents . . . show that some of the kurtaš of Fars were quite simply the result of natural population growth, since doubtless the status of mothers was passed on to their children’. Cf. Dandamaev (1975): 77; Aperghis (2000): 136. ⁵⁴ See Chapter 10, section III, pp. 231–2 above. ⁵⁵ What follows draws heavily on Dandamaev and Lukonin (1989): 170–1.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

The Persian Empire

257

number of Milesians were enslaved and sent to Susa (Hdt. 6.20; cf. 6.101). According to Diodorus (1.46.4), Cambyses drove 6,000 Egyptians into Asia in order to construct palaces, and when Xenophon passed through Babylonia at the end of the fifth century, he observed a number of villages belonging to the king’s mother Parysatis full of slaves (Xen. Anab. 2.4.27; cf. 2.3.17).⁵⁶ The most graphic and important classical source, though, is Diodorus, who records an episode that occurred as Alexander advanced on Persepolis (Diod. 17.69.2–9; cf. Curtius 5.5.5–24; Justin 11.14.11–12): At this point in his advance the king was confronted by a strange and dreadful sight, one to provoke indignation against the perpetrators and sympathetic pity for the unfortunate victims. He was met by Greeks bearing branches of supplication. They had been carried away from their homes by previous kings of Persia and were about eight hundred in number, most of them elderly. All had been mutilated, some lacking hands, some feet, and some ears and noses. They were persons who had acquired skills or crafts and had made good progress in their instruction; then their other extremities had been amputated and they were left only those which were vital to their profession. All the soldiers, seeing their venerable years and the losses which their bodies had suffered, pitied the lot of the wretches. Alexander most of all was affected by them and unable to restrain his tears.

As Dandamaev has noted, ‘it is obvious that these Greeks comprised part of the kurtaš’.⁵⁷ Perhaps there is some exaggeration here about the ubiquity of mutilation; but this source provides a very different perspective on the royal economy from that of the ration texts of a century and a half earlier. One might have been inclined to dismiss the episode as anti-Persian propaganda had it not aligned so well with what we can see in the Fortification Archive. However complex the kurtaš workforce was, then, we can probably agree with Dandamaev and Briant that enslaved war captives, sometimes treated in the most brutal fashion, constituted a central component of the royal economy of Fars.⁵⁸ *

* *

⁵⁶ Briant (2002): 549 thinks that these villagers were a kind of basilikoi laoi of the Hellenistic type, but his argument relies on a mistranslation of the passage. The key line reads ταύτας Τισσαφέρνης Κύρῳ ἐπεγγελῶν διαρπάσαι τοῖς Ἕλλησιν ἐπέτρεψε πλὴν ἀνδραπόδων, which Briant renders as ‘Tissaphernes, by way of insulting Cyrus, gave over these villages—except that it was forbidden to enslave the inhabitants.’ But the Greek says that he turned over the villages ‘apart from the slaves’. Paola Ceccarelli points out to me that the mistake is present in the French original of Briant’s book because it lifts the translation (and error) from the Budé translation. ⁵⁷ Dandamaev and Lukonin (1989): 171; Briant (2002): 434 agrees. Briant also notes the Lycian mentioned in Diod. 17.27.28, who had been captured in war, enslaved, and sent to Persia to work as a shepherd. In Plutarch’s version (Alex. 37.1) the man’s father had been Lycian, his mother Persian. Diodorus’ testimony supports the idea that successive Persian kings enslaved peoples on the frontiers of their empire and brought them to Fars; this aligns with the suggestion of Aperghis (2000): 129 that ‘It is likely that in the period after the initial deportations there may have been additional ones on a smaller scale to “top up” the workforce.’ ⁵⁸ On grants of kurtaš to members of the Persian royal family, see Aperghis (2000): 137, who argues convincingly that these were probably grants of usufruct rather than genuine conveyances.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

258

Greek Slave Systems

In several parts of the Persian Empire, at the very least, slave labour played a prominent role in production; this not only involved (at least in Fars) Greek slaves, but was witnessed by no less than Alexander the Great. However, for the most compelling example of a major slave system beyond Greece and Rome we must look forwards in time from this period and to Carthage in the west, our final case study.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

13 Punic Carthage Perched on the North African littoral where Cape Bon juts into the Mediterranean towards Sicily, the city state of Carthage ruled over a considerable empire in the Western Mediterranean until its destruction by Rome in 146 BC.¹ Carthage’s strength derived in no small part from the intensive exploitation of its rich agricultural hinterland, much of it cultivated by a massive slave population. Yet one of the most remarkable aspects of modern scholarship on ancient slavery is the neglect with which this subject has been treated: it would be no exaggeration to say that an entire large-scale slave system (almost certainly a major ‘slave society’) has been effectively written out of the standard history books on ancient Mediterranean slavery.² This chapter, short as it is, cannot provide exhaustive discussion of this issue. However, it will set out and discuss some of the main evidence for slave status and slavery in the economy of Carthage and its empire. This evidence should be sufficient to re-establish

¹ Carthage was well within the ambit of the classical Greeks, certainly more so that Babylonia or Assyria: Sicilian Greeks had guest-friendship contacts with leading Carthaginians (Hdt. 7.165), whilst a fourth-century decree from Boeotia grants proxenia to a Carthaginian (R&O no. 43), and Greeks bought Carthaginian (and western Phoenician) exports such as textiles (Hermippus fr. 63 K-A), salted fish (Eupolis fr. 199 K-A; Zimmerman Munn 2003), and hunting nets (Xen. Cyn. 2.4). Aristotle took a keen interest in its political institutions (Arist. Pol. 1272b24–1273b25), and some even thought that fifth-century Athenians harboured ambitions of conquering Carthage and expanding Athenian dominion into Africa (Thuc. 6.15). (Some fourth-century Athenians did accompany an abortive campaign against Carthage: Diod. 20.40.5–6.) ² See the Introduction, p. 3 above. This trend is not reflected in specialist studies of Carthage, where scholars have never doubted the large-scale use of slave labour, e.g. Gsell (1920): 172 (‘les esclaves, amenés de partout, étaient, nous le savons, fort nombreux’); Matilla Vicente (1977): 99–104; Lemaire (2003): 221–2; Miles (2010): 219; Hoyos (2010): 69–70; Ameling (2012). (The Picards, though, had little to say on the topic: Picard and Picard 1961: 87.) Two general studies on slavery recognize its scale at Carthage: Patterson (1991): 21; Flaig (2009): 55–6. Flaig (2009): 55 rightly remarks ‘Erheblich unterschätzt wird die phönizische Sklavenhaltergesellschaft im westlichen Mittelmeer. Die Nachrichten sind spärlich, Urkunden fehlen, daher entgeht diese Sklaverei leicht der historischen Aufmerksamkeit.’ Shortly before I completed this book, Noel Lenski shared an essay with me (Lenski 2018a) that contains some similar remarks on Carthage; we seem to have been working independently along parallel lines, and the fact that we come to similar conclusions about Carthaginian slavery is heartening.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

260

Greek Slave Systems

Carthage’s place in the eyes of historians of slavery as a major slave system of antiquity.

I. SLAVE STATUS I N CARTHAGE We earlier noted Finley’s avoidance of Carthaginian slavery, inconvenient as it was to his general model of the rise of slavery and evolution of the concept of freedom. The most that he wrote on this issue was in a footnote to The Ancient Economy (1973), where he noted Mago’s fame as an agronomist, elusively adding that ‘the labour situation in North Africa is still open to discussion’.³ Finley explicitly retracted this comment in a section on ‘Further Thoughts’ added to the 1985 edition of this book: ‘I now believe that in North Africa certainly, and in Spain and Gaul probably, local varieties of dependent labour survived the Roman conquest as they had done in the east.’⁴ But what exactly were these ‘local varieties of dependent labour’? To back up this claim Finley cited two works by C. R. Whittaker on labour conditions in North Africa.⁵ It is important before we move on to address the question of slavery in the Carthaginian economy that we examine these arguments—as well as the Greek, Latin, and Punic sources for dependent status in Carthage—to determine whether we can speak at all of Carthaginian slavery or whether we are dealing with some different, sui generis Punic form of dependent labour, as Finley supposed. To understand this problem, we must first recall Polybius’ revealing discussion of the Carthaginians’ relations with their hinterland and subject communities in the context of the Mercenary War (240–238 BC). He tells us that the Carthaginians relied on the produce of their land for their own private fortunes, but upon tribute levied on the Libyans for their public finances (Polyb. 1.71.1).⁶ The level of this tribute rose to excessive levels during the First Punic War (264–241 BC): 50 per cent of crop yields, plus various taxes imposed on the townsmen, with no exemptions allowed for the poor (Polyb. 1.72.1–7). This, naturally, caused a great deal of resentment; the situation became exacerbated when, shortly after losing the war, the Carthaginians billeted their mercenary armies throughout their territory pending demobilization and payment. When payment was delayed due to Carthage’s huge war indemnities to Rome, the unpaid mercenaries, led by Spendius and Mathus, revolted and were supported by the overtaxed Libyan population. ³ Finley (1999): 223 n. 18. ⁴ Finley (1999): 179. ⁵ Whittaker (1978; 1980). Finley also cited Whittaker (1978) in Finley (1980): 115. For cogent criticisms of Whittaker (1980), see de Ste. Croix (1981): 563–4 n. 13. ⁶ See Hoyos (2010): 63; 142–8.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Punic Carthage

261

Whittaker’s 1978 article studied the condition of the Libyan subjects and Numidian ‘allies’ of the Carthaginians, but he avoided the issue of slave labour on the estates of the Carthaginian elite, precisely the area which required study were Carthage to be granted (or denied) the rank of ‘slave society’.⁷ This omission meant that his study of ‘land and labour in North Africa’, far from supporting Finley’s position on the lack of major slave societies outside Greece and Rome, simply did not come to grips with the ‘slave society’ question. Nevertheless, it allowed Finley to justify ignoring Carthaginian slavery altogether, and the topic has remained marginal if not completely absent from Anglo-American studies of ancient slavery ever since. A survey of the evidence leaves us in no doubt that the Carthaginians not only had a system of slavery (rather than some mysterious Punic form of dependency), but an exceedingly large one at that. For one thing, Greek writers make it clear that Carthage was closely linked to the trade in human beings, one of the key indicators that we are dealing with slavery sensu stricto. For example, we learn from Pseudo-Aristotle (De mirabilibus auscultatibus 88) that the Balearic islanders with whom Carthage traded were known to traffic in slaves (cf. Diodorus Siculus 5.17.3).⁸ Diodorus, perhaps drawing on the great Western Greek historian of the early Hellenistic period, Timaeus of Tauromenium, adds further details to our picture when he informs us that slaves from Corsica—well within the ambit of Carthaginian power in its heyday—were renowned for their quality (Diod. 5.13.5).⁹ To be sure, the slave-trading network of the Carthaginians is almost entirely shrouded in obscurity owing to a paucity of sources, but one passage makes it clear that we are dealing with a large-scale, high-volume slave-trading system. Appian (Punica 2.9) mentions that during the Second Punic War, Hasdrubal Gisco bought (ἠγόραζε) 5,000 slaves as rowers for the Carthaginian fleet. One might compare Xenophon’s advice to the Athenians to buy some 10,000 public slaves for their mines, a purchase that he recommended staggering over several years so that the market could supply high-quality slaves at reasonable prices (Xen. Vect. 4.23–4). That the Carthaginians could purchase in one fell swoop half as many slaves as Xenophon advised the Athenians to buy over several years is, therefore, a remarkable testament to the vibrancy of the slave supply in the

⁷ Though he did admit in passing that ‘chattel slavery, of course, existed, particularly in the Carthaginian chora and in the wealthy households of Carthaginian grandees’ (Whittaker 1978: 338). ⁸ At 5.17.2 Diodorus notes that the Balearic Islanders imported a lot of wine; for the archaeological evidence for Carthaginian transport amphorae in Ibiza, see Bechtold and Docter (2010): 99. ⁹ Jacoby thought the passage to be Timaean, and logs it as FGrHist 566 F 164. Cf. Hoyos (2010): 70, who suggests that Cassius Dionysius’ recommendation of Epirote slaves may derive from Mago (on whom, see section II, pp. 263–4 below).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

262

Greek Slave Systems

Western Mediterranean at this time.¹⁰ It should come as no surprise that the Aegean slave market—stimulated by the demand of a highly commercialized city-state culture—had its counterpart in the west due to similar conditions, though we know far fewer details with regard to the latter. That Carthage stood at the head of a complex network of trading routes reaching into the Iberian and African interiors and encompassing the western Mediterranean islands is undisputed, and it is this context that makes large-scale slave trading credible.¹¹ That we are dealing with a system with more similarities than differences from the Greek model emerges when we look at the evidence of Punic inscriptions. We find, for instance, the consecration of slaves to a deity, a phenomenon known from the Greek world, and perhaps in this case a form of manumission.¹² Likewise, as in Greece Punic slaves could be released after having paid their masters the cost of the manumission (e.g. CIS I 269), or alternatively without paying at all (Bulletin Archéologique 1943–5: 160).¹³ Furthermore, some attention was given to the importance of publicizing the ex-slave’s new status: J. G. Février has argued that freedmen might not only erect a stela thanking certain deities (usually Tanit and Baal Hammon) for their freedom, but also have their new status recorded in a public registry.¹⁴ Some slaves seem to have been manumitted by testament in Carthage, just as in Greece.¹⁵ The Carthaginian experience of slavery paralleled that of Greece in other ways. Plautus (Cassina 71) refers to slave ‘marriage’ practices and says that slaves can have wives in Greece and Carthage;¹⁶ and Appian tells us of an occasion in 149 BC when the Carthaginians manumitted and enrolled large

¹⁰ Flaig (2009): 55: ‘eine solche riesige Zufuhr setzt eine stattliche Lieferzone in Nordafrika voraus’. But cf. Hoyos (2010): 71 for an alternative view. ¹¹ That is not to deny the role of war captives in Carthage’s slave supply. For example, Scipio and Masinissa released numerous enslaved Roman war prisoners who had been captured in Spain, Sicily, and Italy and sent to work the Carthaginian countryside by Hannibal (App. Pun. 15). Populations enslaved in warfare are also treated in the Second Treaty between Rome and Carthage (348 BC): Polyb. 3.24.4–14; cf. App. Pun. 1.3. Diodorus (20.13.1) mentions that some 20,000 manacles were captured from the Carthaginians after their battle with Agathocles in 310 BC: they had clearly been planning to enslave their captives en masse. ¹² For the Greek world, see Zanovello (2016): 83–130; for the Punic documents, see Février (1951–2); (1961); Lemaire (2003): 222; Hoyos (2010): 21; 68. ¹³ See Février (1951–2). ¹⁴ Février (1951–2): 18. A number of these inscriptions bear towards the end of the document the phrase ‘people of Carthage’ (‫)עם קרתחדשת‬, implying some level of state recognition of the change in status. ¹⁵ See Février (1961) for a slight change in his interpretation on the document studied in Février (1951–2), favouring an interpretation linked to a will. On manumission by testament in Greece, see Canevaro and Lewis (2014): 103–110, focusing on the wills of the philosophers. ¹⁶ Pace Ameling (2012), we need not read this as evidence for legal rights held by slaves: see Lewis (2013) for reasons of method explaining why we need not make such assumptions.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Punic Carthage

263

numbers of slaves in the armed forces due to a military emergency, an event reminiscent of the run-up to the battle of Arginusae.¹⁷ On the other hand, it would be ill advised to push these similarities too far and suggest that they drew on Greek legal traditions; for the most part they represent general features that can be found in many historical slave systems. In the Punic language the word for slave is more or less identical to the Hebrew term for slave, ‫עבד‬, and its semantic range is similar to its Hebrew equivalent, as it can be used as a mark of social deference or service to a god in addition to its concrete legal meaning.¹⁸ Western Semitic legal traditions, like those discussed in Chapter 12, probably lie behind Carthaginian slave law. Classical authors too note peculiar features of Carthaginian slave law that the Greeks did not share: Plato (Leg. 674a), for instance, mentions a law banning the consumption of wine by male and female slaves (cf. [Arist] Oec. 1.1344a33–4). At any rate, the important point to emphasize is that we are undoubtedly dealing with slave status in the strict sense and that Finley was mistaken to dismiss dependent labour at Carthage in terms of some mysterious local form of non-slave, but unfree labour.

II. S LAVERY IN THE P UNIC ECONOMY When studying the role of slavery in the Carthaginian economy, it is important for us to begin (as in previous chapters) by mapping out the economy’s main contours; once this framework is in place, we may turn to the question of how slavery fits into this larger picture. Surveys of Carthage’s hinterland combined with study of the numbers and distribution of transport amphorae enable us to document changes over time in the exploitation of the city’s agricultural territory. Growth in rural settlement can be charted from an early period.¹⁹ Palaeobotanical studies afford a glimpse into the crop types cultivated in this region,²⁰ tallying with the accounts of Diodorus (20.8.4) and Appian (Pun. 117), who attest to gardens and fruit orchards, olive groves, flocks, and herds. The richness of the soil in this region was famous, and Carthaginian agronomy reached a high degree of technical sophistication: Greek and Roman writers produced translations and adaptions of Carthaginian agronomic treatises; the most famous was that of Mago, which was translated into Greek by a certain Cassius Dionysius and into

¹⁷ Carthage: App. Pun. 93; cf. Livy 21.45.7; Arginusae: Xen. Hell. 1.6.24; Ar. Ran. 693–4; Hellanicus FGrHist 323a F 25; Diodorus 13.97.1. ¹⁸ See Lemaire (2003): 219, with examples from Punic epigraphy. ¹⁹ See Docter (2009): 182–4 for an up-to-date overview. ²⁰ Docter (2009): 185–6.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

264

Greek Slave Systems

Latin by one Decimus Junius Silanus, with the costs defrayed by the senate (Varro, RR 1.1.10).²¹ Of course, Carthage is most famous for its maritime trade, and the study of amphora distribution can tell us a good deal about Carthaginian imports and exports. B. Bechtold and R. Docter have charted a dwindling reliance on foreign imports of wine and olive oil as the Carthaginian hinterland came to be more heavily exploited, but this process does not simply reflect a move towards autarky: Carthaginian transport amphorae can subsequently be found in large numbers across many Western Mediterranean sites, especially Sicily and Spain, as well as the Balearics. The city clearly became a major producer of cash crops, especially wine and olive oil, which were exported widely.²² S. Lancel has compared Carthage at the height of its prosperity to sixteenthcentury Venice.²³ The parallels are certainly notable, but unlike the free farmers of the Venetian contado, it is clear that a sizeable slave population cultivated much of the Carthaginian chora: slavery must be counted as a central element lying behind the picture of agricultural expansion revealed by the archaeological data. Let us turn to the location of slavery in Carthaginian society. The city was ruled by a wealthy elite, and its mixed constitutional arrangements were praised as early as the fourth century BC by Aristotle (Pol. 1272b24–1273b25), who included it in his research programme into the political arrangements of various poleis scattered around the Mediterranean and the Black Sea. Much of the rich agricultural hinterland of Carthage comprised large estates owned by the aristocracy, as Diodorus (20.8.2–4) explains in the context of Agathocles’ invasion in 310 BC:²⁴ But Agathocles, eagerly seeking to rid his soldiers of their low spirits, was leading his force against the city called Megalopolis, which belonged to the Carthaginians. The intervening countryside, through which it was necessary to travel, had been divided into gardens and orchards of all sorts, watered by many canals and irrigated in every place. There were country houses at frequent intervals, expensively built and stuccoed, clearly indicating the wealth of their owners. And the barns were full of everything needed to satisfy their desires, since the inhabitants had stored up over a long period of peace an abundance of food. And some of the ²¹ On Mago, see Lancel (1995): 273–9; on advice for slave management probably deriving from Mago, see Hoyos (2010): 65–7. ²² Bechtold and Docter (2010). On cereal production, see Lancel (1995): 270. ²³ Lancel (1995): 120–1. ²⁴ On Agathocles’ invasion, see Miles (2010): 149–54. It is regrettable that we do not know with any precision the degree to which elite wealth in Carthage was based on landholding versus investments in maritime trade. Hoyos (2010): 124 sums up what can be known: The rich élite who ran Carthage’s affairs were pretty certainly not just a merchant oligarchy but a more complex blend. Merchants of course bought property, in Carthage and outside, with some of their profits: this was a reliably safe investment. Meanwhile some wealthy Carthaginians no doubt held most of their wealth in land, and their opportunities for acquiring more land would grow as the city’s control over Libya did.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Punic Carthage

265

countryside was planted with vines, whilst other parts bore olives and were packed with trees bearing other kinds of fruit. On each side of the plain grazed herds of cows and flocks of sheep, and the neighbouring meadows were full of horses at pasture. In general there was in these places prosperity of every kind, since the most illustrious Carthaginians had marked out there their estates, and with their wealth had beautified them for their enjoyment.²⁵

We learn more about Carthage’s rich chora from Polybius (1.29.7), who narrates a Roman razzia into the district around Aspis in 255 BC during the First Punic War: The Romans, after taking control of Aspis, and having left a garrison to guard the city and its territory, hereafter also sent a delegation to Rome to announce news of what had occurred, and to ask what they should do next and how to take advantage of the situation; and after this, they hurriedly broke camp and proceeded with their whole force to ravage the territory. And because nobody stood in their way, they wrecked many wonderfully furnished houses, seized great spoils of livestock, and brought back to their ships over twenty thousand slaves.

This campaign was far from a systematic ransacking of the Carthaginian chora; yet it netted a substantial haul of 20,000 slaves.²⁶ Further evidence backing up this picture can be found in the treaty following Carthage’s defeat in the Second Punic War (218–201 BC). The very first clause states that the Carthaginians shall remain in control of the towns of Libya that they controlled before the war of Rome, as well as the territory that they held, with their flocks, slaves, and other property (κτήνη καὶ σώματα καὶ τὴν ἄλλην ὕπαρξιν: Polyb. 15.18.1). An overall estimate of slave numbers is impossible. But in a speech delivered in the Roman senate during the debate over the surrender of Carthage that resulted in the aforementioned treaty, one of Scipio’s associates pleaded for clemency; he pointed out that if the Romans returned to fighting, their Carthaginian foes remained far from an easy prospect, and, if backed into a corner, could prove very dangerous: particularly notable were their numerous warships, their large numbers of elephants, their generals Hannibal and Mago, and their large slave population (App. Pun. 9.59). Furthermore, it seems rather unlikely that this large slave population was a recent acquisition: Justin (21.4.6) tells us of Hanno the Great’s attempted coup d’état during the 340s, when he raised a force of some 20,000 slaves (again, there is no indication that this was any more than a fraction of the overall slave population) but was defeated and tortured to death. And even earlier—that is, shortly after the disastrous Athenian expedition to Sicily (415–413 BC) during the Peloponnesian War—the ²⁵ A Carthaginian country villa of this sort has been excavated at Gamarth: Docter (2009): 183. ²⁶ Kostas Vlassopoulos rightly points out to me, though, that the term somata used here is ambiguous: these captives could comprise formerly free persons as well as individuals who had previously been slaves.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

266

Greek Slave Systems

Carthaginians sacked the Doric polis of Selinus and brought over 5,000 captives back to Africa as slaves (Diod. 13.58). Though a lack of sources shrouds the early development of Carthage’s slave system in obscurity, it is clear that it was already operating on a large scale at least as early as the beginning of the fourth century, and quite possibly rather earlier than that. As for figures of individual slaveholdings, we lamentably have only Polyb. 38.8.4: this passage pertains to events in 147 BC, just before Carthage was finally destroyed by the Romans. Polybius tells us that Scipio Aemilianus offered Hasdrubal the Boetharch a deal: he promised to protect his personal safety, that of his family, and that of the family of ten of his friends; furthermore, Scipio stated that Hasdrubal might keep ten talents of his own fortune and a hundred of his own slaves. Evidently, this was a small fraction of his overall wealth (cf. Polyb. 38.8.11) and the hundred slaves a fraction of his overall holdings. The richest Carthaginians, then, must have had slaveholdings well in excess of their classical Athenian counterparts. Even less clear is the role of slavery in Carthage’s overseas territories. E. Flaig has underscored Polybius’ discussion (ap. Str. 3.2.10) of the silver mines of Carthagena, which were taken over by the Romans and were in his day worked by some 40,000 men. Flaig argues that the Romans simply continued earlier Carthaginian practices and that in relation to these mines we are looking at slavebased silver mining on a larger scale than that ever achieved by the Athenians at Laurion.²⁷ These passages thoroughly support Flaig’s conclusion: It does not matter if the documentation is lacking or not: the Phoenician fringe of the Western Mediterranean was by no means a “periphery”, but was, in terms of slavery, a very powerful metropolis, a slave society on a grand scale, certainly more important than that of Athens in the Classical period, and in some respects—viz. landlordism and the concentration of slaves—not unlike the Roman system of the late Republic.²⁸

And as the senate’s eagerness to produce a Latin translation of Mago’s agrarian treatise shows, Carthage’s large-scale system of agricultural slavery lived a shadowy Nachleben beyond 146 BC in the agrarian slaving techniques that the Romans cribbed and then applied to their own expanding system of slavery in Italy. It is perhaps fitting to end this book’s selection of case studies with Carthage, for it shows that the largest slave society of antiquity, Rome, did not spring up in splendid isolation, but drew at least in part on the technical expertise in agrarian slavery of its Western Semitic neighbours. ²⁷ Flaig (2009): 55–6. ²⁸ Flaig (2009): 56: Es ist gleichgültig, ob uns die Dokumente fehlen oder nicht: Der phönizische Rand des Westmittelmeers war keineswegs eine , sondern unter dem Aspekt der Sklaverei eine sehr leistungsfähige Metropole, eine Sklavenhaltergesellschaft großen Stils, sicherlich bedeutender als die athenische der klassichen Zeit und in mancher Hinsicht—Großgrundbesitz und Sklavenkonzentration—der spätrepublikanisch-römischen nicht unähnlich.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Part IV Why Slavery?

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

14 Differentials in the Magnitude of Slaveholding Towards an Understanding of Regional Variation

This book has, so far, presented detailed historical soundings via a series of Greek and Near Eastern case studies, varying in terms of chronology and geographical location, as well as the quantity and quality of available evidence. I hope that what has been presented so far is sufficient to dispel the black-andwhite contrast painted by Finley between the practices of the Greeks and those of their eastern neighbours. However, I have not argued that all of the societies examined in these case studies qualify as ‘slave societies’ on either the definition of Finley or that of Hopkins: Carthage surely does qualify, and probably monarchical Israel as well; good cases, I think, can also be made for the importance of slavery to elite income in certain regions of the Persian Empire. But whilst slavery played a not insignificant role in generating elite income in Assyria during the eighth–seventh centuries and in Babylonia from the seventh–fourth centuries, these regions are unlikely to make the cut as ‘slave societies’ on either of the aforementioned definitions. It is in such cases that the deficiencies of the black-and-white Finleyan approach to ‘slave society’ noted in Chapter 4 become obvious, and where a different approach, such as Lenski’s intensification model, can prove more useful. This allows us to avoid discarding such cases simply because they do not make the cut, and realize that they do, nonetheless, display many of the same features as full ‘slave societies’, though developed to a lesser degree. However, it is one thing to move towards a more nuanced approach that places historical societies along a scale of differing intensity rather than straightjacketing them in the traditional Finleyan categories, and quite another to explain why various ancient societies might occupy different locations on that scale. How might we explain these regional variations? Why was slavery more thoroughly integrated into some ancient economies than others? This concluding chapter aims to make some progress towards an explanation. The

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

270

Greek Slave Systems

emphasis on progress rather than a complete solution is intentional: the lacunose nature of our evidence, plus the challenges of technical competence that confront any one historian who tries to take all of these regions into his or her purview, means that any claim at having achieved such a solution would be premature. That point notwithstanding, some progress towards an integrated model of why slavery flourished in some regions to a greater degree than others is, to my mind, possible. What I present here, then, is designed to push the debate forwards, without any pretence of crossing the finishing line. I hope that scholars who specialize in the study of the various regions covered in this book (and others that have not been) will be able to engage critically with this framework and fine-tune (or indeed challenge) my arguments. * * * The task undertaken here is not a new one: Herman Nieboer aimed at a general explanation for the rise of slavery in historical societies over a century ago¹; closer to our own time were the efforts of M. I. Finley to account for the rise of slavery in Greece, which, if wrongheaded in its immediate purposes (see Chapter 5 above), did at least provide the skeleton of a general interpretative framework.² Nor can we ignore the more nuanced, though less famous efforts of István Hahn to achieve something similar, but with a keener eye on the Near Eastern world and a wider matrix of causal factors.³ More recently still, Walter Scheidel has made significant progress on this question, and Kyle Harper’s work has added further nuances.⁴ Section I of this chapter will begin at the local, indeed individual level by setting out a model of labour choice based principally on the work of Finley, Hahn, Scheidel, and Harper, grounding it in an Athenian context for illustrative purposes. Section II will then step back from this focused perspective and add two further variables to the mix: political geography and economic geography. I argue that these factors were key in conditioning the relative costs of slave labour in different areas of the ancient world and thus directly impacted on the choices made by individuals who sought to exploit the labour of others in this or that locale. Section III examines the forces that shaped the Aegean slave trade and builds an interpretative framework regarding the slave supply that can be applied to other regions of the ancient world. Section IV argues that the role of markets

¹ Nieboer (1900). Approaches that emphasize the factors of labour and land continue to produce useful insights, e.g. Bernard (2016). ² That is, three key factors necessary for the emergence of a ‘slave society’: (i) private ownership of land; (ii) a rudimentary level of commodity production and markets; (iii) the unavailability of an internal labour supply. See Finley (1980): 86. ³ Hahn (1971) outlines six factors: (i) primitive technology, meaning that large labour inputs are needed; (ii) private ownership of the means of production; (iii) commodity production and markets; (iv) productive units of sufficient size to require numerous labourers; (v) the ability to organize labour; (vi) access to slaves. ⁴ Scheidel (2008); Harper (2010); (2011a): 149–62.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude

271

and the slave trade left an indelible mark on the histories of different regions of the ancient world: denizens of some regions were able to exploit the opportunities afforded by market exchange to import slaves, and significant economic and lifestyle benefits ensued for them. The areas from which slaves were imported in significant quantities, however, were doubly blighted: not only did the slave trade rob them of significant numbers of people; it also incentivized localized predation and impeded processes of state formation in the long run. Section V expands the scope of the model built in sections I–IV to the Near Eastern world, and sets out an argument that can, in part, explain the differences in slaveholding practices between the various societies studied in Part III of this book.

I. CHOOSING A LABOUR FORCE The following framework summarizes those factors emphasized in recent work that influenced the choice between slaves and other forms of labour in ancient economies.

a.) Monetary Costs of Slavery versus Other Forms of Labour The need to acquire labour could be the result of a labour shortage (as per the Finley model) or could reflect the existence of market demand for various products, allowing for expansion of the overall labour force to produce them.⁵ One of Scheidel’s main contributions to our understanding of choices made in ancient labour markets is his emphasis on a basic economic fact: the relative monetary costs of slave labour versus its alternatives.⁶ For the owner of an estate, workshop, bank, or other enterprise in the ancient Mediterranean world, multiple forms of labour were potentially available; the most common options were slaves, tenants, and wage labourers. Differing market conditions across the ancient world, however, affected the relative costs of these options in specific locales. We know of cases where wages were relatively high and slave prices relatively low, such as classical Athens. Roman Egypt, on the other hand, provides a different vista, with a thick rural labour market and relatively low wages, but relatively high slave prices.⁷ The fact that slavery played a greater role in the Athenian economy than in Roman Egypt is a direct consequence of this fact, and it required only a relatively simple calculus in order to grasp which form of labour would prove more profitable in the ⁵ Hahn (1971): 34; Scheidel (2008): 116; Harper (2011a): 152. ⁶ Scheidel (2005): 14–15; (2008). ⁷ Scheidel (2005); (2008): 123–4.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

272

Greek Slave Systems

long run.⁸ These regional differences in market conditions were, as Harper has emphasized, functions of both demand and supply; and any model that seeks to explain these market conditions must reckon with both of these variables.⁹ I will add some further thoughts on this issue in section II below. For now it is key to grasp that the relative monetary costs of slave labour and its alternatives constitutes the most fundamental underlying cause of regional variation in the exploitation of slaves across the ancient Mediterranean world.

b.) Institutional Advantages of Differing Labour Forms That observation, however, requires further nuance. A second point that has been emphasized in recent work, especially that of Harper, is that although slave labour and wage labour may be in competition, they are not straightforward substitutes: inherent, institutional aspects of slavery may place it at an advantage in the context of specific activities and at a disadvantage in others.¹⁰ For example, the permanent duration of slavery lent it well to occupations that required long-term knowledge acquisition, and also to occupations that were physically hazardous or otherwise unpalatable. Mining at Laurion, for example, was both hazardous and—in certain respects—technically complex.¹¹ Slavery equipped the would-be employer with the legal instruments both to retain the worker’s services for as long as he wished and to compel the slave to perform tasks that were dangerous or degrading, and thus avoided by all but the most desperate of free workers. On the other hand, slavery required a high capital input (i.e. buying the slave) and a potentially lengthy period of training in certain occupations. This made buying slaves ill-suited to short-term tasks.¹² Certain employers might opt for a flexible, mixed workforce: we noted in Chapter 5 that writers as early as Hesiod recommended a core workforce of slaves on small farms, supplemented by hired seasonal labour that could easily be turned out when it was no longer needed. Such strategies optimized the institutional advantages of both labour types in the context of the agricultural year, with its attendant fluctuations in labour requirements. It is important to note, though, that, despite the legal similarities between Greek and non-Greek ⁸ I do not find convincing the idea of Kyrtatas (2007) that, because we find no textual traces of weighing up the profitability of slavery versus its alternatives, such a procedure never occurred to the Greeks. For calculations of profit and risk in classical Greece, see Christesen (2003); Faraguna (2008); Bresson (2016): 16–25. On economic rationalism in the Greek world, see also Lewis (forthcoming c). ⁹ That goes for alternatives to slavery as well as to slave prices: Scheidel (2017), for instance, shows that the real costs of slaves in Imperial Rome and China were roughly similar, but that slaves were used more heavily in the former than the latter. Real slave prices, then, need to be contextualized alongside alternative labour forms and their costs. ¹⁰ Harper (2011a): 159–61; but see already Finley (1980): 74. ¹¹ See Rihll (2010). ¹² Epstein (2008).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude

273

forms of slavery set out earlier in this book, there was a certain degree of regional variation in terms of the legal features of other, potentially competing forms of labour. The kind of bound tenancy found in the Neo-Assyrian heartland, for instance, has no mainland Greek parallel, and the ability to constrain the tenant’s movement inherent in this specific institution made it qualitatively different from tenancy in Attica. The relative advantages of slavery, therefore, have to be treated on a regional basis, with adequate attention paid to the specific institutional features of the various local labour alternatives.

c.) Cultural Variables A further complicating point is the issue of ideology and honour. At Athens, for example, a lengthy process of historical evolution forged an ideological outlook among the citizenry that was deeply opposed to long-term employment by another citizen.¹³ This acted as a barrier to the entry of free labour in the labour market, minimizing the amount of free hired labour in the Athenian economy and skewing choices towards the acquisition of slave labour. Certain occupations, such as wet-nursing, were so associated with slaves that only in desperate circumstances would a free person practise them (Dem. 57.35). Another example can be seen in Athens’ demosioi, public slaves. Paulin Ismard has recently argued that the Athenians preferred to delegate expertiseheavy tasks associated with the running of the democracy to slaves, forestalling any political capital citizens could potentially amass through the performance of expert roles, and thereby maintaining civic equality among the citizenry.¹⁴ The degree to which this occurred will no doubt be debated; but it is obvious that slaves were useful in performing politically sensitive roles like barring unruly citizens from the theatre or performing executions, roles that free citizens could never legitimately exercise.¹⁵ These cultural factors distorted the labour market away from the sort of simple scenario sketched in (a) above.

d.) Dynamics of Labour Use Beyond fixed institutional aspects of different forms of labour (b above), choice between them might be conditioned by specific variables that play out on the ground in a given locale. For example, we noted in Chapter 8 that, for a member of the Athenian liturgical class, the choice of a labour force could be affected by particular features of his wealth portfolio: if he inherited a single plot of land or several plots in close proximity, he might choose to invest ¹³ See Chapter 8, section I, p. 169 above. ¹⁵ See Harris (2013a): 38–44.

¹⁴ Ismard (2015).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

274

Greek Slave Systems

more heavily in slaves than would be the case were his landholdings small, numerous, and scattered; in the latter case, we might expect a heavier reliance on leasing.¹⁶ The factor of risk could lead to a spreading out of investments in several areas; the factor of minimizing time commitments could, on the other hand, lead to consolidation of interests or a greater reliance on managers. Our model of labour choice, then, should accommodate such individual circumstances and not boil down labour choices to a single, regional stereotype.¹⁷

II. P OLITICAL AND E CONOMIC GEOGRA PHY We have noted different market conditions that affected the price of slaves in different locales, but I have not explained regional variations thereof in any detail beyond vaguely invoking the forces of demand and supply. We can better understand these variations, however, if we contextualize them in geographical terms. At this point I would like to introduce to the discussion an approach recently pioneered by Jeffrey Fynn-Paul, who has made a rule of thumb distinction between two kinds of historical society: those which erected ‘no-slaving zones’ and those which did not, which Fynn-Paul terms ‘slaving zones’.¹⁸ The former were created through two kinds of historical process: state formation (or the formation of empires) and the spread of certain monotheistic religions. To stay with our Athenian example, Solon’s reforms provide a good example of the former process: he banned loans on the security of the body, minimized the avenues by which a free resident of Attica could fall into slavery, and even claimed to have repatriated some Athenians who had been enslaved and sold abroad.¹⁹ His innovations triggered reconfigurations in the Athenian slave supply, and by the classical period imports of barbarian slaves constituted a far larger slice of the slave supply than can have possibly been the case in Solon’s time; harsh penalties, too, were imposed on those who contravened the law and attempted to enslave insiders.²⁰ This whole process forced the Athenians to acquire their slaves either from the reproduction of existing slaves or from sources beyond the borders of Attica. Yet this kind of institutionalization of a ‘no-slaving zone’ was hardly a Greek invention. Franz van Koppen has documented something similar in Mesopotamia a millennium earlier than Solon, where Old Babylonian slave sale contracts from the time of Abiešuh guarantee that the slave being sold was either home-born or a ¹⁶ See Chapter 8, section II, pp. 174–5 above. Cf. Harper (2011a): 149–62. ¹⁷ See Harper (2011a): 149–79. ¹⁸ Fynn-Paul (2009). ¹⁹ [Arist.] Ath.Pol. 2.2; 4.4; 6.1; 9.1; Plu. Sol. 23.2; Solon fr. 36 [West] and Harris (2006): 249–69. ²⁰ [Arist.] Ath.Pol. 52.1; Din. 1.23; Harris (2013a): 51–2.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude

275

foreigner from outside the Babylonian kingdom: ‘the sale of slaves was only then legally possible if they fell into these groups’.²¹ As for monotheistic religions, Islam provides a prime example: Islamic law outlawed the enslavement of co-religionists, and as the frontiers of the Islamic world—the Dar ul Islam—expanded, they pushed the frontiers of the Dar ul Harb (‘the Abode of War’)—or what Fynn-Paul would call ‘slaving zones’—ever farther away from the core territories in Arabia from which Islam emerged. Medieval and early modern Islamic societies thus gained the majority of their slave imports from the pagan (i.e. non-Islamic) societies of the Slavic world and the African interior.²² No doubt there is something of the Idealtypus about these contrasting concepts, and historical realities are unlikely to fit neatly into these two boxes if they are strongly defined, the former type as hermetically sealed from all sources of internal enslavement, the latter marked by general mayhem.²³ We might note, for example, regions like Israel in the Iron Age II period, where there were certainly legal and intellectual currents aiming at the construction of a ‘no-slaving zone’, at least in terms of the imagined community of the Israelite/Judahite ethnic group. These currents, as we noted in Chapter 9, were not entirely successful, most strikingly attested by Jeremiah 34:8–22 (cf. Neh. 5:1–13). We might, then, think of locating real historical societies along a sliding scale, with comprehensive protection for insiders at one extreme, utter vulnerability to predation at the other, but few historical societies fully approximating to either of the two Idealtypen. Nevertheless, Fynn-Paul’s approach retains considerable heuristic value for modelling economic dynamics at the macro-scale. Let us imagine the situation, then, as it pertains to Athens in the classical period. The Athenians created a true ‘noslaving zone’, insofar as their legal system reduced the enslavement of free residents of Attica to negligible levels. Attica itself was nested within a mosaic of other Greek polities, which probably had (to greater or lesser degrees, depending on the region) some version of ‘no-slaving zones’ themselves.²⁴ The residents of these surrounding Greek polities were, of course, fair game as potential slaves during times of war, though it should be noted that warfare was intermittent and usually localized; furthermore, a growing cultural trend over time led to disapproval over the enslavement of fellow Greeks (a trend which, if it dampened the true potential of other Greek communities as

²¹ van Koppen (2004): 12. ²² Fynn-Paul (2009): 20–4. ²³ This point is, indeed, accepted by Fynn-Paul (2009): 7–12. ²⁴ Note, though, that even in the time of Thucydides some regions of Greece adhered to the old Homeric system of free-for-all raiding and kidnap (Thuc. 1.5); this was not truly halted until the coming of the Romans. See Scholten (2000); Gabrielsen (2001). Asylia agreements inserted no-go zones for raiders into an external political terrain in which—for the Greeks of certain traditional regions—all foreigners were viewed as fair game for seizure: Kvist (2003).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

276

Greek Slave Systems

sources of slaves, never quite got rid of the problem).²⁵ These political factors thus tended to shift the arena of potential supply towards the nearest ‘slaving zones’, and Attica was located in an almost ideal position in terms of access to slaves, for it was hemmed in to the east by the major slaving zone of Anatolia and to the north by the major slaving zone of Thrace. It also had ready access through its network of merchants and markets to the farther-flung slaving zones of the Black Sea and the Levant.²⁶ This is where economic geography becomes important in our overall assessment of the problem. One of the major factors emphasized in recent work that utilizes the insights of New Institutional Economics is that of transaction costs.²⁷ Commodities do not move in a frictionless market, and various kinds of costs can impede the smooth allocation of resources. In relation to the slave trade, one of the most significant costs relates to the logistics of transport. As we noted in Chapter 8, the costs of land transport have often been exaggerated in terms of local trade; but if we look at longdistance trade, the traditional view that transport over land could be highly expensive still rings true. The costs associated with transporting slaves are several.²⁸ Even though slaves can move under their own steam in coffles, they still have to be fed and supervised; costs mount at every point the slave is sold from one merchant to another, for each merchant has his profit margin about which to worry.²⁹ Furthermore, taxes on sales at the various stages of the journey increase the costs of the process and must perforce be passed on to the buyer. Long journeys, especially over land, can lead to the slave’s physical deterioration, which impacts on the price the slave will fetch at market and can even lead to losses of life among slaves along the route.³⁰ It should be clear,

²⁵ The number of Greek slaves attested in the Delphic manumissions is one striking sign of this: Lewis (forthcoming d). Disapproval of enslaving Greeks: Garlan (1987): 17–23. ²⁶ Thrace: Velkov (1964); Anatolia and Syria: Lewis (2011); Black Sea: Fischer (2016). ²⁷ Bresson (2016): 19–22. ²⁸ Lewis (2015a): 321–3. ²⁹ This has been graphically documented in Jeff Eden’s brilliant recent study of the slave trade in central Asia during the eighteenth–nineteenth centuries AD. To give a flavour of how this factor worked in practice, we might consider the fate of a man from Herat named ῾Abdullāh, who was kidnapped in 1838. As Eden (2017): 929 relates: ῾Abdullāh was sold for 8 tillas to a Turkmen named Berdi-Kilich, who in turn sold him at a profit to a Khivan two months later, earning 12 tillas. The sale evidently took place in the Turkmens’ camp near Merv. The Khivan, named D___, immediately took ῾Abdullāh to the town of Urgench and sold him to another Khivan, this one named Palvān Niyāz, for 18 tillas. After five days, Palvān Niyāz sold him to another Khivan, identified as ‘yuzbāshi Yakub’, for 25 tillas. In less than three months, ῾Abdullāh had already been sold four times. Eden (2017): 932 further notes that by the time of ῾Abdullāh’s final sale (for 37 tillas) his market value had increased to nearly 500 per cent of his original purchase price. ³⁰ Scheidel (2005): 10 discounts the importance of transport costs in the slave trade, but does not note (a) costs of supervision; (b) costs of feeding the slaves; (c) deterioration in the physical condition of slaves transported long distances by land; (d) resales of slaves of the sort described in n. 29 above. He does, however, note the effect of market tolls.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude

277

then, that relatively short transaction chains that can deliver a slave from his or her point of origin to their final point of sale with the fewest instances of resale and the fewest tariffs—and preferably covering as much of the travel by sea rather than by land—were relatively efficient. Lengthy journeys by land routes, on the other hand, made the whole process more expensive and less attractive for merchants than dealing in other commodities. All this was structured by political geography: when a state banned the enslavement of insiders, it altered the architecture of the labour supply: new slaves had to be sought from the nearest ‘slaving zone’ or from the reproduction of existing slaves; and the proximity of the former, as well as the scale of potential supply it could provide, mattered a great deal for the price of slaves paid by end buyers. When approaching the problem of price differentials for slaves, then, it may help to envisage the broader geography—political and economic—into which the slave supply fitted.

III. THE DYNAMIC FORCES OF THE AEGEAN SLAVE TRADE Without essaying a comprehensive sketch of the Aegean slave trade, we may for now note several dynamics that shaped its structure and attempt to gauge their effects. Once this framework has been set out, we can use it to understand the problems of slave supply in other, less well-documented ancient societies.³¹ The first dynamic of which we may take note is the sheer power of dispersal that the networks of merchants and markets, criss-crossing the Aegean and branching outwards into the Black Sea to the north and the wider Eastern Mediterranean to the south, enabled. Instrumental in all of this was the sea itself. Port towns functioned as nodes drawing all manner of goods—slaves included—from adjoining hinterland regions; these nodes in turn linked local regional trade to networks of interregional maritime commerce. The relative efficiency of waterborne transport in comparison to overland travel significantly enhanced the potential distances these goods could be traded. A striking illustration of how this affected the trade in slaves can be seen in a Hellenistic

³¹ This builds on several preparatory studies, whose conclusions I will not repeat in full here: Lewis (2011); (2015a); (forthcoming d). One should not be too pessimistic about the state of our evidence: in terms of the geography of the slave trade and the origins of slaves, we have considerably more evidence at our disposal than, e.g., scholars of the early medieval Mediterranean world: cf. McCormick (2001): 249–54.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

278

Greek Slave Systems

funerary monument from the Cycladic island of Rheneia, which bears the following inscription (SEG 23:381):

4

8

12

16

20

Ἰσίδωρε Μαιῶτα, Δαμᾶ Μαιῶτα Ἰσίδωρε Ἀπαμεῦ, Βῖθυ Ἰστρια⟨ν⟩έ Καλλιόπη Ὀδησσῖτι, Ὁμόνοια Ἑρμόλαε Ῥωσεῦ Ἀντίπατρε Μαζακηνέ Ἀσκληπιάδη Σιδῆτα Ἀπολλωνίδη Μαρισηνέ Νικηφόρε Ἰοπεῖτα Μενέλαε Μαραθηνέ Ποσῆ Μαραθηνέ Ἡρακλείδ⟨η⟩ Μαιῶτα Νικία Μαιῶτα Ἀμμωνία Κυρηναία καὶ θυγάτ⟨η⟩ρ Ἀπολλωνία Νικήρατε Ἀπαμεῦ Λαοδίκη Ἀπάμισα Δάμων Μύνδιε Ζαιδε Ναβαταῖε Δαμᾶ Ἰστριανέ οἱ Πρωτάρχου χρηστοὶ χαίρετε.

Isidoros of Maeotis; Dama of Maeotis; Isidoros of Apamea; Bithys of Istros; Kalliope of Odessos; Homonoia; Hermolaos of Rhosos; Antipatros of Mazaka; Asklepiades of Side; Apollonides of Marisa; Nikephoros of Joppa; Menelaos of Marathos; Poses of Marathos; Herakleides of Maeotis; Nikias of Maeotis; Ammonia of Cyrene and her daughter Apollonia; Nikeratos of Apamea; Laodike of Apamea; Damon of Myndos; Zaidos of Nabataea; Damas of Istros; the khrestoi of Protarchos Farewell.

It is not known what manner of accident precipitated the (apparently) simultaneous deaths of these twenty-two slaves of Protarchos.³² But their various origins—from locations as disparate as the Eurasian steppes, Cyrene in North Africa, and the Nabataean desert—testify to the powerful currents of commerce in the Hellenistic Eastern Mediterranean and their ability to subsume and relocate individuals across a vast geography.³³ On the other hand, complete Brownian motion in the distribution of individuals via the slave trade was hampered by other forces. We might note two in particular: slave agency and the costs of transport. Aegean slavery was forged through a push-and-pull between market forces, on the one hand, and, on the other, the efforts of slaves themselves to forestall this motion, to set down roots, to build families, and to disburden themselves of the methods of legal leverage their owners held over them by striving towards manumission. Epigraphy provides eloquent testimony to these conflicting forces. Take, for instance, the following epitaph from Lindos on Rhodes (IG XII,1 881):

³² See Couilloud (1974): 192 for suggestions. ³³ For notes on the toponyms in this document, see Lewis (forthcoming d).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude Φρόνιμος Γαλάτας. Ἀρτεμισία Σύρα. Εὐφρόνιος ἐγγενής.

279

Phronimos, a Galatian. Artemisia, a Syrian. Euphronios, locally born.

This inscription captures both the hard economic and legal facts that shaped the life trajectories of individual slaves, as well as hallmarks of their own agency. Two individuals, one from Galatia, another from Syria, are uprooted and cast into commercial circulation by processes unknown; they are bought, and end up at Rhodes. The fact that they bear Greek names is a likely sign of renaming by their new owner or owners. Yet, despite their different backgrounds, they meet one another, form a union, and have a child, Euphronios; and they left a trace of their lives that is still with us today, in the form of this gravestone. It is also an eloquent testimony of how ethnicities can intermingle, yet disappear within a generation; for if the upper portions of this stone were damaged we might never have guessed at Euphronios’ mixed origins. Occasionally, onomastics can furnish clues to the ethnicity of home-born slaves’ parents; but such instances are the exception rather than the rule.³⁴ Second, we should note that ancient Greek interregional trade was far from frictionless. Since both the Greek settlements that imported slaves and the locations of key slaving zones that supplied them were widely distributed in geographical terms, proximity to such a supply zone—paired with the volume of slaves exported from each supply zone—appears to have significantly shaped the ethnic make-up of any given Greek slave system. This follows a basic economic logic, neatly summarized by Bresson: Transport costs, which are a function of the unit value of the product and the distance, constitute a constraint that structures trade networks. Observation shows that even today, in most cases, the intensity of trade relations is a function of distance: for economies of the same size trade partners are distributed on a gradient corresponding to the distance between them. This was the case a fortiori in an economy like that of ancient Greece, because transport costs were relatively much higher than they are now, and there were considerable differences in the value of the goods transported.³⁵

³⁴ For instance, SGDI II 1884, an inscription from Delphi dating to 165/4 BC, records the manumission of Θραικίδας τὸ γένος ἐνδογενῆ (‘MacThracian, of home-born stock’). In the vast majority of instances, though, the designations endogenes and oikogenes obscure any trace of the ethnicity of the slave’s parents. ³⁵ Bresson (2016): 364. That is not to say that some slaves could not travel great distances, as the Protarchos inscription illustrates. Rather, it is to note that this dampening effect could limit the diffusion of slaves by market mechanisms so that a majority probably ended up closer to their initial point of dispersal. High-value slaves might be trader further than was usual: cf. Theophr.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

280

Greek Slave Systems

The Hellenistic evidence bears this out most visibly. For example, we might compare the number of Thracians and Phrygians at Rhodes and Delphi. At Delphi we find only four Phrygians out of all the manumission documents recording slaves of non-Greek origins (a sum of 179 individuals), whereas we find some thirty-four Thracians. If we look at Rhodes during the same period, the pattern is reversed: we find only three Thracians, but ten times as many Phrygians (out of a sum of 158 slaves of known foreign origins). It stands to reason that the greater proximity of Delphi to Thrace, and the greater proximity of Rhodes to Phrygia, played some part in this differential distribution of Thracian and Phrygian slaves.³⁶ The most striking evidence for this phenomenon derives from Ptolemaic Egypt, where nearly all the slaves of known origins came from Syria and the Levant.³⁷ Had we similar evidence for the origins of slaves at, for example, Corcyra, it would be surprising if we did not find many more Illyrians there than the handful that ended up at Athens (IG I³ 421.39, 43). These differing variables (location, transport costs) help to explain such regional variations. However, these factors alone provide for only a static model. It is through thinking in terms of diachronic shifts in demand and supply that we can set this model in motion and start to explain not just regional differences but also historical change. The number of slaves emanating from Syria, for instance, seems to have increased by the second century BC, for although Syrian slaves were commonplace in Attica during the classical era, they by no means outnumbered Thracians or Phrygians.³⁸ However, the picture changes during the Hellenistic period: Syrians not only completely dominate the Ptolemaic evidence, but are the largest single group in the Delphic manumissions (fortynine individuals), and are one of the largest ethnic components of the Rhodian dossier of epitaphs.³⁹ In other words, the sheer scale of supply of Syrian slaves during this period led to their significant presence in locales all over the Eastern Mediterranean, despite the greater distance between, for example, Delphi and the Levant in comparison to the more proximate supply zones of Thrace and Anatolia. The most notable and drastic change of market conditions, however, related to demand and was heralded by the gradual intrusion of Italian demand into the Eastern Mediterranean slave market during the Hellenistic period.⁴⁰ It is probably the entry of Italian demand into the market, paired with growing market integration across the Mediterranean, Char. 21.3 on rare and (presumably) expensive Ethiopian slaves. (For the representation of black African slaves in Hellenistic art, see Baratta 2014.) ³⁶ For an overview of the evidence relating to slave origins at Delphi and Rhodes, see Lewis (forthcoming d). ³⁷ Bieżuńska-Małowist (1974): 54–8; Scholl (1990): 213; Lewis (forthcoming d). ³⁸ See the evidence collected in Lewis (2011). ³⁹ Lewis (forthcoming d). ⁴⁰ Bresson (2016): 348; cf. pp. 371–4 for shifts and reorientations of trading networks; Scheidel (2005): 9–10.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude

281

that explains certain diachronic changes in the evidence for manumission at Delphi. Keith Hopkins and P. J. Roscoe have charted a growth in Delphic manumission fees over time, probably tracking a parallel growth in the market price of slaves as demand for slaves grew relative to a less flexible supply; in step with (and perhaps as a result of) this trend, we find an increasingly heavy reliance on the reproduction of slaves over the acquisition of new slaves on the market.⁴¹ ‘One explanation which immediately comes to mind is that the demand for slaves in Roman Italy outran supply . . . Central Greece was hooked by Roman conquest into a pan-Mediterranean economy, and so slave prices rose.’⁴² The factor of market integration may in part explain the price differentials between classical Athens and imperial Rome noted by Scheidel:⁴³ for the earlier period, a lack of market integration and a lack of western competition left the Aegean Greeks with a dominant share of the supply of slaves from Thrace, Anatolia, and Syria.⁴⁴ The situation was not so favourable several centuries later, when the tug of the Italian market and more integrated networks of commerce between the Eastern and Western Mediterranean shifted the variable of demand relative to supply, diminishing the advantage the Aegean city states had previously enjoyed and driving up prices.⁴⁵ (That is not to suppose that supply remained a fixed, inelastic ⁴¹ Hopkins and Roscoe (1978); but Duncan-Jones (1984) provides some valuable cautionary discussion regarding their figures: the average manumission fees for the later periods are drawn from a much smaller pool of data than for the earlier periods. Given that the scatter of slave prices in any synchronic snapshot will inevitably be reasonably wide due to a variety of factors (see Chapter 8, p. 170 n. 15 above), one would ideally need a large pool of prices to derive a tolerably representative average. Hopkins and Roscoe’s average prices for the later periods, drawing on a smaller data pool, are therefore less reliable. ⁴² Hopkins and Roscoe (1978): 159. ⁴³ Scheidel (2005). ⁴⁴ Note that these regions still constituted the most conspicuous sources of slave supply for Mediterranean markets well into the Imperial period: Harris (1980): 126–8. ⁴⁵ Note also that whereas recent estimates of the proportion of the population of Roman Italy made up of slaves have been revised downwards to around 15–25 per cent (Scheidel 2011: 288–9), there is no reason to question the well-established view that slaves may have made up over a third of the population of Attica in the fourth century BC (Bresson 2016: 459–60); this dovetails with a picture wherein the Aegean Greeks of the classical period were positioned more advantageously in terms of the slave market than their later Roman counterparts. Although we know almost nothing of the proportions of home-born versus imported slaves in classical Attica, the evidence of Hellenistic Delphi and Rhodes is suggestive: at Delphi during the period of the first four priesthoods (201–153 BC), of the slaves of known origins less than a third were homeborn, and over two-thirds imported (Westermann 1955: 32); this changed noticeably, however, during the ensuing period, perhaps due to changing market conditions (see above). At Rhodes, only thirty of the individuals of known origins from the period c.300 BC–AD 100 are described as ἐγγενής: the other 158 were imported (Lewis forthcoming d). Now, we should bear in mind that these dossiers are rather small, which affects their reliability. On the other hand, if home-born slaves were more likely to be manumitted than bought slaves (cf. Hopkins and Roscoe 1978: 166–7), then they are perhaps over-represented in these records (the Delphic material is, of course, comprised of manumission inscriptions; as for Rhodes, it is likely that the slaves who were able to erect grave monuments came from the more successful end of the spectrum and were thus more likely to have been manumitted, similarly skewing the sample). This suggests

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

282

Greek Slave Systems

factor across this time period, only that the shift in demand was more dramatic than whatever putative shifts in supply occurred, a process we can assume, but not quantify.)

I V . TH E E F F E C T O F S L A V E T R A D I N G O N S U P P L IE R Z ONES In his recent book The Making of the Ancient Greek Economy, Alain Bresson has done much to illuminate the core–periphery commercial relations between the Greeks and their non-Greek neighbours. Often these were of a mutually beneficial character.⁴⁶ But as Bresson notes, it was in the slave trade that real exploitation of peripheral zones was to be found.⁴⁷ This is a process that requires further exploration, and we may begin with some comparanda. Fynn-Paul’s work on slaving zones has unearthed a disturbingly consistent pattern in world history regarding the effects of market demand once it intrudes into a slaving zone: 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.⁴⁸

A detailed example of this can be seen in early modern Africa. Here, market demand from the Islamic world and Europe catalysed indigenous African forms of slavery and stimulated major—and cataclysmic—social transformations in the African interior. In his book Transformations in Slavery, Paul Lovejoy describes the process as follows: Of the approximately 2.2 million slaves shipped from Africa to the Americas between 1500 and 1700, almost 60 percent came from that region [i.e. west-central Africa]. In the eighteenth century, some 2.4 million slaves were exported from that these figures should be taken as maxima. As we noted above, market conditions were probably more favourable to slave-owners in the classical period, meaning that the incentives to allow slave families were less strong in, for example, fourth-century Athens. The consensus on the Roman side seems to hold that its empire’s slave supply relied on reproduction far more heavily (Harper 2011: 38–9), which was more sustainable in the long run given the empire’s sheer scale, but also more expensive. To be sure, we should not forget that many slaves in classical Athens may have wanted and struggled to form families, surely succeeding in many cases; but the Delphic and Rhodian comparanda count against any attempt to play up the importance of slave families and reproduction to the classical Athenian slave supply. Cf. Schmitz (2012). ⁴⁶ Bresson (2016): 348–51. ⁴⁷ Bresson (2016): 349. ⁴⁸ Fynn-Paul (2009): 11.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude

283

west-central Africa alone; the total migration from 1500 to 1800 exceeded 3.6 million people. Where did the people come from? How were they enslaved? . . . three phenomena made the migration possible. The first was the disintegration of the Kongo kingdom, which experienced civil war in the last half of the seventeenth century. Kongo was then a major source of slaves. The second involved Portuguese expeditions into the interior in an attempt to expand the colony and capture slaves. The third was the rise of Imbangala war bands that operated in the area where the Mbundu lived to the south of Kongo. The trend was toward political instability. Whereas once the kingdom of Kongo had dominated the area, now no central authority could enforce law and order. Instead, the period was one for the warlords, whether they were Kongo nobles, Imbangala chiefs, Portuguese commanders, or new warrior princes farther inland. The common element was the emphasis on war and plunder, and the consequence was enslavement on a massive scale. Without doubt, the majority—perhaps the vast majority—of the 1.3 million people dispatched from west-central Africa between 1500 and 1700 were a product of the interaction between these Imbangala, Kongo, and Portuguese warlords. The eighteenth century required the inclusion of areas farther inland to maintain the level of departures, but this meant that other warlords were added to the list of participants in the campaigns to supply slaves for export.⁴⁹

This level of chaos, it should be emphasized, did not represent a kind of African ‘state of nature’. It was external influence and—above all—external demand for slaves that acted as a catalyst stimulating warfare and raiding in collapsed-state zones. Here, local elites could grow rich by predating on both local people and their neighbouring enemies, trading slaves for a variety of products hitherto unavailable to them. External demand for slaves thereby inhibited processes of state formation in supplier zones and stimulated predation and violence. Did a similar pattern occur in antiquity? In fact, some ancient writers noted precisely this process and pinpointed its causes. Strabo, discussing the explosion of piratical activity in the eastern Mediterranean during the second century BC, wrote as follows (14.5.2): The reason for this was that the Romans, having become wealthy after the destruction of Carthage and Corinth, were making use of large numbers of slaves; and seeing how easy it was to make a profit, the pirates flourished in droves, themselves both raiding and trading in slaves. The kings of both Egypt and Cyprus collaborated in these things, because they were enemies of the Syrians; nor were the Rhodians friendly towards them, so that they gave them no assistance.

Political fragmentation in the eastern Mediterranean, zones of collapsed political authority, large numbers of dispossessed people (particularly sailors),

⁴⁹ Lovejoy (2011): 73–4.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

284

Greek Slave Systems

and a strong increase in demand for slaves from the west combined into a perfect storm that—until it was checked by Pompey in 67 BC—transformed the Eastern Mediterranean into a slaving zone pockmarked by the depredations of pirate gangs. Nobody was safe; even Julius Caesar was kidnapped (though as a high-value captive he was ransomed, not sold).⁵⁰ A similar process occurred in late antique Africa. Kyle Harper has drawn attention to a letter of Augustine discovered a generation ago that describes the wanton destruction of parts of Roman North Africa by slave raiders: these were ‘draining the greater part of the human race’ (Aug. Ep. 10*.2) and selling them on the Mediterranean market.⁵¹ Harper convincingly shows that this was not a freak occurrence, but represents the relocation of a standard process— one that usually occurred beyond the borders of the empire—to within its territory: ‘Augustine’s letter was written at a moment when one of the supply lines of the Roman slave trade was forced to cannibalize upon the empire rather than its usual victims.’⁵² As governmental structures weakened in this region, rendering permeable the ‘no-slaving zone’ normally demarcated by the empire’s border, the freewheeling predation that usually occurred beyond this frontier was pulled inwards by the tug of market forces, spreading chaos in its wake. What is particularly telling is that Augustine himself was able to pinpoint the cause of all this chaos: market demand, i.e. the fact that merchants hungry for profit were ready and present in North Africa, paying large amounts of money to the slave raiders for their captives. As he points out, had the merchants not been present, none of the resultant chaos would have occurred (mercatores autem si non essent, illa non fierent). What of the earlier period, c.450–250 BC? The slaving zones of Thrace and Phrygia, the principal reservoirs of slave labour for the classical Athenians, provide much food for thought. A valuable insight into Thracian warfare is sketched by Xenophon, who, along with the mercenary army returned from Cunaxa, joined forces with the Thracian warlord Seuthes II for a raid on some enemy villages at the turn of the fourth century BC (Anab. 7.3.43–7.4.2). After taking a thousand captives, two thousand large animals, and ten thousand small animals (Anab. 7.3.48), Seuthes fired the villages in order to strike fear into his enemies; on the next day, he sent the captives to the Greek port of Perinthus to be sold (Anab. 7.4.2).⁵³ This represents one of the larger actions of tribal warfare, but there were probably many other smaller-scale raids regularly occurring in Thracian territory. One wonders whether Xenophon had Seuthes in mind when he wrote (Symp. 4.36) that some tyrants were so

⁵⁰ See Rauh (1998); Gabrielsen (2003). ⁵¹ Harper (2011a): 92–4. ⁵² Harper (2011a): 94. ⁵³ From another port perched on the edge of Thrace, Abdera, we possess a fragmentary law of the fourth century BC regulating the sale of slaves and pack animals: see SEG 47:1026. Cf. Archibald (2015): 386. For the trading interests of the Odrysians, see Baralis (2009): 118–20.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude

285

enamoured of money that they would destroy families, kill men, and enslave whole towns in the pursuit of profit.⁵⁴ At any rate, Herodotus’ observation (5.3.1; cf. Thuc. 2.97.5)—that the Thracians were the most populous people in the world after the Indians, and would be invincible were they able to unite under one leader—underscores two important facts: the demographic density of potentially enslaveable persons in Thrace and the endemic warfare between hostile tribal groups that produced a constant stream of commodified humanity for Greek buyers.⁵⁵ The trade in slaves allowed the Thracians to import various products from the Aegean world in return. ‘You’re a noble Thracian, bought for salt,’ claims one of Menander’s characters (fr. 891 K-A), and Pollux (Onom. 10.14) informs us that such Thracian slaves were called halonetoi, ‘salt-bought’.⁵⁶ To judge from the sheer frequency of references to Thracian slaves in the Greek city states,⁵⁷ it seems difficult to avoid the conclusion that the Aegean Greeks did not merely siphon off an existing supply of war captives from Thrace: the enormous demands for cheap labour from the Greek city states surely accelerated violence in the Thracian hinterland. That would help to explain its relative lack of political development compared to neighbouring regions, the city-state culture of southern Greece, for example. Such an explanation may also provide a new angle on the question of the internal social dynamics of Phrygia in the centuries after the collapse of its Iron Age state. Recently, Peter Thonemann has made the attractive argument that the Phrygians, following this early collapse, avoided rebuilding state institutions as a means of escaping from some of the deleterious effects of living within a state structure.⁵⁸ Perhaps this is correct. But another possibility is that demand for Phrygian slaves provided a continual incentive for predation by local strongmen and inhibited the regrowth of state institutions that could have placed their victims off limits from, for example, seizure or enslavement for debt.⁵⁹ Excavations at centres of elite activity in Gordion

⁵⁴ The recent find in 2004 of a bronze statue head—probably depicting a later ruler, Seuthes III—vividly conveys the menace that such warlords wished to project. For an image, see Martinez et al. (2015) catalogue no. 82. ⁵⁵ Note also that the Thracians were reputed to sell their children as slaves: Hdt. 5.6.1. ⁵⁶ Θρᾷξ εὐγενὴς εἶ πρὸς ἅλας ἠγορασμένος. On halonetoi, see Alexianu (2008) and (2011), though his claim that the quote from Menander attests to noble Thracians owning halonetoi seems to me a misinterpretation: cf. Theophr. Char. 28.2, where the idea seems to be that what passes as noble among Thracians is only fit for slaves at Athens. On the salt trade, see Carusi (2015); for Thasian exports of wine to Thrace, see Tzochev (2015a); for an overview of Thracian trade, see Tzochev (2015b), esp. pp. 420–1 on imports and exports. ⁵⁷ See Velkov (1964), and for a useful list of references, Velkov (1967): 75–8. ⁵⁸ Thonemann (2013). A similar idea regarding the avoidance of coercive state institutions lies behind the recent book of Paulin Ismard (2015); these ideas ultimately go back to Clastres (1974). ⁵⁹ For enslavement for debt in Eastern Anatolia in the Neo-Assyrian period, see SAA V 149.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

286

Greek Slave Systems

and Daskyleion paint a revealing picture: steady, regular imports of Greek wine and fine wares over a long period.⁶⁰ What was offered in exchange? We have already noted the fame of Phrygian slaves as early as the time of Ezekiel (27:13) at the turn of the sixth century BC, and the regular export of Phrygian slaves continued to occur for many centuries thereafter.⁶¹ In sum, I am in complete agreement with Braund when he writes that ‘there seems every reason to suppose that the slave trade at the coast served to generate instability and conflict in the interior’.⁶² If we draw these strands of argument together, we can begin to appreciate the particularly advantageous position of the Aegean Greek city states in the wider Easter Mediterranean world, insofar as access to slaves was concerned. The various factors of geography, demography, economics, and politics combined in the Greeks’ favour. Above all, they stood in close geographical relation to the slaving zones of Anatolia and the eastern Balkans, regions with large populations of enslaveable individuals. The institutions of many of the city states both strengthened the legal protections from enslavement for their own citizens and facilitated those trade networks that provided easy access to these nearby reservoirs of slave labour. Much recent work has charted the material efflorescence of Greek city-state culture in the longue durée from c.800–300 BC.⁶³ Numerous factors have been enlisted to explain that efflorescence: the trend towards democracy and republican forms of political regime; strong citizen rights; low tax demands on the majority of the citizen population; creative methods of enlisting the competitive habits of elites to fund civic expenses in return for social and political capital; strong property regimes with reliable third-party enforcement; and so on.⁶⁴ Without wishing to dispute the significance of any of these factors, I think the list is incomplete without the addition of slavery and the slave trade as major institutional forces driving economic growth. These had a twofold effect. By flooding the Greek world with cheap slave labour, they undergirded the economic prosperity of citizens of the Greek poleis, and enabled a growing commitment on behalf of average citizens to civic life and civic responsibilities. But these gains for the Greeks have to be viewed in a wider context, namely the entropic effect of the slave trade on supplier zones. These regions, and the slaves exported from them, paid a hefty price for Greece’s material and cultural efflorescence.

⁶⁰ Gordion: Lewis (2015a): 319–20; Daskyleion: Yaldır (2011). ⁶¹ For the classical period, see Lewis (2011); for the Hellenistic period, see Lewis (forthcoming d); for the Roman period, see Harris (1980). ⁶² Braund (2011): 115. ⁶³ Ober (2015) and Bresson (2016) are prominent examples. ⁶⁴ Ober (2015); Bresson (2016); see also Harris (2006): 143–62; (2015), who stresses property rights and third-party enforcement.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude

287

V. BEYOND THE G REEK WORLD The preceding discussion has underscored several factors that are particularly significant to the development of slavery in a given locale. The cost of slaves is crucial and is a function of demand and supply, but also reflects the transaction costs involved in moving the slave from his or her point of origin to the ultimate buyer. Demand and supply involve a demographic component, i.e. the number of potential buyers relative to the number of available slaves. If there are many potential buyers, but few slaves, slave prices are likely to be high. Political geography is also important, given that politics shaped who was and who was not likely to be enslaved (and where), structuring the transaction costs involved in importing slaves. And the price of slaves must be contextualized in terms of the whole labour market and the relative costs of other forms of labour. To distil the above discussion into a matrix of variables, the ideal environment for slavery to proliferate might display the following features: (i) strong state institutions guaranteeing citizen rights, law and order, and market infrastructure; (ii) capital formation, especially if it is distributed relatively equitably across the citizen body;⁶⁵ (iii) high levels of commitment to civic activities and a strong disinclination among citizens to work for one another; (iv) markets for slave-produced goods; (v) geographical proximity to a slaving zone; (vi) commercial networks linking buyers and sellers, ideally by sea; (vii) high demographic density within the slaving zone (i. e. strong potential supply).⁶⁶ With this framework in mind, let us now consider those slave systems that we have examined in Part III of this book. Although our evidence for the systems of slave supply found in those other regions of the Eastern Mediterranean world studied in this book is decidedly worse than what we possess for the Greek slave supply, some tentative suggestions can be outlined nonetheless. Perhaps the closest parallel to the Aegean Greek situation is that of Carthage: treated just like a Greek polis in the eyes of Aristotle, Carthage shared a number of those structural variables that stood in the Greek city states’ favour in terms of the successful exploitation of slave labour. With a strong commitment to its citizenry, an expansive network of interregional trade interlocking with less developed peripheral regions (and exploited for a long time in a monopolistic manner), it is easy to see why Carthage’s slave system grew to the remarkable extent noted by Greek and Roman writers. The reasonably developed extent of slavery in the Levant and Achaemenid Egypt can be ⁶⁵ Cf. Ober (2010): 98: the “motor” of consumption powered by a tiny elite is relatively feeble. It is only with the emergence of a substantial and stable “middling” class of persons living well above the level of subsistence, and therefore willing and able to purchase goods unnecessary for their mere survival, that societal consumption becomes a strong driver of economic growth. ⁶⁶ Drawing on Hahn (1971); Scheidel (2008); Fynn-Paul (2009); Harper (2010; 2011a); Lewis (2015a); Lenski (2018b).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

288

Greek Slave Systems

explained in part through their integration with networks of Eastern Mediterranean trade: these gave city states like Tyre access to the slaves of Greece, Phrygia, and Cappadocia at least as early as the time of Solon, and enabled Persian satraps like Aršama to buy Cilician slaves for his Egyptian estates.⁶⁷ The situation with regard to the Mesopotamian slave systems of Assyria and Babylonia was very different. Their rich documentary records, as noted in Chapters 10 and 11, cast a dazzling light on certain aspects of social and economic life whilst leaving other areas in relative gloom; the geography of the slave supply is one area that is, regrettably, poorly illuminated due to a lack of evidence.⁶⁸ What we do have in abundance, though, is slave prices; and these show that, compared to—let us say—Athens during the classical era, slaves were relatively speaking much more expensive in Assyria and Babylonia, and slave ownership more upwardly distributed.⁶⁹ Several factors may explain this. The unquantifiable factors of demand and supply may simply have worked less favourably in these territories, for they were densely populated and highly commercialized (thus exerting high labour demands), yet were mostly surrounded by politically developed neighbours. They had not the ease of access to peripheral slaving zones that the Aegean Greeks enjoyed; perhaps too those external supplies that were available were not as abundant. And whilst it would be imprudent to speculate about the importance of these external suppliers given the poor state of the evidence, a few thoughts about transport costs may also be worth setting out. As we have already noted, Greeks, Phoenicians, and Egyptian-based Persians imported Anatolian slaves at one time or other during the first millennium BC; and all of these could rely on seaborne merchants to transport slaves from coastal ports in Asia Minor to their respective markets. We know that the Assyrians obtained at least some of their slaves from Anatolia;⁷⁰ but they had access only via lengthy overland trade routes that linked the Assyrian heartland to Anatolia through the northern Jezirah.⁷¹ Likewise, the modicum of evidence we have for foreign

⁶⁷ See Chapter 12, sections I–II, pp. 251–3 above. ⁶⁸ See Jursa (2010): 227–8; we are better informed about the Old Babylonian period, largely because sale contracts in that era mention the origins of slaves: van Koppen (2004). Had the formula persisted in later contracts, we would have a far richer knowledge of the Neo-Babylonian slave supply. ⁶⁹ Assyria: see Chapter 10, section II, pp. 229–30 above; Babylonia: see Chapter 11, section II, p. 239 above. ⁷⁰ See Radner (1997): 227–30; Jiménez, Adalı, and Radner (2015) for the Neo-Assyrian period; Bayram and Çeçen (1996) for the Old Assyrian period. Zadok (2010): 420 notes PNA 2: 771b, a slave whose personal name translates as ‘Muškean’, i.e. ‘Phrygian’. This is the Assyrian equivalent of slaves named Phryx in Greek sources. Most Assyrian slave sale documents do not mention the origin of the slave in question, but for an exception (a slave from Tabal), see SAA XIV 91. ⁷¹ For these roads and routes, see Kessler (1997); Astour (2000); Wilkinson et al. (2005): 32–7. For the Old Assyrian period, see Barjamovic (2011). Radner (1997): 227–30 discusses several

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude

289

slaves in Babylonia relates to individuals who must have been transported overland from sometimes quite far-flung regions.⁷² Whereas the demand– supply relationship in these regions is difficult to assess, we can at least then partially explain higher relative slave prices through higher transport costs.⁷³ Two other factors have to be taken into account regarding the ancient Near East. First (and if the extensive evidence for a large population of wage labourers in Babylonia is anything to go by), we may discount there the cultural barriers to working as employees of fellow citizens that we have noted at Athens.⁷⁴ Indeed, the labour market in Babylonia was more like that of Roman Egypt than classical Athens, insofar as slaves were outcompeted by the presence of a substantial number of free wage labourers whose wages were relatively low compared to the high costs of buying slaves. Second—most particularly in the cases of the Assyrian and Persian empires—is the effect of the command branch of the economy. This has traditionally been exaggerated, and recent attempts to rebalance our view of ancient Near Eastern economies away from the old redistributive palatial economy model are well taken;⁷⁵ but, such revisions notwithstanding, it is obvious that the mass deportations of populations enacted by Assyrian kings, and later (to a lesser extent) under the Persians, were well beyond the muscle of any Greek city state. These latter might reduce enemy cities to ruins, massacre the men, and sell the women and children into slavery; but this was on a relatively petty scale in comparison to

fascinating letters that provide an insight into the Anatolian slave trade. SAA V 149 is a plea to the Assyrian king to halt abuse in Hargu (in Eastern Anatolia), where indebtedness was rife and large numbers of people were being sold into slavery. SAA V 150 is a letter asking the Assyrian king to halt the illegal activities of a slave trader named Atarham, active in Hargu, who had ignored an edict of the Crown prince concerning slave trading in Assyrian emporia: Deszö and Vér (2013): 358 believe that the infraction was exporting slaves from imperial territory. ⁷² Dandamaev (1984): 107–11 collects evidence for slaves from Egypt, Iran, Gandhara (Kandahar), as well as slaves with Aramaic, Phoenician, and Hebrew names. In comparison with our Greek evidence, however, this evidence is very scanty indeed. ⁷³ We might note too the apparent prominence of slave families in the Neo-Babylonian period: Dandamaev (1984): 406 remarked on the fact that the majority of the hundred or so slaves divided among the heirs of Itti-Marduk-balat ụ (Dar. 379) lived in family groups, each with between one and four children. This contrasts sharply with the snapshots we possess of Greek slaveholders’ households such as Kephisodorus (IG I³ 421.33–49) and Protarchos (SEG 23:381), which are dominated by imported slaves. This fits with the trend noted in Delphi (see section III, pp. 280–1 above) where slave families tend to be less common in systems where slave prices are low and the supply of slaves plentiful. ⁷⁴ Jursa (2010): 660–81. But things may have worked differently in other regions: we find, for instance, in Israel an egalitarian ideology very much out of step with Mesopotamian norms (Faust 2006) in which the independent smallholder was held up as the ideal. Carthage’s republican institutions may also hint at a rather different ideology from what we find in Mesopotamia; these may have affected labour choices in practice. ⁷⁵ See, e.g., Jursa (2010): 19–26.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

290

Greek Slave Systems

the mass relocations of very large populations enacted by Near Eastern kings.⁷⁶ This affected the development of slavery in different ways. In the Assyrian countryside the settlement of large numbers of deportees and their incorporation as full Assyrians, on the one hand, and the allocation of large numbers of captives to members of the elite not as slaves but as bound tenants, on the other, probably offered cheaper labour alternatives in the countryside to what the slave market could usually provide, though without the same flexibility. On the other hand, the royal economy of Fars provides an example of the large-scale integration of enslaved war captives into a regional economy solely through the command mode, without relying on the vicissitudes of the market. Examples of this sort—or, for that matter, Sparta’s system of helotage, deliberately withdrawn from interregional trade—remind us that not everything can be explained through a market-centred model. In certain less commercially developed regions, agricultural work regimes could function well in the context of self-reproducing slave populations; whilst regions like Crete could, in addition, rely on the old archaic custom of the seaborne raid to snatch slaves from surrounding areas. * * * It is now time to draw the arguments set out in this volume together and to set out some conclusions. The following ten conclusions represent points of departure from orthodox or currently popular views and are proposed as potential modifications to our picture: 1. Despite creative attempts to argue the contrary, it is necessary to return to a basic understanding of slavery’s distinctive, defining feature being its legal, institutional aspect. Slavery is the ownership of human beings. To recognize this is not to deny the slave agency or to eschew a dynamic understanding in favour of a static one. Historical societies are made of fixed and moving parts; to claim that the fixed parts are irrelevant simply because they might be less stimulating to this or that scholar’s taste is to blind oneself to important aspects of reality. Nor should this legal understanding be viewed as the whole picture, for there is, of course, room for a variety of methodologies in studying so multifaceted a phenomenon. Rather, it is to recognize that the societies of the ancient Eastern Mediterranean world created their own indigenous legal institutions that distinguished between slaves and everybody else, and these were based on the criterion of property, not ‘domination’ or ‘social death’. The question ‘who is or is not a slave’ must focus above all on these legal institutions. (Other questions, such as the life experiences of slaves, are, of course, open to other methodologies; and I hope this study ⁷⁶ Assyria: see Chapter 10, section III, pp. 231–3 above; Persia: see Chapter 12, section III, pp. 254–8 above.

OUP CORRECTED PROOF – FINAL, 28/6/2018, SPi

Differentials in Slavery’s Magnitude

291

can find a pew in a broad church populated with other, complementary approaches that are applied to addressing different kinds of question.) Different societies produced varied vocabularies to describe this difference in legal status; but if we drill down to the level of sociolegal practices, we can see that in all of these societies the institutionalization of the ownership of human beings displays certain common features. These similarities of practice, understood through empirical study, are more important than the intellectual packaging given them by ancient philosophers and jurists. 2. The Greeks contributed nothing new to the conceptual distinction between slaves and non-slaves. Individuals in Near Eastern societies were just as capable of distinguishing slaves from other persons as the Greeks. Although an antonym to slavery is not necessary for making this distinction, the terminology of freedom was not a Greek idiosyncrasy, for the Babylonians too made use of comparable terms. This faculty of distinguishing free from slave had nothing to do with the relative economic role of slavery in this or that society; and the intellectual progression posited by Finley from a ‘fuzzy’ period in which various statuses shaded into each other to one in which free and slave resolved into sharp contradistinction in sixth-century Greece, has no evidentiary basis. 3. Laws, customary or written, legitimized this most unequal of relationships, sanctioning and undergirding the power of the owner over his or her slave. This most extreme of power relationships constituted a potent metaphor when mapped onto other unequal aspects of society. It is in the metaphorical extension of the terminology and imagery of slavery and freedom that we find the greatest divergence between Greece and the Near East. In Greece we progress from an early period in which the terminology of slavery and freedom is restricted to personal status, through a gradual semantic mapping-out of that vocabulary to politics (in the sixth century) and to social and dispositional domination (in the fifth century). In Greece, metaphors of slavery are usually (but not invariably) negative. In the Near East, however, metaphors of slavery tend to emphasize loyalty and deference, and are usually (but not invariably) positive. 4. Greeks and other ancient peoples were not bamboozled as a result of these processes of semantic extension; for the terminology of slavery and freedom was polysemous and its meanings calibrated by context. A glance at legal distinctions made in Athenian and Babylonian courtrooms shows a practical approach to this issue. The blurring of conceptual boundaries between the legal understanding of slavery and its terminology’s manifold other uses is a product of the minds of modern scholars, not those of their ancient subjects.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

292

Greek Slave Systems

5. Although Finley’s concepts of ‘slave society’ and ‘society with slaves’ represented an important intellectual advance, several caveats must be observed regarding their use. First, they inevitably create a binary, blackand-white picture of the role of slavery in ancient economies, whereas a sliding scale of intensification of the sort proposed by Lenski is better suited to capturing the full diversity of real-life circumstances. Second, this black-and-white bifurcation is compounded by the enormous unevenness in our knowledge of ancient societies; for no society poorly attested in the sources can qualify as a slave society, even if it was one in fact. Everything that we can see of ancient slave systems is filtered through the genres of the available evidence, and the light refracted through these generic prisms varies in its intensity from one society to the next, given the uneven quantities of evidence that survive. For some slave systems (e.g. Athens, Rome) we have much evidence from a wide variety of genres; for others (e.g. Assyria, Babylonia) we have much evidence from a narrow range of genres. Yet others (e.g. Israel, Carthage) are attested through a limited amount of evidence from a limited range of genres. By far the bulk of the picture of ancient Eastern Mediterranean slavery is simply invisible to us due to a lack of sources. These facts are best kept at the forefront of our minds and should cause one to doubt the validity of sweeping generalizations about ‘Greek’ or ‘Near Eastern’ slavery. 6. Regarding the early history of Greek slavery and its development in the longue durée of c.700–400 BC, the Finleyan understanding of the problem must be completely revised. Slavery was of key importance to elites c.700 BC; what followed was not a ‘rise’ of Greek slavery or an ‘invention’ of slave society, but the expansion and adaption of slavery across a variety of regions. In Attica, the region we know best, we find major changes in (a) the sources of slaves and (b) the uses to which they were put. Although non-Greek slaves are evident in archaic texts, they clearly became far more important to the slave supply of Attica (and other regions) over time. Demographic expansion in the free population was probably tracked by demographic expansion in the slave population; and growing diversity in the economy led to an expansion of the roles that slaves performed. The treatment of slaves in customary law was increasingly supplemented and codified in written enactments, allowing ever more complex bodies of legislation concerning slavery to accumulate. 7. Whilst slavery in Attica displays features that were probably quite typical of many of the more commercialized Greek poleis, it should not stand as the archetype of classical Greek slavery; indeed, we are best to avoid archetypes altogether. Classical Greek slavery was diverse, a point that follows if we recognize that Spartan helotage and systems of forced

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Differentials in Slavery’s Magnitude

293

labour in other regions (e.g. Crete, Thessaly, Heraclea Pontica, Syracuse) were not forms of ‘serfdom’ or some vague species of Unfreiheit. They were systems of private slavery; but systems adapted to the broader social, cultural, and institutional architecture of the regions in which they were embedded. The old ‘serfdom’ view is largely based on a rejection of contemporary evidence in favour of late sources, as well as an improbable piece of Quellenforschung concerning the authorship of the metaxu slogan in Pollux 3.83, repeated largely unquestioned over the last fifty years until it calcified into a quasi-fact. 8. Stepping back from the Greek world and viewing Greek slave systems in a wider Eastern Mediterranean context, we can see a variety of similarities and differences. In terms of the role of slavery in generating elite wealth, it is simply false to posit a clear fault line between the Greeks and their neighbours. In some parts of the Eastern Mediterranean we can find local elites importing slaves and relying on their labour to a high degree. In others, slavery did not develop to the same extent, and elites leant more heavily on other forms of labour such as tenancy and wage labour. In most parts of the Eastern Mediterranean, we have simply no knowledge of the details of systems of labour, slave or otherwise. The ‘test pit’ approach pursued in this book has sampled those regions where—at different times—it is possible to address the problem of how innovative or unusual Greek slaving practices were. The resulting picture shows that the contrast is far less pronounced than was previously thought. 9. Explaining regional diversity of this sort requires a broad conceptual toolbox. In economic terms, neither a ‘formalist’ approach (emphasizing individual rational choice) nor a ‘substantivist’ approach (emphasizing culture-specific codes of behaviour) is sufficient on its own. A hybrid methodology is required, one sensitive both to the idiosyncratic cultural norms and institutional structures of specific ancient societies and to generalized laws of demand, supply, market mechanisms, transaction costs, and so forth (and, furthermore, one that does not deny the existence of the profit motive in ancient economies, however much that might have been constrained or mitigated in practice by social obligations and ethical strictures). What is more, politics and geography have not hitherto been fully recognized as key forces shaping the slave systems of the ancient world. State formation and the choice to protect citizens from enslavement had economic consequences. So did the happenstance variable of proximity to this or that slaving zone and its relative demographic density. There is, then, no monocausal explanation for regional diversity in slaveholding practices, but a shifting set of variables, all of which need to be taken into account.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

294

Greek Slave Systems

10. Apart from the minority in the ancient world that held a moral aversion to slavery, individuals in all the societies we have studied aimed to acquire slaves. Slavery was not an undifferentiated ‘mode of production’ constituted by chain gangs for males, domestic drudgery for females, spurred along by the whip: slavery was a legal institution, and sophisticated systems of incentives and sanctions allowed its pragmatic extension to almost every conceivable occupation. Slavery enabled the slaveholder to extract labour from an individual for his or her entire lifetime without payment. The reason why we find it developed to greater or lesser degrees in this or that area is a matter of practical constraints: does it pay to buy slaves more than to hire workers or otherwise compel other kinds of labour? What social costs and benefits might the available labour alternatives involve? Here, a nexus of variables shaped the decisions that individuals made about labour; and in turn shaped a mosaic of slave systems across the Eastern Mediterranean world and the lives of the slaves who laboured in them.

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

APPENDIX

The Meaning of oiketes in Classical Greek In his remarkable recent book Slavery in the Late Roman World, AD 275–425, Kyle Harper has discussed the semantic range of the term οἰκέτης in Late Antique Greek and, through an examination of all attestations of the term collected in a TLG search of Greek literary sources from the fourth–sixth centuries AD, has cogently argued that the word is a concrete status term for slaves in general: it is not used for free servants, nor is it limited to ‘domestic’ or ‘household’ slaves, and it does not appear to be used metaphorically, unlike that other major slave word, δοῦλος, which frequently is (cf. Chapter 2 of this volume).¹ Here, I provide a parallel study of the term in classical Greek. The resolution of two important historical debates hinges on the correct translation of this term. First, E. M. Wood has claimed that oiketes means ‘servant’. Whilst slaves might be included in this category, the term was not, according to Wood, limited to slaves sensu stricto. This ‘fact’ allowed her to dismiss from the debate over agricultural slavery in Attica all passages that employ this term.² Second, in his discussion of the Athenian slave population, Raymond Descat has recently argued that the oiketai mentioned in Athen. Deip. 6.93.272c–d should be translated as ‘household members’, a suggestion that has been followed in a recent study by Hans van Wees, with the implication that the slave population in late fourth-century Athens was perhaps two or three times larger than previous estimates had supposed.³ Since the correct translation of this term has real consequences for our understanding of these problems, I hope that a new and thorough examination of the problem will prove useful. One could go about this task in two ways: by looking at etymology or by looking at usage. It is obvious that, in terms of etymology, the word οἰκέτης derives from the same root as the words for ‘house’, ‘household’, or ‘estate’ (οἶκος, οἰκία, etc.).⁴ It is, therefore, no wonder that many have supposed the term to mean ‘household slave’ or ‘domestic slave’.⁵ A preliminary objection may be lodged against this translation: even if (as seems certain) the term was linked with the household in its original sense, it does ¹ Harper (2011a): 513–18. ² See Chapter 8, pp. 181–3 above. ³ Descat (2004): 368–70; (2011): 209; van Wees (2011): 106; cf. Wiedemann (1987): 13; Andreau and Descat (2006): 34–5. See also Dmitriev (2016), who applies a similar understanding of the word to the debate over the presence of slaves in the Athenian hubris law, believing it protected slaves insofar as they were members of a household. ⁴ Frisk (1970): 360–1 s.v. οἶκος; Chantraine (1974): 781–2 s.v. οἶκος; Beekes (2013): 1055–6, s.v. οἶκος. ⁵ LSJ⁹ s.v. οἰκέτης gives two definitions: ‘household slave’ and, in the plural, ‘household’. Montanari (2014) s.v. οἰκέτης provides similar definitions: ‘domestic, member of a family’ and ‘slave, servant’. Frisk (1970): 360 defines the term thus: ‘Hausgenosse, Diener, Haussklave’. Likewise, Chantraine (1974): 781 writes ‘le mot répondrait au français domestique, désigne les gens, les serviteurs de la maison, équivaut dans l’ensemble des emplois à δοῦλος, mais peut parfois s’en distinguer, cf. Pl. Lois 763a, 777a, etc.’. On these passages of Plato, see section IV. 4 below.

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

296

Appendix

not follow that the same meaning remained unaltered over time.⁶ McMahon highlights the standard problem with the etymological fallacy: for instance, hierarchy was first used in English for the medieval classification of angels into various ranks . . . In the seventeenth century, hierarchy was extended to the ranking of clergymen, and thereafter to any system of grading. Hughes . . . quotes a letter to the Daily Telegraph from 1976 which clearly harks back to the earlier meanings, complaining that ‘in your issue of March 3 you refer to the “Soviet Communist Hierarchy”. I eagerly await your publication of a photograph of Mr. Brezhnev wearing a cope and chasuble—or even wings.’⁷ One could point to the word ‘holiday’ in the same respect: despite the word’s obvious original import, the religious element has essentially disappeared from its modern usage. To claim that oiketes must mean ‘domestic slave’ or such like is to take a similarly erroneous position as the Telegraph’s correspondent, and fall squarely into the trap of the etymological fallacy, taking the fossilized core of the word as proof of a fossilized meaning at a later point in time.⁸ The meaning of specific words may change over time. In order to grasp the semantic range of a term in a given time period, therefore, we must examine its usage in all possible contexts. To this end, I have examined all attestations of the term in Greek literary texts from the fifth century down to the third century BC. I have found scarcely any difference in the basic application of this term from that mapped out by Harper. Three points are key, after which I will address possible objections: (i) οἰκέτης is not limited to a narrow meaning of ‘domestic’ or ‘household’ slave but, on the other hand, (pace Descat and van Wees) its semantic range is not so broad as to indicate ‘member of the household’ in a general sense; (ii) the word never applies (pace Wood) to free servants; and (iii) its basic difference from δοῦλος lies in the fact that the latter (and its related terms) are frequently used in a metaphorical sense,

⁶ Gschnitzer (1964): 1296 writes: ‘Er hat einen, wenn man so sagen will, familiären Ursprung und bezeichnet den Sklaven vor allem in seinen menschlichen Beziehungen zum Herrn . . . Wie δημότης der Angehörige einer Gemeinde, so ist οἰκέτης der Angehörige eines Hauses.’ His citations of passages from Herodotus as instances where the term was apparently applied to slave and free members of a household are not, however, convincing: see section IV. 2 below. Regarding the original meaning of the word, the usage of the term οἰκεύς in Homer is suggestive: here, the word is used as a general term for ‘slave’ and is also used of slaves living in the countryside, such as Eumaeus (Od. 14.4, 63). A similar usage to the Eumaeus example is evident at Gortyn, where the term ϝοικεύς can be used of male slaves working as herdsmen in the countryside: IC IV 72 IV 34–6. The archaic term οἰκεύς is also used in a Solonian law on damages, again evidently concerning slaves in general: see Leão and Rhodes (2015): 53–4. Odysseus attempts a ruse at Od. 4.245 by flagellating himself so that he looks like a slave (οἰκεύς). I have found no usage in Homer suggesting that individuals so labelled were ever anything other than slaves. Even in Homer’s time, then, the semantic range of οἰκεύς was not (pace Kyrtatas 2007: 1057) limited to ‘domestic slave’, nor would it seem (pace Kästner 1981: 298) to mean ‘servant’ or ‘household member’ in a broad sense. See further Ndoye (2010): 209–13. ⁷ McMahon (1994): 175. ⁸ Cf. the remarks of Denniston (1934): v–vi on etymology; on the term oiketes, Jameson (1992): 142, remarks against Wood’s etymological arguments that the term’s semantic range shifted so that it ‘it had become a general term for a servile worker. In the history of language this is, of course, a perfectly familiar phenomenon. We no longer expect all “maids” to be virgins, nor all “nurses” to give milk.’

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Appendix

297

whereas οἰκέτης almost always has a concrete meaning and is applied to those who are legally enslaved.

I. οἰκέτης Does Not Mean ‘Domestic Slave’ A number of passages make it clear that oiketes is not restricted to the narrow meaning of ‘domestic slave’. For example, when Thucydides (8.40.2) writes about the slave population of Chios, he says that the Chians have more oiketai than any other polis except Sparta. It is clear from the context that he is talking about slaves in general and not claiming that the Chians and Spartans had exceptionally large numbers of domestic slaves.⁹ Likewise, when Plato writes of the dangers of having oiketai who all speak the same language, he is evidently concerned about revolts of slaves in general, not simply domestic mutiny (Pl. Leg. 777c3). Several authors refer to Athenian laws concerning oiketai; it would be improbable to suppose that the Athenians had different laws for domestic slaves from those governing slaves employed elsewhere. For example, Hypereides (3.22), when referring to sales of oiketai, notes a law holding that the responsibility for a slave’s misdeeds lies with the man who owned him at the time the misdeed occurred. This hardly makes sense in a domestic setting and must refer to slaves in general. Aeschines (3.44) refers to a law that banned the manumission of oiketai in the theatre; this can hardly apply to domestic slaves alone. The Aristotelian Athenaion Polteia (57.3) notes that trials involving the murder of oiketai are to take place in the Palladium. This must have applied to slaves in general, and in Isoc. 18.52 we find such a case where the victim is referred to not as an oiketis, but as a therapaina. And Lycurgus (fr. 11.1) refers to andrapodistai (enslavers) as persons who deprive others of their oiketai, but it seems unlikely that he is referring only to the theft of domestics. In all these cases, it is obvious that these authors are using the word as a catch-all term for slaves. Likewise, several passages show oiketai working outside the domestic environment. Aristophanes (Pax 1249) refers to weighing figs to give to one’s oiketai who work in the fields. Lysias (7.19) gives a more specific example of oiketai working in the countryside, in this case removing olive trees. Aeschines (1.97) refers to a group of shoemaking oiketai who worked under the supervision of a superintendent and paid Timarchus an apophora of two obols per day each; and in the speech Against Pantaenetus, we hear of oiketai working in a silver mine extracting ore, about as far from a domestic setting as one might imagine (Dem. 37.28–9). When Aeschines (1.54) refers to the demosios Pittalacus as an oiketes, he obviously does not mean a domestic, for Pittalacus can hardly have been much use to the polis had he remained at home. Evidently, by the classical period the term oiketes (whatever its original import) was not limited to slaves working in the domestic sphere alone, but applied to all slaves in ⁹ Stephen Hodkinson points out to me that Thucydides here underscores the Chian slaves’ knowledge of the countryside (ἐπιστάμενοι τὴν χώραν), which must mean that these oiketai were largely involved in agriculture, not domestic service. In a general study of the term oiketes in Thucydides, K. M. Kolobova (n.d.) comes to a similar conclusion, pointing out similar usages with regard to slaves at Plataea and Corcyra. (I quote from p. 18 of a German translation of the Russian study, held by the Mainz Akademie der Wissenschaften und der Literatur; my thanks to Jason Porter for a copy.): ‘So zeugt das Studium des Termimus oiketai bei Thukydides zweifellos davon, dass man den Begriff dieser Kategorie von Sklaven nicht durch die Vorstellung von Haussklaven einengen darf, den Sklaven als Dienern.’

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

298

Appendix

general. And the fact that the archaic term oikeus appears to have had a similar semantic range means that this broader meaning was not a recent development.

II. οἰκέτης Does Not Mean ‘Servant’ Despite the popularity of rendering this term with the legally ambiguous word ‘servant’ in modern English translations of classical Greek texts (examples of which are too numerous to render citation worthwhile¹⁰), I have found no example where the term is used for a person in service of any kind who is demonstrably of non-slave status. (For the argument that the term can be used of household members more generally, see section IV, ‘Possible Objections’, below.) On the contrary, the word is regularly used as a synonym for other slave terms, often simply to vary diction: δοῦλος (Aeschin. 1.17; Ant. 1.30; Dem. 9.3; 17.3; 24.124; 45.35; 45.76; 45.86; Isoc. 1.21; 4.150; 12.97; 17.49; Thuc. 3.73.1; Pl. Leg. 720c7; 776d3; 777a4, 6; 777d4; 777e6; 794b7), ἀνδράποδον (Hyp. 3.15; Pl. Resp. 578d9; 578e4, 7), παῖς (Dem. 37.40), θεράπων (Is. 6.16; Lyc. Leocr. 34; Hdt. 2.113), and ἀργυρώνητος (Dem. 17.3; Isoc. 4.123). Several writers use οἰκέται (masculine plural) alongside θεράπαιναι (feminine plural) to mean ‘male and female slaves’ (Aeschin. 1.99; [Dem.] 59.42; Is. 8.9–10; Lyc. Leocr. 29, 30, 32; Timaeus FGrHist 566 F11).¹¹ In Plato’s Euthyphro, we find a free servant (πελάτης) distinguished from a slave (οἰκέτης: Pl. Euthyphr. 4c5).¹² Likewise, Demosthenes (18.258–9) attacks Aeschines’ lowly origins and his drudge-like duties as a youth helping his father sweep the schoolroom, saying that this is the task of an oiketes, not a freeborn boy. At 49.52, Demosthenes specifically distinguishes between slaves (oiketai) and free hired servants (misthotoi). Lycurgus (Against Leocrates 65) refers to Athens’ early laws, where no distinction in penalty was allegedly made between that for murdering a slave (oiketes) and a free man (eleutheros). If the former term had ever been used for servants more generally (i.e. both free and slave), Lycurgus surely would have picked a more precise term. Whereas I have found no example where oiketes is demonstrably used for a free person, there are plenty of passages that diagnostically indicate slave status. The laws concerning oiketai mentioned at Hyp. 3.22, [Arist.] Ath.Pol. 57.3, Aeschin. 3.41, and Lycurg. fr.11.1 all clearly deal with slaves (see section I, p. 297 above). In a number of passages, the term oiketes is used for slaves potentially subject to judicial torture for evidence (e.g. Dem. 29.38; 30.27; 37.41; 52.22; Is. 8.10; Lycurg. 1.29–32; Aeschin. 2.127; Lys. 4.16); and Demosthenes says that oiketai are legally capable of being tortured for evidence, but hired servants (misthotoi) are not ([Dem.] 49.52). In Plato’s will (D.L. 3.42) we find oiketai bequeathed by the philosopher, evidently slaves, for how else could he bequeath them were they not his property?¹³ Likewise, in a fragment of Lysias (fr. 1 [Thalheim]) we find an oiketes pledged as security for a loan; and other passages mention oiketai being sold ([Dem.] 53.6; Hyp. 3.15; Xen. Mem. 2.5.2) and manumitted (Dem. 24.124; Isoc. 6.88). In several passages, oiketai are specifically contrasted with their masters or with free men. Ctesias (fr. 4 ap. Athen. ¹⁰ Though note that it is also used in basic textbooks for teaching Greek: see, e.g., Balme, Lawall, and Morwood (2016): 53, translating Xen. Oec. 7.35. ¹¹ On terminology for female slaves, see Faraguna (1999): 58–9. ¹² For an argument that the πελάτης was a debt bondsman, see Harris (2006): 261–2. ¹³ On the authenticity of this document, see Canevaro and Lewis (2014): 103–6.

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Appendix

299

Deip. 14.44.639c) writes of a festival where slaves (oiketai) and their masters (despotai) reverse their usual roles; Aeneas Tacticus (40.3) writes of oiketai belonging to despotai;¹⁴ Hypereides (3.16) specifically contrasts the legal position of oiketai and eleutheroi; and Aristotle (De generatione animalium 744b19–22) writes that in running a household, good food is given to the free, leftovers and inferior food to the slaves, and the slops to animals (ἐν δὲ ταῖς οἰκονομίαις τῆς γινομένης τροφῆς ἡ μὲν βελτίστη τέτακται τοῖς ἐλευθέροις, ἡ δὲ χείρων καὶ τὸ περίττωμα ταύτης οἰκέταις, τὰ δὲ χείριστα καὶ τοῖς συντρεφομένοις διδόασι ζῴοις). Translators should, therefore, abandon the convention of euphemistically translating oiketes as ‘servant’. More than any other word, this word denotes ‘slave’ as a technical status term.

III. οἰκέτης is Rarely Used Figuratively, but Sometimes it is Deployed in Vituperative Speech The term doulos has a broader semantic range than oiketes, as have some other slave terms.¹⁵ (Harper has pointed out the same pattern in late antique Greek.¹⁶) One nontechnical usage of slave terminology that is worth considering can be found in forensic oratory, which provides us with several examples of individuals being called ‘slaves’, even though they were evidently legally free persons. As Kamen has noted: in a speech of Isaios, the speaker calls an associate of his opponents, the hetaira Alkê, a doulê, although it is suggested elsewhere in the speech that she is a freedwoman (Is. 6.49). In another case, Demosthenes twice refers to the freedman Lykidas, the former slave of Khabrias, as a doulos (Dem. 20.131–3), even though Lykidas is not only freed but also a proxenos, a fairly lofty position.¹⁷ Although some scholars have seized upon these passages as evidence for ‘privileged slaves’,¹⁸ it is rather the case that these instances are symptomatic of rhetorical mud-slinging. This could involve one of two things: either drawing attention to the servile origins of actual freedpersons who had risen up the social ladder, or casting groundless aspersions on the parentage of free Athenians.¹⁹ Of the former variety, two well¹⁴ On the meaning of δεσπότης in Attic Greek of the classical period, see Harris (2006): 278–9. ¹⁵ On figurative usage of δοῦλος terms, see Brock (2007): 209–10. A good example is Isoc. 1.21, which refers to the sort of man who rules over his slaves (οἰκετῶν ἄρχειν), but is a hypocrite since he is enslaved to his desires (ἡδοναῖς δουλεύειν; cf. Isoc. 4.150). Mactoux (1980): 148 writes ‘contrairement à oikétès qui n’était jamais au sens figuré, thérapôn est employé au sens propre comme au sens figuré’. Not quite: I have found the following figurative usages of the term oiketes: Isoc. 7.26 (rich people should act like oiketai of the community; very similar to the rhetoric of being a ‘public servant’ used by modern British politicians); Pl. Theaet. 173c3 (in philosophical discourse, arguments are viewed as the oiketai of their speakers). But Mactoux’s point is well taken: such usages are extremely rare. ¹⁶ Harper (2011a): 513. ¹⁷ Kamen (2009): 48. Cf. [Dem.] 45.75–6 and 83, where Apollodorus calls the freedman Phormion a doulos. ¹⁸ Cohen (2000): 130–54. ¹⁹ On servile invective, see MacDowell (1993) and Kamen (2009). Kamen (2009): 47–8 explains instances where Athenian politicians are accused of servile ancestry in forensic oratory as a comic convention that has been transferred into the courtroom. There is perhaps something to this suggestion. However, it seems to me that there is a more basic reason why scope existed in

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

300

Appendix

known examples are illustrative: Pittalacus in Aeschines Against Timarchus (who is referred to as a public slave at 1.54: ἄνθρωπος δημόσιος οἰκέτης τῆς πόλεως) and Lampis in Dem. 34 Against Phormion, a ship’s captain (naukleros) who is called the oiketes of Dion at Dem. 34.5, and apparently escaped a sinking ship into a small boat along with the ‘other slaves of Dion’ (Dem. 34.10: καὶ αὐτὸς μὲν ἀπεσώθη ἐν τῷ λέμβῳ μετὰ τῶν ἄλλων παίδων τῶν Δίωνος). Scholars eager to find evidence for slaves holding legal rights in fourth-century Athens trumpet these references as proof of their case, but it is far more probable that the orators in these instances refer to these freedmen as if they were still slaves to hammer home their lowly origins.²⁰ Two aspects of Athenian law further help to explain this phenomenon. First, Athens’ law on slander (kakegoria) was extremely specific and did not cover the wide range of defamatory comments actionable under modern slander laws.²¹ Second, trials took place on a single day, and litigants were constrained by the water running out of the clepsydra. These two facts meant that all manner of accusations could be hurled inside or outside court without making the perpetrator of these remarks subject to legal action and that, when in court, litigants did not have the time to address vituperative remarks if they meant to honour their oath to stick to discussing the substance of the charge.²²

IV. Possible Objections Hans van Wees, following Raymond Descat, has recently claimed that the word οἰκέται ‘literally means “members of the household” and is not infrequently used in Athens for these sorts of slanderous remarks to be levelled: men (especially elite men) generally kept the doings of their womenfolk strictly behind closed doors, yet used female slaves for sex. This provided fertile soil for speculation and rumour-mongering regarding the parentage of children. For example, see Ar. Thesm. 565–6 for the notion of substituting a slave baby for a stillborn free child, more or less the same practice as the notorious bedpan trick allegedly perpetrated by James II in 1688 (cf. Ar. Thesm. 340, 407; Pl. Resp. 537e–538a; Dem. 21.149). In Theophrastus’ Characters (19.2) the nauseating (δυσχερής) man claims that his scabrous state is an inherited illness and that it is impossible to pass off another’s baby as his own, since the substitute would presumably not display the distinctive family ailments. In the case of an Athenian who was born with ginger hair, there was ample scope for pasquinade: Cratinus, for example, made fun of Hipponicus by calling him ‘Scythian’ on account of his red hair (Hesychius s.v. Σκυθικός); and one could always allude to the whiff of scandal if a man’s slave were said to have had an affair with his wife, as happened to Euripides: see Ar. fr. 596 K-A; Ran. 1046; 1407–10. As we know from recent events concerning President Obama, even in societies with sophisticated forms of personal documentation, ludicrous ‘birther’ rumours can still attract a lot of attention and debate. ²⁰ See Cohen (2000): 130–54, who argues that scholars with preconceived ideas about slavery ignore the obvious import of these and other passages, which apparently prove that Athenian slaves had all manner of rights against abuse and sexual exploitation. Refreshing as this challenge is, Cohen routinely ignores passages inconvenient to his case that show that slaves had no rights at all (cf. the evidence set out in Chapter 1 of this book, pp. 39–48 above); and he adduces other passages (such as those concerning Pittalacus and Lampis) as proof of his case, despite the fact that they can be interpreted in more than one way. On the status of Pittalacus and Lampis, see the sensible comments of Todd (1993): 192–4 and Fisher (2001): 190–1. ²¹ It included the terms murderer, father- or mother-beater, and the suggestion that someone had thrown away their shield: see Lysias 10.9 with Harris (2013a): 171. ²² For the importance of keeping to the point in Attic oratory, see Rhodes (2004). Rhodes’ approach is fine-tuned and his conclusions strengthened in Harris (2013a): 101–37.

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Appendix

301

the broadest sense to include wives, children and free servants as well’.²³ Based on this view, he interprets the claim of Ctesicles (FGrHist 245 F1 ap. Ath. Deip. 6.272c–d) that the Athenians during the rule of Demetrius of Phaleron had 400,000 oiketai—which scholars from David Hume to Westermann and beyond have rejected as an impossible figure for slaves—to mean 400,000 ‘household members’, i.e. women, children, and slaves. On this view, the slave population would still amount to c.323,000 slaves.²⁴ Alain Bresson has recently shown the impossibility of Descat’s interpretation of Ctesicles’ figures on demographic grounds.²⁵ I will concentrate therefore on van Wees’ linguistic arguments. Van Wees (2011: 106) cites the following passages as supporting his translation: Athen. Deip. 6.93.267e; Hdt. 8.4; 8.41; 8.142; Xen. Cyr. 4.2.2; 4.3.1–2; 4.2.37; Pl. Leg. 763a; 777a; 853e. Let us deal with each author in turn.

1. Athen. Deip. 6.93.267e According to van Wees, ‘Athenaeus reports as “common usage” that “one who lives in a household is an oiketes even if he is free”.’²⁶ Indeed he does: ὅτι δὲ οἰκέτης ἐστὶν ὁ κατὰ τὴν οἰκίαν διατρίβων κἂν ἐλεύθερος ᾖ κοινόν. One must examine this claim, however, in the context of the whole discussion of 6.93.267b–e. Athenaeus begins by having his dinner guest Democritus cite the Hellenistic philosopher Chrysippus of Athens on the difference between oiketai and douloi: διαφέρειν δέ φησι Χρύσιππος δοῦλον οἰκέτου γράφων ἐν δευτέρῳ Περὶ Ὁμονοίας διὰ τὸ τοὺς ἀπελευθέρους μὲν δούλους ἔτι εἶναι, οἰκέτας δὲ τοὺς μὴ τῆς κτήσεως ἀφειμένους. ὁ γὰρ οἰκέτης, φησί, δοῦλός ἐστι κτήσει κατατεταγμένος. This we may translate as: ‘but Chrysippus says in the second book of On Concord that that there is a difference between doulos and oiketes, since freedmen are still douloi, whereas oiketai are those who are not let loose from ownership. For the oiketes, he says, is a doulos held in ownership.’ This statement, though initially perplexing, is in fact a rather good description of the respective semantic ranges of doulos and oiketes in classical Attic: the latter term is used for slaves in a concrete, property sense, whereas the former (and related terms) can be used far more broadly for those dominated in some way or other and for those in some manner tainted by servile ancestry or working in a ‘slavish’ occupation.²⁷ The contrast that Chrysippus is making, then, is between a word that means ‘slave’ in a narrow technical sense concerned with legal ownership and one that means ‘slave’ in a much broader sense and can therefore apply to freedmen, who are tainted socially by their servile background. Athenaeus then has his speaker Democritus digress on a number of rare terms for slaves, at times showboating with etymological speculation, for example claiming that the ἀμφίπολος is a slave girl who is ‘about’ (amphi) her mistress, whereas a πρόπολος is one who walks ‘in front’ [of her mistress] (προπορευομένην). He then (at 6.93.267d) returns to the issue of oiketai by discussing two plays, one (the Laertes) by the fifth-century polymath Ion of Chios, the other (a satyr play, the Omphale) by the fifth-century poet Achaeus of Eretria. In the latter play we find the phrase ‘ὡς εὔδουλος, ὡς εὔοικος ἦν’ (= Achaeus fr. 32 TGF), which one would expect to mean either ‘how ²³ Van Wees (2011): 106. Cf. Kyrtatas (2005): 71, ‘The most common word applied for domestic slaves, «οἰκέτης», was also used for the free members of a household.’ ²⁴ Van Wees (2011): 107. ²⁵ Bresson (2016): 459–60. ²⁶ Van Wees (2011): 106. ²⁷ See Canevaro and Lewis (2014): 100–2.

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

302

Appendix

good to his slaves, and to his household, he was’ (following LSJ ⁹ on the meaning of these words²⁸) or (as seems more likely, following the pattern of εὔπορος, ‘rich’, ‘well provided for’, e.g. Arist. Pol. 1279b8, etc.) ‘how wealthy in slaves, how wealthy in property he was’.²⁹ Athenaeus, however, has his speaker Democritus opt for a rather tortuous translation of εὔοικος: he believes that Achaeus is ‘saying, in a peculiar manner, how kind he was to his douloi and oiketai’ (ἰδίως λέγων ὡς χρηστὸς ἐς τοὺς δούλους ἐστὶ καὶ τοὺς οἰκέτας), and the following line (ὅτι δὲ οἰκέτης ἐστὶν ὁ κατὰ τὴν οἰκίαν διατρίβων κἂν ἐλεύθερος ᾖ κοινόν) makes it clear that he takes oiketai here to mean free servants. From this piece of elucidatory grandstanding we can get a clear sense of the tenuous kind of explanation favoured in this dinner conversation, where the rare and obscure view scores more points than the quotidian and obvious. To reinforce the notion that this strained interpretation is the right one, the speaker claims that it is a commonplace (κοινόν) for free men to be referred to as oiketai. The claim is demonstrably false, for of the hundreds of attestations of this term extant in Greek literature, such a usage never crops up once. As Harper rightly points out, ‘maybe Athenaeus was right that οἰκέτης could have the sense of a house-servant or a free man, but surely it is noteworthy that the only example that he could adduce is a strained interpretation of a line of classical poetry’.³⁰ We might go a little further, though, and question the very exercise of abstracting a comment belonging to one of Athenaeus’ diners and treating it as a straightforward fact. The intellectual sparring between the diners is regularly characterized by dubious argumentation and mutual criticism thereof; and as Swain points out, ‘Ulpianus is also criticized for his preposterous etymological usages, which for example make him achrêstos (‘useless’) in the sense of ‘unused’ (3.97d–e). According to Cynulcus there is a whole tribe of sophists—the

²⁸ Both terms are extremely rare and their meanings uncertain; as we shall see, Athenaeus’ speaker fallaciously argues that the term εὔοικος means ‘good to (free) servants’ in support of his claim that the term oiketes can be used of free servants. LSJ⁹ s.v. seem to follow the explanation of the phrase given by Athenaeus’ speaker. The only other attestation of εὔδουλος is in Pollux 3.80, who cites Pherecrates (fr. 212 Kock = fr. 244 K-A): καὶ εὔδουλος δὲ ὁ τοῖς δούλοις εὖ χρώμενος παρὰ Φερεκράτει, καὶ τὸ δρᾶμα Δουλοδιδάσκαλος, ‘eudoulos—he who deals well with his slaves, as in Pherecrates, and the play The Slave Trainer’. Pollux includes εὔδουλον at 9.162 in a list of words with the prefix εὐ-, but with no explanation of the term. As for εὔοικος, its other attestations amount to (i) an entry in the twelfth-century AD Etymologicum Magnum (389.24 [Gaisford]), a gloss on the word εὐβύριον, mentioned by Euphorion and meaning ‘well furnished with houses’ or ‘with good houses’, a definition also found in the twelfth-century AD Etymologicum Symeonis, both clearly derived from the identical entry in the earlier Etymologium Genuinum of the ninth century AD; (ii) a line of Oppian’s Halieutica (3.370), describing a fish trap, which, to the trapped fish, is not εὔοικος and which must, therefore, mean something like ‘homely’ (and certainly has nothing to do with slaves); (iii) Dio Cassius 44.39, who claims that Julius Caesar was εὐοικότατος but at the same time εὐδαπανώτατος in his private affairs, retaining his patrimony as far as possible, but being generous with that which he had acquired; here εὐοικότατος must mean something along the lines of ‘a most prudent manager of his estate’ (again, and given the rest of Dio’s sentence, the word has nothing to do with slaves). ²⁹ Harper (2011a): 517 translates the phrase as ‘rich in slaves, rich in property.’ As I have shown in n. 28 above, it is most unlikely that εὔοικος has anything to do with slaves; the close connection between εὔδουλος and εὔοικος in Achaeus fr. 32 TGF makes it unlikely that we should translate his phrase as ‘kind to slaves’ for the first term, but ‘wealthy in his estate’ for the second term. Harper’s translation seems to me the most elegant and credible solution. ³⁰ Harper (2011a): 517.

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Appendix

303

Oulpianeioi—who display the same faults.’³¹ In our passage, the speaker is not Ulpian but Democritus; but the faults of Ulpian are to a greater or lesser extent shared by all the deipnosophistai. We should be very wary, therefore, of taking such unsubstantiated claims at face value.

2. Hdt. 8.4; 8.41; 8.142 Van Wees writes that: Herodotus repeatedly speaks of Greeks evacuating ‘their children and their oiketai’ as the Persian army approaches (8.4, 41), where, of course, the wives are included under oiketai. Conversely, the Spartans at one point promise that for the duration of the war they will look after the Athenians’ ‘women and oiketai who are of no use in war’ (8.142), where surely young children and elderly parents are foremost among the oiketai.³² Examined in context, however, these passages far from prove van Wees’ point. In Hdt. 8.4 the Euboeans, seeing the approach of the Persian fleet, entreat the naval commander Eurybiades to delay departing so that they can remove their children and oiketai (τέκνα τε καὶ τοὺς οἰκέτας). To be sure, it is odd that Herodotus fails to mention the wives of the Euboeans; but if van Wees is right that oiketai means ‘household members’, then it is equally odd that Herodotus does mention their children, who, on his view, surely rank among the oiketai and need not be mentioned separately. The argument cuts in both directions, then; but the passage proves nothing: we need not assume that Herodotus took special care to list every last category of persons to be evacuated, and it may well mean nothing more startling than ‘children and slaves’. (Hdt. 8.41 presents an effectively identical usage.) Van Wees’ argument appears arbitrary given his next citation, where Herodotus mentions women and oiketai: so this time, the term covers slaves, servants, and children, but not women, who are mentioned separately. Either Herodotus is expressing himself pleonastically on all three occasions or he simply means ‘women and slaves’ and is again not providing an exhaustive list of dependants.³³ Although these expressions could potentially fit van Wees’ translation (if we assume pleonasm in all three cases), they also could be translated and explained simply as referring to slaves. Since the expressions are too vague to rule out either version for sure, they hardly count as proof of a special meaning for oiketai of ‘household members’.³⁴

3. Xen. Cyr. 4.2.2; 4.3.1–2; 4.2.37 Van Wees writes that: Xenophon gives something close to a definition of the term when he claims that in Asia soldiers tend to bring their oiketai on campaign since it is their custom to ‘go to war alongside those with whom they live together’ (Cyrop. 4.2.2;

³¹ Swain (1996): 50. ³² Van Wees (2011): 106. ³³ Cf. Thuc. 2.4.2 and 5.82.6, where free women are distinguished from oiketai. ³⁴ LSJ⁹ s.v. οἰκέτης 2 also cites Hdt. 8.106, which again could support van Wees’ view, but could also be interpreted in terms of slaves and proves nothing.

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

304

Appendix

cf. 4.3.1–2). He even uses the term for the household excluding slaves and servants, when he refers to servants preparing food ‘for their masters and their oiketai’ (4.2.37).³⁵ These passages are the weakest of all of van Wees’ examples. Masters usually live together with their slaves, and there is no mystery in them bringing their slaves on campaign, both to keep an eye on them and to be served by them; and in this case Xenophon is writing about the nomadic Hyrcanians, who had no fixed homes in which to leave their slaves behind: of course they brought them along! Xenophon at Cyr. 4.3.1–2 does not mention oiketai, but merely repeats his earlier comment on the alleged tendency of eastern peoples to take all their property with them to war. As for Cyr. 4.2.37, this passage simply refers to the preparation of food by some servants or slaves for their ‘masters and his slaves’ (τοῖς δεσπόταις καὶ τοῖς οἰκέταις). Not all the slaves present here are involved in the cooking; but both masters and slaves alike are, naturally, consumers of rations—slaves do not subsist on thin air. There is therefore an entirely banal and commonsense way in which we can interpret this passage, and we do not need to resort to the unusual interpretation given it by LSJ⁹ and followed by van Wees.

4. Pl. Leg. 763a; 777a; 853e These three passages are cited by LSJ⁹ (s.v. οἰκέτης 2) as instances where oiketes is supposedly opposed to doulos.³⁶ As Harper rightly points out, they ‘are all unconvincing’, since ‘the first is a pleonasm, the second a synonym, the third an attempt to vary the diction’.³⁷ * * * It should be stressed that against this small handful of citations, whose vagueness provides room for argument that they might mean ‘household members’, but none of which is probative, one must weigh several hundred attestations of the term in classical Greek literature that fit the normal pattern described above.³⁸ Given these facts, there are strong grounds for contesting the arguments of Wood, Descat, and van Wees, which are based upon a small number of tendentious passages. Wood’s argument about slavery in agriculture—flawed in other respects (cf. Chapter 8 pp. 183–93 above)—can also, therefore, not rely on her translation of oiketes as ‘servant’; and the arguments of Descat and van Wees about the Attic slave population in the time of Demetrius of Phaleron likewise rest on weak linguistic grounds, not to mention the pragmatic demographic problems with their view. However, to give these scholars their fair due, the error is not one of their own

³⁵ Van Wees (2011): 106. ³⁶ And by van Wees (2011): 106. ³⁷ Harper (2011a): 517. ³⁸ A particularly good example of the proper usage of oiketes is Pl. Resp. 578d–579a, which is unintelligible if van Wees is correct in supposing that the term can mean ‘members of a household’. Here, a free man is whisked off ἐκ τῆς πόλεως αὐτόν τε καὶ γυναῖκα καὶ παῖδας θείη εἰς ἐρημίαν μετὰ τῆς ἄλλης οὐσίας τε καὶ τῶν οἰκετῶν. Note how the oiketai are grouped with the property, not with the family members. Plato also uses several slave terms later in the passage to vary the diction, calling the slaves not only oiketai, but also andrapoda, therapontes, and douloi, yet always referring to the same group of individuals. Cf. Ar. Nub. 5–7 with Kästner (1981): 299. Here, when Strepsiades mentions oiketai, he obviously means his slaves, not his household members in general.

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Appendix

305

invention: the kernel of the error can be found in Liddell and Scott’s lexicon; and to be fair to Liddell and Scott, an error entirely understandable given the magnitude of their task.³⁹

³⁹ I have not been able yet to check a copy of the first edition of their lexicon (1843); the third edition of 1849, however, has the following entry: ‘strictly, an inmate of one’s house; but most usu. a house-slave, menial’. The entry also cites Herodotus, saying that in his work ‘οἰκέται also for one’s family, women and children’, as well as citing Pl. Leg. 763a and 853e as cases where the term oiketes is apparently opposed to doulos. However, the error would seem to go back at least as far as Franz Passow’s Handwörterbuch der griechischen Sprache (1819), the lexicon on which Liddell and Scott based their own work. Passow s.v. οἰκέτης has ‘Hausgenosse, aber den herrschenden Sprachgebrauch nach Haussklave, Knecht, Diener.’ Passow also cites Herodotus as (allegedly) using the plural to mean wives and children, though he does not mention Plato.

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography Abraham, Kathleen. 2004. Business and Politics under the Persian Empire. The Financial Dealings of Marduk-nāṣir-apli of the House of Egibi (521–487 B.C.E.) (Bethesda, MD). Acton, Peter. 2014. Poiesis: Manufacturing in Classical Athens (Oxford and New York). Acton, Peter. 2015. ‘Industry structure and income opportunities for households in classical Athens’. In The Ancient Greek Economy: Markets, Households and City-States, edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 149–65). Adams, Colin. 2007. Land Transport in Roman Egypt (Oxford). Ahlström, Gösta W. 1992. The History of Ancient Palestine (Sheffield). Ahmad, Ali and Postgate, Nicholas. 2007. Archives from the Domestic Wing of the North-West Palace at Kalhu/Nimrud (London). Akrigg, Benjamin and Tordoff, Robert (eds.). 2013. Slaves and Slavery in Ancient Greek Comic Drama (Cambridge and New York). Alexianu, Marius. 2008. ‘Une Catégorie d’esclaves thraces: les halônetoi’. In La Fin du statut servile? Affranchissement, libération, abolition. XXXe colloque du GIREA, Besançon 2005, edited by Antonio Gonzales (Besançon: 487–92). Alexianu, Marius. 2011. ‘Lexicographers, paroemiographers and slaves-for-salt barter in ancient Thrace’, Phoenix vol. 65.3/4: 389–94. Allain, Jean. 2008. The Slavery Conventions. The Travaux Préparatoires of the 1926 League of Nations Convention and the 1956 United Nations Convention (Leiden). Allain, Jean. 2015. ‘Property in persons: Prohibiting contemporary slavery as human rights’. In Property Rights and Human Rights in a Global Context, edited by Ting Xu and Jean Allain (Oxford: 93–119). Allain, Jean and Bales, Kevin. 2012. ‘Slavery and its definition’, Global Dialogue vol. 14: 1–15. Allain, Jean and Hickey, Robin. 2012. ‘Property and the definition of slavery’, International and Comparative Law Quarterly vol. 61.4: 915–38. Ameling, Walter. 1998. ‘Landwirtschaft und Sklaverei im klassischen Attika’, Historische Zeitschrift vol. 266.2: 281–315. Ameling, Walter. 2012. ‘Karthago’. In Handwörterbuch der antiken Sklaverei. CD-ROM, edited by Heinz Heinen (Akademie der Wissenschaften und der Literatur, Mainz). Amigues, Suzanne. 2007. ‘L’Exploitation du monde végétal en Grèce classique et hellénistique. Essai de synthèse’, Topoi vol. 15.1: 75–125. Anderson, Perry. 1974. Lineages of the Absolutist State (London). Ando, Hiromu. 1988. ‘A study of servile peasantry of ancient Greece: Centring around hectemoroi of Athens’. In Forms of Control and Subordination in Antiquity, edited by Toru Yuge and Masaoki Doi (Leiden: 323–30). Andreau, Jean and Descat, Raymond. 2006. Esclave en Grèce et à Rome (Paris).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

308

Bibliography

Aperghis, Gerassimos. 2000. ‘War captives and economic exploitation: Evidence from the Persepolis Fortification Tablets’. In La Guerre dans les économies antiques, edited by Jean Andreau, Pierre Briant, and Raymond Descat (Saint-Bertrand-deComminges: 127–44). Archibald, Zosia. 2015. ‘Social life of Thrace’. In A Companion to Ancient Thrace, edited by Julia Valeva, Emil Nankov, and Denver Graninger (Chichester: 385–98). Arciniega, Alberto Prieto. 1989. ‘Aproximación a las formas de dependencia en los territorios de Baetulo e Iluro’ In Esclavos y semilibres en la antigüedad clásica. Coloquio No. 15 del G.I.R.E.A. (Madrid: 179–85). Astour, Michael C. 2000. ‘Overland trade routes in ancient Western Asia’. In Civilisations of the Ancient Near East, vol. 3, edited by Jack M. Sasson (New York: 1401–20). Avigad, Nahman. 1953. ‘The epitaph of a royal steward from Siloam village’, Israel Exploration Journal vol. 3: 137–52. Azoulay, Vincent. 2006. ‘L’Archidamos d’Isocrate: une politique de l’espace et du temps’, Revue des Études Grecques vol. 119: 504–31. Bäbler, Balbina. 1998. Fleißige Thrakerinnen und wehrhafte Skythen. Nichtgriechen im klassischen Athen und ihre archäologische Hinterlassenschaft (Stuttgart and Leipzig). Badian, Ernst. 1981. ‘The bitter history of slave history’, New York Review of Books (22 October 1981: 49–53). Baker, Heather D. 2001. ‘Degrees of freedom: slavery in mid-first millenium BC Babylonia’, World Archaeology vol. 33.1: 18–26. Baker, Heather D. 2004. The Archive of the Nappāḫu Family (Vienna). Baker, Heather D. 2009. Review of Galil 2007, Journal of the American Oriental Society vol. 129.2: 343–4. Baker, Heather D. 2014. ‘House size and household structure: quantitative data in the study of Babylonian urban living conditions’. In Documentary Sources in Ancient Near Eastern and Greco-Roman History: Methodology and Practice, edited by Heather D. Baker and Michael Jursa (Oxford: 7–23). Baker, Heather D. 2016a. ‘Slavery and personhood in the Neo-Assyrian Empire’. In On Human Bondage: After Slavery and Social Death, edited by John Bodel and Walter Scheidel (Malden, MA, Oxford, and Chichester: 15–30). Baker, Heather D. 2016b. ‘Urban craftsmen and other specialists, their land holdings, and the Neo-Assyrian state’. In Dynamics of Production in the Ancient Near East, edited by Juan Carlos Moreno García (Oxford and Philadelphia, PA: 53–73). Bakewell, Geoff. 2008. ‘Trierarchs’ records and the Athenian naval catalogue IG i³ 1032’. In Orality, Literacy and Memory in the Ancient Greek and Roman World, edited by in E. Anne MacKay (Leiden: 141–60). Bakir, Abdel Mohseen. 1952. Slavery in Pharaonic Egypt (Cairo). Balme, Maurice, Lawall, Gilbert, and Morwood, James. 2016. Athenaze: An Introduction to Ancient Greek. Book I. Rev. edn (Oxford). Baralis, Alexandre. 2009. ‘The chôra formation of the Greek cities of Aegean Thrace. Towards a chronological approach to the colonization process’. In Meetings of Cultures in the Black Sea Region: Between Conflict and Coexistence, edited by Pia Guldager Bilde and Jane Hyarl Petersen (Aarhus: 101–30).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

309

Baratta, Giulia. 2014. ‘Una statuetta bronzea di schiavo accovacciato da son Catlar (Minorca, Isole Baleari)’, Mitteilungen des Deutsches Archäologischen Instituts (Madrid) vol. 55: 316–30. Barjamovic, Gojko. 2011. A Historical Geography of Anatolia in the Old Assyrian Colony Period (Copenhagen). Bayram, Sebahattin and Çeçen, Salih. 1996. ‘The institutions of slavery in ancient Anatolia in the light of new documents’, Belleten vol. 229: 605–30. Beaulieu, Paul-Alain. 2000. ‘A finger in every pie. The institutional connections of a family of entrepreneurs in Neo-Babylonian Larsa’. In Interdependency of Institutions and Private Entrepreneurs, edited by A. C. V. M. Bongenaar (Istanbul: 43–72). Beaulieu, Paul-Alain. 2005. ‘Eanna’s contribution to the construction of the north palace at Babylon’. In Approaching the Neo-Babylonian Economy: Proceedings of the START Project Symposium Held in Vienna, 1–3 July 2004, edited by Heather D. Baker and Michael Jursa (Münster: 45–73). Beaulieu, Paul-Alain. 2006. ‘Official and Vernacular Languages: The Shifting Sands of Imperial and Cultural Identities in First-Millennium B.C. Mesopotamia’. In Margins of Writing, Origins of Cultures, edited by Seth L. Sanders (Chicago: 191–220). Beaulieu, Paul-Alain. 2007. ‘Late Babylonian intellectual life’. In The Babylonian World, edited by Gwendolyn Leick (New York and London: 473–84). Bechtold, Babette and Docter, Roald. 2010. ‘Transport amphorae from Punic Carthage: An overview’. In Motya and the Phoenician Ceramic Repertoire between the Levant and the West, 9th–6th Century BC, edited by Lorenzo Nigro (Rome: 85–116). Beekes, Robert. 2013. Etymological Dictionary of Greek. Vol. II (Leiden). Bendor, Shunya. 1996. The Social Structure of Ancient Israel (Jerusalem). Benedict, Burton. 1980. ‘Slavery and Indenture in Mauritius and Seychelles’. In Asian and African Systems of Slavery, edited by James L. Watson (Berkeley and Los Angeles: 135–68). Benson, Bruce L. 1989. ‘Enforcement of private property rights in primitive societies: Law without government’, The Journal of Libertarian Studies 9.1: 1–26. Benson, Peter. 2002. ‘Philosophy of property law’. In The Oxford Handbook of Jurisprudence and Philosophy of Law, edited by Jules Coleman and Scott Shapiro (Oxford: 752–814). Beringer, Walter. 1960. ‘Die ursprüngliche Bedeutung von δουλοσύνην/ς ἀνέχεσθαι in Odysee 22.423’, Athenaeum vol. 38: 65–97. Beringer, Walter. 1961. ‘Zu Begriffen für “Sklaven” und “Unfreie” bei Homer’, Historia vol. 10: 259–91. Beringer, Walter. 1982 ‘ “Servile status” in the sources for early Greek history’, Historia vol. 31.1: 13–32. Bernard, Seth. 2016. ‘Debt, land, and labor in the early republican economy’, Phoenix vol. 70.3/4: 317–38. Bielman, Anne. 1999. ‘De la capture à la liberté. Remarques sur le sort et le statut des prisonniers de guerre dans le monde grec classique’. In Guerres et sociétés dans les mondes grecs, edited by Patrice Brun (Paris: 179–99). Bieżuńska-Małowist, Iza. 1974. L’Esclavage dans l’Égypte gréco-romain. Première partie: période ptolemaïque (Wrocław). Bintliff, John. 2012. The Complete Archaeology of Greece (Chichester and Malden, MA).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

310

Bibliography

Birks, Peter. 1985. ‘The Roman law concept of dominium and the idea of absolute ownership’, Acta Juridica: 1–37. Blavatskaja, Tatjana V., Golubcova, Elena S., and Pavlovskaja, Aleksandra I. 1972. Die Sklaverei in hellenistischen Staaten im 3.-1. Jh. v. Chr. (Wiesbaden). Bogaert, Raymond. 1995. Review of Cohen 1992. Gnomon vol. 67: 604–9. Bömer, Franz. 1990. Untersuchungen über die Religion der Sklaven in Griechenland und Rom. Teil III: Die wichtigsten Kulte der griechischen Welt (Stuttgart). Bongenaar, A. C. V. M. 1997. The Neo-Babylonian Ebabbar Temple at Sippar: Its Administration and Its Prosopography (Istanbul). Borowsky, Ron and Masson, Michael E. J. 1996. ‘Semantic ambiguity effects in word identification’, Journal of Experimental Psychology vol. 22.1: 63–85. Bradley, Keith R. 1994. Slavery and Society at Rome (Cambridge). Bradley, Keith R. and Cartledge, Paul A. (eds.). 2011. The Cambridge World History of Slavery. Vol. I: The Ancient Mediterranean World (Cambridge). Braund, David. 2008. ‘Royal Scythians and the slave trade in Herodotus’ Scythia’, Antichthon vol. 42: 1–19. Braund, David. 2011. ‘The slave supply in classical Greece’. In The Cambridge World History of Slavery. Vol. I: The Ancient Mediterranean World, edited by Paul A. Cartledge and Keith R. Bradley (Cambridge: 112–33). Braund, David and Tsetskhladze, Gocha R. 1989. ‘The export of slaves from Colchis’, Classical Quarterly vol. 39.1: 114–25. Bremner, Jan. 2015. ‘Stigmata: From tattoos to saints’ marks’. In Bodies in Transition: Dissolving the Boundaries of Embodied Knowledge, edited by D. Boschung, A. Shapiro, and F. Waschek (Paderborn: 137–51). Bresson, Alain. 2000. La Cité marchande (Bordeaux). Bresson, Alain. 2016. The Making of the Ancient Greek Economy: Institutions, Markets, and Growth in the City-States (Princeton, NJ, and Oxford). Briant, Pierre. 2002. From Cyrus to Alexander. A History of the Persian Empire (Winona Lake, IN). Bridge, Edward J. 2010. ‘Polite language in the Lachish letters’, Vetus Testamentum vol. 60: 518–34. Bridge, Edward J. 2012. ‘Female slave vs female slave: ‫ אמה‬and ‫ שפחה‬in the HB’, Journal of Hebrew Scriptures vol. 12: 1–21. Bright, John. 1981. History of Israel. 3rd edn (Philadelphia, PA). Brixhe, Claude and Bile, Monique. 1999. ‘La Circulation des biens dans les Lois de Gortyne’. In Des dialectes grecs aux lois de Gortyne, edited by Catherine DobiasLalou (Nancy: 75–116). Brock, Roger. 2007. ‘Figurative slavery in Greek thought’. In Peur de l’esclave—Peur de l’esclavage en Méditerranée ancienne/Fear of Slaves—Fear of Enslavement in the Ancient Mediterranean (XXIXe colloque du GIREA, Rethymnon, 4–7 novembre 2004), edited by Anastasia Serghidou (Besançon: 217–24). Brock, Roger. 2013. Greek Political Imagery from Homer to Aristotle (London and New York). Brown, Francis, Driver, Samuel R., and Briggs, Charles A. 1906. The Brown-Driver-Briggs Hebrew and English Lexicon (Peabody, MA). Brulé, Pierre. 1978. La Piraterie crétoise hellénistique (Paris).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

311

Bryce, Trevor. 2012. The World of the Neo-Hittite Kingdoms (Oxford). Bunimovitz, Shlomo and Faust, Avraham. 2010. ‘Re-constructing biblical archaeology: Toward an integration of archaeology and the Bible.)’. In Historical Biblical Archaeology and the Future: The New Pragmatism, edited by Thomas E. Levy (London: 43–54). Burchfiel, K. J. 1994. ‘The myth of “prelaw” in early Greece’. In Symposion 1993. Vorträge zur griechischen und hellenistischen Rechtsgeschichte, edited by Gerhard Thür (Cologne: 79–104). Bury, John B. 1962. History of Greece for Beginners (London and New York). Bussi, Silvia. 2001. Economia e demografia della schiavitù in Asia minore ellenisticoromana (Milan). Cairns, Douglas (ed.). 2001. Oxford Readings in Homer’s Iliad (Oxford). Cairns, Douglas and Fulkerson, Laura. 2015. ‘Introduction’. In Emotions between Greece and Rome, edited by Douglas Cairns and Laura Fulkerson. BICS Suppl. 125 (London: 1–22). Callender, Dexter E. 1998. ‘Servants of god(s) and servants of kings in Israel and the ancient Near East’, Semeia vol. 83–4: 67–82. Cameron, George G. 1948. Persepolis Treasury Tablets (Chicago). Campagno, Marcelo, Gallego, Julián, and Mac Gaw, Carlos García. (eds.). 2013. Rapports de subordination personnelle et pouvoir politique dans la Méditerranée antique et au-delà (Besançon). Canevaro, Mirko. 2013. The Documents in the Attic Orators. Laws and Decrees in the Public Speeches of the Demosthenic Corpus (Oxford). Canevaro, Mirko. 2016a. Demostene, Contro Leptine. Introduzione, traduzione e commento storico (Berlin and Boston, MA). Canevaro, Mirko. 2016b. ‘The popular culture of the Athenian institutions: “Authorized” popular culture and “unauthorized” elite culture in classical Athens’. In Locating Popular Culture in the Ancient World, edited by Lucy Grig (Cambridge: 39–65). Canevaro, Mirko. 2018. ‘The public charge for hybris against slaves: The honour of the victim and the honour of the hybristês’, Journal of Hellenic Studies vol. 138. Canevaro, Mirko and Harris, Edward M. 2012. ‘The documents in Andocides’ On the Mysteries’, Classical Quarterly vol. 62.1: 98–129. Canevaro, Mirko and Lewis, David M. 2014. ‘Khoris oikountes and the obligations of freedmen in late classical and early Hellenistic Athens’, Incidenza dell’antico vol. 12: 91–121. Cantarella, Eva. 1979. Norme e sanzione in Omero. Contributo alla protostoria del diritto greco (Milan). Cardellini, Innocenzo. 1981. Die Biblischen ‘Sklaven’ Gesteze im Lichte des keilscriftlichen Sklavenrechts: Ein Beitrag zur Tradizion, Überlieferung und Redaktion der alttestamentlichen Rechtstexte (Bonn). Carlsen, Jesper. 2002. ‘Estate managers in ancient Greek agriculture’. In Ancient History Matters: Studies Presented to Jens Erik Skydsgaard on his Seventieth Birthday, edited by Karen Ascani, Vincent Gabrielsen, Kirsten Kvist, and Anders Holm Rasmussen (Rome: 117–26).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

312

Bibliography

Carr, David. 2010. The Formation of the Hebrew Bible: A New Reconstruction (Oxford and New York). Cartledge, Paul A. 1979. Sparta and Lakonia: A Regional History, 1300–362 BC (London). Cartledge, Paul A. 1985. ‘Rebels and sambos in classical Greece: A comparative view’. In Crux: Essays in Greek History Presented to G. E. M. de Ste. Croix on his 75th Birthday, edited by Paul A. Cartledge and F. David Harvey (London: 16–46). Cartledge, Paul A. 1988. ‘Serfdom in classical Greece’. In Slavery and Other Forms of Unfree Labour, edited by Léonie J. Archer (London: 33–41). Cartledge, Paul A. 1993a. The Greeks: A Portrait of Self and Others (Oxford). Cartledge, Paul A. 1993b. ‘Like a worm i’ the bud? A heterology of classical Greek slavery’, Greece and Rome vol. 40: 163–80. Cartledge, Paul A. 2002. ‘The political economy of Greek slavery’. In Money, Labour and Land: Approaches to the Economies of Ancient Greece, edited by Paul A. Cartledge, Edward E. Cohen, and Lin Foxhall (London: 156–66). Cartledge, Paul A. 2003. ‘Raising hell? The helot mirage—a personal re-view’. In Helots and their Masters in Laconia and Messenia: Histories, Ideologies, Structures, edited by Nino Luraghi and Susan Alcock (Cambridge, MA, and London: 12–30). Cartledge, Paul A. 2011a. ‘The helots—a contemporary review’. In The Cambridge World History of Slavery. Vol. I: The Ancient Mediterranean World, edited by Keith R. Bradley and Paul A. Cartledge (Cambridge: 74–90). Cartledge, Paul A. 2011b. Ancient Greece: A Very Short Introduction (Oxford). Cartledge, Paul A. and Spawforth, Tony. 1989. Hellenistic and Roman Sparta: A Tale of Two Cities (London and New York). Carusi, Cristina. 2015. ‘Vita humanior sine sale non quit degere. Demand for salt and salt trade patterns in the ancient Greek world’. In The Ancient Greek Economy: Markets, Households and City-States, edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 337–55). Champion, Craig. 1997. ‘Gortyn, law code of ’. In The Historical Encyclopedia of World Slavery, Vol. I, edited by Junius P. Rodriguez (Santa Barbara, CA, Denver, CO, and Oxford: 310). Chaniotis, Angelos (ed.). 1999. From Minoan Farmers to Roman Traders. Sidelights on the Economy of Ancient Crete (Stuttgart). Chaniotis, Angelos. 2005. ‘The great inscription, its political and social institutions, and the common institutions of the Cretans’. In La grande iscrizione di Gortyna, edited by Emanuele Greco and Mario Lombardo (Athens: 175–94). Chandezon, Christophe. 2003. L’Élevage en Grèce (fin Ve-fin Ier s.a. C.) (Bordeaux). Chantraine, Pierre. 1974. Dictionnaire étymologique de la langue grecque. Vol. III (Paris). Chatzidimitriou, Athina. 2005. ΠΑΡΑΣΤΑΣΕΙΣ ΕΡΓΑΣΤΕΡΙΩΝ ΚΑΙ ΕΜΠΟΡΙΟΥ ΣΤΗΝ ΕΙΚΟΝΟΓΡΑΦΙΑ ΤΩΝ ΑΡΧΑΪΚΩΝ ΚΑΙ ΚΛΑΣΙΚΩΝ ΧΡΟΝΩΝ (Athens). Chatzidimitriou, Athina. 2010. ‘Transport of goods in the Mediterranean from the Geometric to the Classical period. Images and meaning’, Bollettino di archeologia online vol. 1: 1–20. Chen-Bo Zhong and Liljenquist, Katie. 2006. ‘Washing away your sins: threatened morality and physical cleansing’, Science vol. 313: 1451–2.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

313

Chirichigno, Gregory C. 1993. Debt Slavery in Israel and the Ancient Near East (Sheffield). Chrimes, Kathleen M. T. 1949. Ancient Sparta: A Re-Examination of the Evidence (Manchester). Christesen, Paul. 2003. ‘Economic rationalism in fourth-century BC Athens’, Greece and Rome vol. 50: 31–56. Christesen, Paul. 2004. ‘Utopia on the Eurotas: Economic aspects of the Spartan mirage’. In Spartan Society, edited by Thomas Figueira (Swansea: 309–37). Christesen, Paul. 2010. ‘Spartans and Scythians, a meeting of mirages: The portrayal of the Lycurgan politeia in Ephorus’ Histories’. In Sparta: The Body Politic, edited by Anton Powell and Stephen Hodkinson (Swansea: 211–63). Clastres, Pierre. 1974. La Société contre l’état. Recherches d’anthropologie politique (Paris). Clay, Albert T. 1908. Legal and Commercial Transactions Dated in the Assyrian, NeoBabylonian and Persian Periods, Chiefly from Nippur (Philadelphia, PA). Cohen, Edward E. 1992. Athenian Law and Society: A Banking Perspective (Princeton, NJ). Cohen, Edward E. 1998. ‘The wealthy slaves of Athens: Legal rights, economic obligations’. In Le Monde antique et les droits de l’homme, edited by Hugette Jones (Brussels: 105–29). Cohen, Edward E. 2000. The Athenian Nation (Princeton, NJ). Collombier, Anne-Marie. 2002. ‘Le Destin d’Eumée (Odysée XV, 400–484): traffics d’esclaves dans les poèmes homériques’. In Routes et marchés d’esclaves: Actes du 26e colloque du GIREA, Besançon, 27–29 septembre 2001, edited by Marguerite Garrido-Hory (Besançon: 7–19). Conophagos, Constantin. 1980. Le Laurium antique et la technique grecque de la production de l’argent (Athens). Conteneau, Georges. 1927. Contrats néo-babyloniens I: de Téglath-phalasar III à Nabonide (Paris). Conteneau, Georges. 1929. Contrats néo-babyloniens II: achéménides et séleucides (Paris). Coogan, Michael. 2009. ‘The social worlds of the Book of Job’. In Exploring the Longue Durée: Essays in Honour of Lawrence E. Stager, edited by J. David Schloen (Winona Lake, IN: 77–81). Couilloud, Marie-Thérèse. 1974. Les Monuments funéraires de Rhénée, Exploration archéologique de Délos XXX (Athens). Crielaard, Jan Paul. 1995. ‘Homer, history and archaeology: Some remarks on the date of the Homeric world’. In Homeric Questions, edited by Jan Paul Crielaard (Amsterdam: 201–88). Crosby, Margaret. 1950. ‘The leases of the Laurion mines’, Hesperia vol. 19: 189–312. Crosby, Margaret. 1957. ‘More fragments of mining leases from the Athenian agora’, Hesperia vol. 26: 1–23. Crouwel, Joost H. 1992. Chariots and Other Wheeled Vehicles in Iron Age Greece (Amsterdam). Cruz-Uribe, Eugene. 1982. ‘Slavery in Egypt during the Saite and Persian periods’, Revue Internationale des Droits de l’Antiquité 3rd ser., vol. 29: 46–71. Culbertson, Laura. 2011. ‘Slaves and households in the Near East’. In Slaves and Households in the Near East, edited by Laura Culbertson (Ann Arbor, MI: 1–17).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

314

Bibliography

Dal Lago, Enrico and Katsari, Constantina. 2008a. ‘The study of ancient and modern slave systems: Setting an agenda of comparison’. In Slave Systems Ancient and Modern, edited by Enrico Dal Lago and Constantina Katsari (Cambridge: 3–31). Dal Lago, Enrico and Katsari, Constantina. 2008b. ‘Ideal models of slave management in the Roman world and the ante-bellum American South’ In Slave Systems Ancient and Modern, edited by Enrico Dal Lago and Constantina Katsari (Cambridge: 187–213). Dalley, Stephanie. 2013. The Mystery of the Hanging Gardens of Babylon (Oxford). Dandamaev, Muhammad A. 1963. ‘Foreign slaves on the estates of the Achaemenid kings and their nobles’, Труды двадцать пятого Международного конгресса востоковедов, Москва 1960 (Proceedings of the 25th Congress of Orientalists, Moscow 1960) (Moscow: 147–54). Dandamaev, Muhammad A. 1975. ‘Forced labour in the palace economy of Achaemenid Iran’, Altorientalische Forschungen vol. 2: 71–8. Dandamaev, Muhammad A. 1979. ‘Le Statut juridique de l’ikkaru néo-babylonien’. In Terre et paysans dépendants dans les sociétés antiques. Colloque international tenu à Besançon les 2 et 3 mai 1974 (Paris: 153–61). Dandamaev, Muhammad A. 1984. Slavery in Babylonia. From Nabopolassar to Alexander the Great (626–331 BC). Translated by Victoria A. Powell. Rev. edn (Dekalb, IL). Dandamaev, Muhammad A. 1996. Review of Chirichigno 1993. The Jewish Quarterly Review vol. 87.1: 170–1. Dandamaev, Muhammad A. and Lukonin, Vladimir G. 1989. The Culture and Social Institutions of Ancient Iran (Cambridge). Davidson, James. 1997. Courtesans and Fishcakes. The Consuming Passions of Classical Athens (Chicago). Davies, John K. 1971. Athenian Propertied Families (Oxford). Davies, John K. 1981. Wealth and the Power of Wealth in Classical Athens (Salem, MA). Davies, John K. 1993. Democracy and Classical Greece. 2nd edn (Cambridge, MA). Davies, John K. 1996. ‘Deconstructing Gortyn: When is a code a code?’ In Greek Law in its Political Setting, edited by Lin Foxhall and Andrew D. E. Lewis (Oxford: 33–56). Davies, John K. 1998. ‘Ancient economies: models and muddles’. In Trade, Traders and the Ancient City, edited by Helen Parkins and Christopher Smith (London and New York: 225–56). Davies, John K. 2001. ‘Hellenistic economies in the post-Finley era’. In Hellenistic Economies, edited by Zosia Archibald, John K. Davies, Vincent Gabrielsen, and Graham J. Oliver (London: 11–62). Davies, John K. 2005a. ‘The Gortyn laws’. In The Cambridge Companion to Ancient Greek Law, edited by Michael Gagarin and David Cohen (Cambridge: 305–27). Davies, John K. 2005b. ‘Gortyn within the economy of archaic and classical Crete’. In La grande iscrizione di Gortyna, edited by Emanuele Greco and Mario Lombardo (Athens: 153–74). Davies, John K. 2015. ‘Towards a general model of long-distance trade: Aromatics as a case study’. In The Ancient Greek Economy: Markets, Households and City-States,

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

315

edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 299–315). Davies, Norman. 2011. Vanished Kingdoms: The History of Half-Forgotten Europe (London). Davies, Peter A. 2017. ‘Articulating status in ancient Greece: Status (in)consistency as a new approach’, Cambridge Classical Journal vol. 63: 29–52. Davies, Philip. 1992. In Search of ‘Ancient Israel’ (Sheffield). Debord, Pierre. 1999. L’Asie mineure au IVe siècle (412–323 a.C.). Pouvoir et jeux politiques (Bordeaux). Degler, Carl N. 1959. ‘Starr on slavery’, Journal of Economic History vol. 19: 271–7. Deller, Karlheinz, Fales, F. Mario, and Jakob-Rost, Liane. 1995. ‘Neo-Assyrian texts from Assur. Private archives in the Vorderasiatisches Museum in Berlin, part II’, State Archives of Assyria Bulletin vol. 9: 3–137. De Ste. Croix, Geoffrey E. M. 1972. The Origins of the Peloponnesian War (London). De Ste. Croix, Geoffrey E. M. 1981. The Class Struggle in the Ancient Greek World (London). De Ste. Croix, Geoffrey E. M. 1988. ‘Slavery and other forms of unfree labour’. In Slavery and Other Forms of Unfree Labour, edited by Léonie Archer (London: 19–32). De Ste. Croix, Geoffrey E. M. 2002. ‘The helot threat’. In Sparta, edited by Michael Whitby (Edinburgh: 211–6). Descat, Raymond. 2004. ‘L’Économie’. In Le Monde grec aux temps classiques, tome 2: le IVe siècle, edited by Pierre Brulé, Raymond Descat, Jean-Luc Lamboley, Sylvie Le Bohec, and Jacques Oulhen (Paris: 353−411). Descat, Raymond. 2006. ‘Argyrônetos: les transformations de l’échange dans la Grèce archaïque’. In Agoranomia: Studies in Money and Exchange Presented to John H. Kroll, edited by Peter van Alfen (New York: 21–36). Descat, Raymond. 2011. ‘Labour in the Hellenistic economy: Slavery as a test case’. In The Economies of Hellenistic Societies, Third to First Centuries BC, edited by Zosia Archibald, John K. Davies, and Vincent Gabrielsen (Oxford: 207–15). Deszö, Tamás and Vér, Ádám. 2013. ‘Assyrians and Greeks: The nature of contacts in the 9th–7th centuries BC’ Acta Antiqua Academiae Scientiarum Hungaricae vol. 53: 325–59. De Vaux, Roland. 1961. Ancient Israel: Its Life and Institutions (London). Dever, William G. 1993. ‘Biblical archaeology: Death and rebirth’. In Biblical Archaeology Today, 1990: Proceedings of the Second International Congress on Biblical Archaeology, edited by Avraham Biran (Jerusalem: 706–22). Dever, William G. 2001. What Did the Biblical Writers Know and When Did They Know it? (Grand Rapids, MI). Diakonoff, Igor M. (ed.). 1980. Ancient Mesopotamia: A Collection of Studies by Soviet Scholars (Schaan). Diakonoff, Igor M. 1992. ‘The naval power and trade of Tyre’, Israel Exploration Journal vol. 42: 168–93. Diels, Hermann A. and Kranz, Walther. 1952. Die Fragmente der Vorsokratiker. 6th edn (Berlin). Dimopoulou-Piliouni, Athina. 2008. ‘Apeleutheroi: Metics or foreigners?’, DIKE vol. 11: 27–50.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

316

Bibliography

Dmitriev, Sviatoslav. 2016. ‘The protection of slaves in the Athenian law against hubris’, Phoenix vol. 70: 64–76. Dobbs-Allsopp, Chip, Roberts, James M., Choon-Leong Seow, and Whitaker, Richard E. 2005. Hebrew Inscriptions: Texts from the Biblical Period of the Monarchy with Concordance (New Haven, CT, and London). Docter, Roald. 2009. ‘Carthage and its hinterland’, Iberia Archaeologica vol. 13: 179–89. Dougherty, Raymond P. 1920. Records from Erech, Time of Nabonidus (555–538 B.C.) (New Haven, CT). Dover, Kenneth J. 1974. Greek Popular Morality in the Time of Plato and Aristotle (Indianapolis, IN, and Cambridge). Drerup, Engelbert. 1898. ‘Über die bei den attischen Rednern eingelegten Urkunden’, Jahrbuch für Klassische Philologie, Supplementband 24: 221–366. Driver, Godfrey R. 1957. Aramaic Documents of the Fifth Century B.C. (Oxford). Dubois, Page. 2010. Slavery: Antiquity and its Legacy (London and New York). Ducat, Jean. 1990. Les Hilotes. BCH Suppl. 20 (Athens). Ducat, Jean. 1993. ‘L’Esclavage collectif en Illyrie à la recherche d’un hilotisme barbare’. In L’Illyrie meridionale et l’Épire dans l’Antiquité II. Actes du colloque international de Clermont-Ferrand (25–27 Octobre 1990), edited by Pierre Cabanes (Paris: 211–17). Ducat, Jean. 1994. Les Pénestes de Thessalie (Besançon). Ducat, Jean. 2006. Spartan Education. Youth and Society in the Classical Period. Translated by Emma Stafford, Pamela-Jane Shaw, and Anton Powell (Swansea). Ducat, Jean. 2015. ‘Les Hilotes à l’époque archaïque’. In La Main-d’ œuvre agricole en Méditerranée archaïque. Statuts et dynamiques économiques, edited by Julien Zurbach (Paris: 165–95). Ducrey, Pierre. 2007. ‘Le Monde antique est-il basé sur la peur? Peur des esclaves, peur de l’esclavage dans le monde gréco-romain’. In Fear of Slaves—Fear of Enslavement in the Ancient Mediterranean. XXIXe colloque du GIREA, Rethymnon, 2004, edited by Anastasia Serghidou (Besançon: 9–20). Duncan-Jones, Richard. 1984. ‘Problems of the Delphic manumission payments 200–1 BC’, Zeitschrift für Papyrologie und Epigraphik vol. 57: 203–9. Duplouy, Alain. 2006. Le Prestige des élites: recherches sur les modes de reconnaissance sociale en Grèce entre les xe et ve siècles avant J.-C. (Paris). Dušek, Jan. 2007. Les Manuscrits araméens du Wadi Daliyeh et la Samarie vers 450–332 av. J.-C. (Leiden). Dusinberre, Elspeth R. M. 2013. Empire, Authority, and Autonomy in Achaemenid Anatolia (Cambridge). Eagleman, David. 2015. The Brain: The Story of You (Edinburgh and London). Eco, Umberto. 1990. The Limits of Interpretation (Bloomington and Indianapolis, IN). Eden, Jeff. 2017. ‘Beyond the bazaars: Geographies of the slave trade in central Asia’, Modern Asia Studies vol. 51.4: 919–55. Ehrenberg, Victor. 1961. The People of Aristophanes (New York). Einstein, Albert. 1916. ‘Ernst Mach’, Physikalische Zeitschrift vol. 17: 101–4. Elayi, Josette. 1981. ‘La Révolte des esclaves de Tyr relatée par Justin’, Baghdader Mitteilungen vol. 12: 139–50.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

317

Eltis, David and Engerman, Stanley L. 2011. ‘Dependency, servility, and coerced labor in time and space’. In The Cambridge World History of Slavery, Vol. 3: AD 1420–AD 1804, edited by David Eltis and Stanley Engerman (Cambridge: 1–21). Epstein, Shimon. 2008. ‘Why did Attic building projects employ free labourers rather than slaves?’, Zeitschrift für Papyrologie und Epigraphik vol. 166: 108–12. Erickson, Brice. 2005. ‘Archaeology of empire: Athens and Crete in the fifth century B.C.’, American Journal of Archaeology vol. 109.4: 619–63. Fachard, Sylvian. 2013. ‘Eleutherai as the gates to Boeotia’, Revue des Études Militaires Anciennes vol. 6: 81–106. Fachard, Sylvian and Pirisino, Daniele. 2015. ‘Roads out of Attica’. In Autopsy in Athens. Recent Archaeological Work on Athens and Attica, edited by in Margaret M. Miles (Oxford: 139–53). Faist, Betina. 2009. ‘An Elamite Deportee’. In Homeland and Exile: Biblical and Ancient Near Eastern Studies in Honour of Bustenay Oded, edited by Gershon Galil, Markham J. Geller, and Alan Millard (Leiden: 59–69). Fales, F. Mario. 1973. Censimenti e catasti di epoca neo-assira (Rome). Fales, F. Mario. 1990. ‘The rural landscape of the Neo-Assyrian Empire: A survey’, State Archives of Assyria Bulletin vol. 4.2: 81–142. Fales, F. Mario. 1996. ‘Prices in Neo-Assyrian sources’, State Archives of Assyria Bulletin vol. 10.1: 11–53. Fales, F. Mario. 2001. L’impero assiro. Storia e amministrazioni (IX–VII secolo a.C.) (Bari). Fales, F. Mario. 2009–10. ‘On Assyrian “lower stratum” families’, State Archives of Assyria Bulletin vol. 18: 163–86. Fales, F. Mario. 2017. ‘Palatial economy in Neo-Assyrian documentation: An overview’. In Palatial Economy in the Ancient Near East and in the Aegean, edited by Pierre Carlier, Francis Joannès, Françoise Rougemont, and Julien Zurbach (Pisa and Rome: 271–94). Fales, F. Mario and Jakob-Rost, Liane. 1991. ‘Neo-Assyrian texts from Assur. Private archives in the Vorderasiatisches Museum in Berlin, part I (with two appendixes by K. Deller)’, State Archives of Assyria Bulletin vol. 5: 3–157. Fales, F. Mario. and Postgate, J. Nicholas (eds.). 1995. Imperial Administrative Records, Part II: Provincial and Military Administration. State Archives of Assyria Volume XI (Helsinki). Falkenstein, Adam. 1952. Review of Cameron 1948. Die Welt des Orients vol. 1: 503–6. Fantalkin, Alexander and Lytle, Ephraim. 2016. ‘Alcaeus and Antimenidas: Reassessing the evidence for Greek mercenaries in the Neo-Babylonian army’, Klio vol. 98.1: 90–117. Faraguna, Michele. 2008. ‘Calcolo economico, archivi finanziari e credito nel mondo greco tra VI e IV sec. a.C.’. In Pistoi Dia Tèn Technèn. Bankers, Loans and Archives in the Ancient World: Studies in Honour of Raymond Bogaert, edited by Koenraad Verboven, Katelijn Vandorpe, and Véronique Chankowski (Leuven: 33–57). Faucher, Thomas and Redon, Bérangère. 2014. ‘Le Prix de l’entrée au bain en Égypte hellénistique et romaine d’après les donées textuelles et numismatiques’. In 25 siècles de bain collectif en Orient, edited by Marie-Françoise Boussac, Sylvie Denoix, Thibaud Fournet, and Bérangère Redon (Cairo: 835–56).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

318

Bibliography

Faust, Avraham. 2006. ‘Early Israel: An egalitarian society’, Biblical Archaeology Review vol. 39.4: 45–63. Faust, Avraham. 2008. ‘Cities, villages and farmsteads: The landscape of Leviticus 25:29–31’. In Exploring the Longue Durée: Essays in Honour of Lawrence E. Stager, edited by J. David Schloen (Winona Lake, IN: 103–12). Faust, Avraham. 2012. The Archaeology of Israelite Society in Iron Age II (Winona Lake, IN). Faust, Avraham and Weiss, Ehud. 2005. ‘Judah, Philistia, and the Mediterranean world: Reconstructing the economic system of the seventh century B.C.E.’, Bulletin of the American Schools of Oriental Research vol. 338: 71–92. Ferone, Claudio. 1997. Lesteia. Forme di predazione nell’Egeo in età classica (Naples). Février, James-Germain. 1951–2. ‘Vir Sidonius’, Semitica vol. 4: 13–18. Février, James-Germain. 1961. ‘Textes puniques et néopuniques relatifs aux testaments’, Semitica vol. 11: 5–8. Figueira, Thomas. 2003. ‘Xenelasia and social control in classical Sparta’, Classical Quarterly vol. 53.1: 44–74. Figueira, Thomas. 2004. ‘The nature of the Spartan klêros’. In Spartan Society, edited by Thomas Figueira (Swansea: 47–76). Finkelman, Paul. 1985. Review of Patterson 1982. The Journal of Interdisciplinary History vol. 15.3: 508–11. Finkelstein, Israel, and Silberman, Neil Asher. 2001. The Bible Unearthed: Archaeology’s New Vision of Ancient Israel and the Origin of its Sacred Texts (New York). Finley, Moses I. 1936. Review of Westermann 1935. Zeitschrift für Sozialforschung vol. 5: 441–2. Finley, Moses I. 1954. The World of Odysseus (London). Finley, Moses I. 1960. ‘The Servile Statuses of Ancient Greece’, Revue internationale des droits de l’antiquité vol. 7: 165–89. Finley, Moses I. 1964. ‘Between slavery and freedom’, Comparative Studies in Society and History vol. 6: 233–49. Finley, Moses I. 1968. ‘Slavery’. In The International Encyclopedia of the Social Sciences. Vol. 14, edited by David L. Sills (New York and London: 307–13). Finley, Moses I. 1969. ‘The idea of slavery: Critique of David Brion Davis’ The Problem of Slavery in Western Culture’. In Slavery in the New World: A Reader in Comparative History, edited by Laura Foner and Eugene D. Genovese (Engelwood Cliffs, NJ: 256–61). Finley, Moses I. 1971. The Use and Abuse of History (London). Finley, Moses I. 1980. Ancient Slavery and Modern Ideology (London). Finley, Moses I. 1981. Economy and Society in Ancient Greece, edited by Brent Shaw and Richard Saller (London). Finley, Moses I. 1999 [orig. 1973]. The Ancient Economy. Updated edition with a foreword by Ian Morris (Berkeley, Los Angeles, and London). Finnegan, Ruth. 1977. Oral Poetry (Cambridge). Fischer, Josef. 2016. ‘Der Schwarzmeerraum und der antike Sklavenhandel. Bemerkungen zu einigen ausgewählten Quellen’. In Akten des 15. Österreichischen Althistorikertages Salzburg, 20.–22. November 2014, edited by in Monika Frass, Herbert Graßl, and Georg Nightingale (Salzburg: 53–71).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

319

Fisher, Nick. 1993. Slavery in Classical Greece (London). Fisher, Nick. 1995. ‘Hybris, Status and Slavery’. In The Greek World, edited by Anton Powell (London: 44–84). Fisher, Nick. 2001. Aeschines, Against Timarchus (Oxford). Fitzhardinge, Laurence F. 1980. The Spartans (London). Flaig, Egon. 2009. Weltgeschichte der Sklaverei (Munich). Fleming, Daniel E. 2004. Democracy’s Ancient Ancestors: Mari and Early Collective Governance (Cambridge). Fleming, Daniel E. 2012. The Legacy of Israel in Judah’s Bible: History, Politics, and the Reinscribing of Tradition (Cambridge). Flower, Michael. 2002. ‘The invention of tradition in classical and Hellenistic Sparta’. In Sparta: Beyond the Mirage, edited by Anton Powell and Stephen Hodkinson (Swansea: 193–219). Fogel, Robert W. and Engerman, Stanley L. 1974. Time on the Cross: the Economics of American Negro Slavery (Boston, MA). Forrest, W. George. 1968. A History of Sparta, 950–192 BC (London). Forsdyke, Sara. 2012. Slaves Tell Tales, and other Episodes in the Politics of Popular Culture in Ancient Greece (Princeton, NJ). Foxhall, Lin. 1992. ‘The control of the Attic landscape’. In Agriculture in Ancient Greece, edited by Berit Wells (Stockholm: 155–9). Foxhall, Lin. 1997. ‘A view from the top: Evaluating the Solonian property classes’. In The Development of the Polis in the Archaic Period, edited by Lynette Mitchell and P. J. Rhodes (London and New York: 61–74). Foxhall, Lin. 1998. ‘Cargoes of the heart’s desire: The character of trade in the archaic Mediterranean world’. In Archaic Greece: New Approaches and New Evidence, edited by Nick Fisher and Hans van Wees (London: 295–309). Foxhall, Lin. 2002. ‘Access to resources in classical Greece: The egalitarianism of the polis in practice’. In Money, Labour and Land: Approaches to the Economies of Ancient Greece, edited by Paul A. Cartledge, Edward E. Cohen, and Lin Foxhall (London: 209–20). Foxhall, Lin. 2007. Olive Cultivation in Ancient Greece: Seeking the Ancient Economy (Oxford). Foxhall, Lin. 2013. ‘Can we see the “hoplite revolution” on the ground?’ In Men of Bronze: Hoplite Warfare in Ancient Greece, edited by Donald Kagan and Gregory F. Viggiano (Princeton, NJ: 194–221). Frisk, Hjalmar. 1970. Griechisches Etymologisches Wörterbuch. Vol. II (Heidelberg). Fynn-Paul, Jeffrey. 2009. ‘Empire, monotheism and slavery in the greater Mediterranean region from antiquity to the early modern era’, Past and Present vol. 205: 3–40. Gabrielsen, Vincent. 1986. ‘Phanera and aphanes ousia in classical Athens’, Classica & Mediaevalia vol. 37: 99–114. Gabrielsen, Vincent. 2001. ‘Economic activity, maritime trade and piracy in the Hellenistic Aegean’, Revue des Études Anciennes vol. 103: 219–40. Gabrielsen, V. (2003) ‘Piracy and the slave trade’. In A Companion to the Hellenistic World, edited by A. Erskine (Oxford: 389–404).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

320

Bibliography

Gabrielsen, Vincent. 2013. ‘Rhodes and the Ptolemaic kingdom: The commercial infrastructure’. In The Ptolemies, the Sea and the Nile: Studies in Waterborne Power, edited by Kostas Buraselis, Mary Stefanou, and Dorothy Thompson (Cambridge: 66–81). Gagarin, Michael. 1988. ‘The first law of the Gortyn Code’, Greek Roman and Byzantine Studies vol. 29: 335–43. Gagarin, Michael. 1996. ‘The torture of slaves in Athenian law’, Classical Philology vol. 91.1: 1–18. Gagarin, Michael. 2010a. ‘Serfs and slaves at Gortyn’, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Romanistische Abteilung vol. 127: 14–31. Gagarin, M. 2010b. ‘Legal procedure at Gortyn’. In Symposion 2009. Akten der Gesellschaft für griechische und hellenistische Rechtsgeschichte, edited by Gerhard Thür (Vienna: 127–45). Gagarin, Michael and Perlman, Paula. 2016. The Laws of Ancient Crete, c.650–400 BCE (Oxford). Galil, Gershon. 2007. The Lower Stratum Families in the Neo-Assyrian Period (Leiden and Boston, MA). Gallagher, William R. 1994. ‘Assyrian deportation propaganda’, State Archives of Assyria Bulletin vol. 8.2: 57–65. Gallant, Thomas. 1991. Risk and Survival in Ancient Greece. Reconstructing the Rural Domestic Economy (Stanford, CA). Gallego, Julián. 2007. ‘Farming in the ancient Greek world: How should the small free producers be defined?’, Studia Humaniora Taurensia vol. 8: 1–21. Garlan, Yvon. 1982. Les esclaves en Grèce ancienne (Paris). Garlan, Yvon. 1987. ‘War, piracy and slavery in the Greek world’. In Classical Slavery, edited by Moses I. Finley (London: 7–21). Garlan, Yvon. 1988. Slavery in Ancient Greece. Translated by Janet Lloyd (Ithaca, NY). Garnsey, Peter. 1996. Ideas of Slavery from Aristotle to Augustine (Cambridge). Gates, Henry Louis, Jr. 2002. The Classic Slave Narratives (New York). Gawlinski, Laura. 2012. The Sacred Law of Andania: A New Text with Commentary (Berlin and Boston, MA). Geary, Dick and Vlassopoulos, Kostas (eds.). 2009. Slavery, Citizenship and the State in Classical Antiquity and the Modern Americas. Special issue of European Review of History: Revue Européenne d’Histoire, Vol. 16.3: 295–436. Gehrke, Hans-Joachim. 1997. ‘Gewalt und Gesetz. Die soziale und politische Ordnung Kretas in der archaischen und klassichen Zeit’, Klio vol. 79: 23–68. Gentner, Dedre and Bowdle, Brian. 2008. ‘Metaphor as structure-mapping’. In The Cambridge Handbook of Metaphor and Thought, edited by Raymond W. Gibbs, Jr (Cambridge: 109–28). Gigon, Olof. 1960–87. Updated edition of Immanuel Bekker, Aristotelis Opera. 3 vols. (Berlin). Ginzberg, Carlo. 2013. ‘Our words, and theirs: a reflection on the historian’s craft, today’, Cromohs vol. 18: 97–114. Glazebrook, Alison and Tsakirgis, Barbara (eds.). 2016. Houses of Ill Repute: The Archaeology of Brothels, Houses and Taverns in the Greek World (Philadelphia, PA). Goette, Hans R. (ed.). 2002. Ancient Roads in Greece. Proceedings of a Symposion Organized by the Cultural Association Aigeas (Athens) and the German Archaeological

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

321

Institute (Athens) with the Support of the German School at Athens, November 23, 1998 (Athens). Goette, Hans R. 2010. ‘Vorstudien zum antiken Marmortransport in Attika und zum Demos Trinemeia’. In Attika: Archäologie einer ‘zentralen’ Kulturlandschaft, edited by Hans Lohmann and Torsten Mattern (Wiesbaden: 131–6). Golden, Mark. 1988. ‘The effects of slavery on citizen households and children: Aeschylus, Aristophanes and Athens’, Historical Reflections/Réflexions historiques vol. 15.3: 455–75. Golden, Mark. 2008. Greek Sport and Social Status (Austin, TX). Gomme, Arnold W., Andrewes, Antony A., and Dover, Kenneth. 1970. A Historical Commentary on Thucydides. Vol. IV: Books V.25-VII (Oxford). Gonzales, Antonio (ed.). 2008. La Fin du statut servile? Affranchissement, libération, abolition. XXXe colloque du GIREA, Besançon 2005. 2 vols. (Besançon). Goody, Jack. 1969. ‘Inheritance, property, and marriage in Africa and Eurasia’, Sociology vol. 3.1: 55–76. Goody, Jack. 1977. The Domestication of the Savage Mind (Cambridge). Goody, Jack. 2006. The Theft of History (Cambridge). Gottschalk, Hans B. 1972. ‘Notes on the wills of the Peripatetic scholars’, Hermes vol. 100: 314–42. Granovetter, Mark. 1985. ‘Economic action and social structure: The problem of embeddedness’, American Journal of Sociology vol. 91.3: 481–510. Greene, Elizabeth S., Leidwanger, Justin, and Özdaş, Harun. 2013. ‘Expanding contacts and collapsing distances in early Cypro-archaic trade: Three case studies of shipwrecks off the Turkish coast’. In Transport Amphorae and Trade of Cyprus, edited by Mark Lawall and John Lund (Aarhus: 21–34). Gropp, Douglas et al. 2001. Wadi Daliyeh II. The Samaria Papyri from Wadi Daliyeh (Oxford). Groß, Melanie and Pirngruber, Reinhard. 2014. ‘On courtiers in the Neo-Assyrian Empire: ša-rēši and mazzāz-pani’, Altorientalische Forschungen vol. 41.2: 161–75. Grote, George. 1862. A History of Greece (London). (Wiesbaden). Gschnitzer, Fritz. 1976. Studien zur griechischen Terminologie der Sklaverei. Vol. 2 (Wiesbaden). Gsell, Stefan. 1920. Histoire ancienne de l’Afrique du nord IV. La civilisation carthaginoise (Paris). Guarducci, Margherita. 1950. Inscriptiones Creticae IV: Titulae Gortynii (Rome). Guía, Miriam V. and Gallego, Julián. 2010. ‘Athenian zeugitai and the Solonian census classes: New reflections and perspectives’, Historia vol. 59.3: 257–81. Gundlach, Rolf. 1994. Die Zwangsumsiedlung auswärtiger Bevölkerung als Mittel ägyptischer Politik bis zum Ende des mittleren Reiches (Stuttgart). Gschnitzer, Fritz. 1963. Studien zur griechischen Terminologie der Sklaverei. Vol. 1 Hackl, Johannes. 2010 ‘Apprenticeship contracts’. In Aspects of the Economic History of Babylonia in the First Millennium BC, edited by Michael Jursa (Münster: 700–25). Hadjisavvas, Sophocles and Chaniotis, Angelos. 2012. ‘Wine and olive oil in Crete and Cyprus: Socio-economic aspects’. In Parallel Lives: Ancient Island Societies in Crete and Cyprus, edited by Gerald Cadogan, Maria Iacovou, Katerina Kopaka, and James Whitley (London: 157–73).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

322

Bibliography

Haggis, Donald et al. 2004. ‘Excavations at Azoria, 2002’, Hesperia vol. 73.3: 339–400. Hahn, István. 1971. ‘Die Anfänge der antiken Gesellschaftsformation in Griechenland und das Problem der sogennanten asiatischen Produktionsweise’, Jahrbuch für Wirtschaftsgeschichte vol. 2: 29–47. Hall, Jonathan. 2007. A History of the Archaic Greek World, ca. 1200–479 BCE (Oxford). Hallock, Richard T. 1969. Persepolis Fortification Tablets (Chicago). Hallock, Richard T. 1978. ‘Selected Fortification Texts’, Cahiers de la Délégation Française en Iran vol. 8: 109–36. Halpern, Baruch. 1995. ‘Erasing history: The minimalist assault on ancient Israel’ Bible Review vol. 11.6: 26–47. Halpern, Baruch. 2001. David’s Secret Demons: Messiah, Murderer, Traitor, King (Grand Rapids, MI). Hann, Chris. 1998. ‘Introduction: The embeddedness of property’. In Property Relations: Renewing the Anthropological Tradition, edited by Chris Hann (Cambridge: 1–47). Hann, Chris. (ed.). 2005. Property Relations: The Halle Focus Group, 2000–2005 (Halle). Hansen, Mogens H. 1985. Demography and Democracy: The Number of Athenian Citizens in the Fourth Century B.C. (Herning). Hansen, Mogens H. 1988. Three Studies in Athenian Demography (Copenhagen). Hansen, Mogens H. 2006a. Polis: An Introduction to the Ancient Greek City-State (Oxford). Hansen, Mogens H. 2006b. Studies in the Population of Aigina, Athens and Eretria (Copenhagen). Hansen, Mogens H. 2006c. The Shotgun Method: The Demography of the Ancient Greek City-State Culture (Columbia, MO, and London). Hardin, James W. 2010. Lahav II. Households and the Use of Domestic Space at Iron II Tell Halif: An Archaeology of Destruction (Winona Lake, IN). Harper, Kyle. 2008. ‘The Greek census inscriptions of late antiquity’, Journal of Roman Studies vol. 98: 83–119. Harper, Kyle. 2010. ‘Slave prices in late antiquity (and the very long term)’, Historia vol. 59.2: 206–38. Harper, Kyle. 2011a. Slavery in the Late Roman World, AD 275–425 (Cambridge). Harper, Kyle. 2011b. ‘Review article: Knowledge, ideology and skepticism in ancient slavery studies’, American Journal of Philology vol. 132.1: 160–8. Harper, Kyle and Scheidel, Walter. 2018. ‘Roman slavery and the idea of “slave society” ’. In What is a Slave Society? The Practice of Slavery in Global Perspective, edited by Noel Lenski and Catherine Cameron (Cambridge: 86–105). Harris, Edward M. 1995. Aeschines and Athenian Politics (Oxford and New York). Harris, Edward M. 1997. ‘A new solution to the riddle of the seisachtheia’. In The Development of the Polis in Archaic Greece, edited by Lynette Mitchell and P. J. Rhodes (London: 55–60). Harris, Edward M. 2002a. ‘Did Solon Abolish Debt Bondage?’, Classical Quarterly vol. 52: 415–30. Harris, Edward M. 2002b. ‘Workshop, marketplace and household: The nature of technical specialisation in classical Athens and its influence on economy and

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

323

society’. In Money, Labour, and Land: Approaches to the Economies of Ancient Greece, edited by Paul A. Cartledge, Edward E. Cohen, and Lin Foxhall (London: 67–99). Harris, Edward M. 2004. ‘Notes on a lead letter from the Athenian agora’, Harvard Studies in Classical Philology vol. 102: 157–70. Harris, Edward M. 2006. Democracy and the Rule of Law in Classical Athens (Cambridge and New York). Harris, Edward M. 2008. ‘Notes on two legal inscriptions’, Zeitschrift für Papyrologie und Epigraphik vol. 167: 81–4. Harris, Edward M. 2012. ‘Homer, Hesiod, and the ‘Origins’ of Greek Slavery’, Revue des Études Anciennes vol. 114.2: 345–66. Harris, Edward M. 2013a. The Rule of Law in Action in Democratic Athens (Oxford and New York). Harris, Edward M. 2013b. ‘How to address the Athenian assembly. Rhetoric and politcal tactics in the debate about Mytilene’, Classical Quarterly vol. 63.1: 94–109. Harris, Edward M. 2013c. ‘Were there business agents in ancient Greece? The evidence of some lead letters’. In The Letter: Law, State, Society and the Episolary Format in the Ancient World, edited by Uri Yiftach-Firanko (Wiesbaden: 105–24). Harris, Edward M. 2013d. ‘Finley’s Studies in Land and Credit sixty years later’, DIKE vol. 16: 123–46. Harris, Edward M. 2014. ‘Wife, household, and marketplace: The role of women in the economy of classical Athens’. In Donne che contano nella storia greca, edited by Umberto Bultrighini and Elisabetta Dimauro (Lanciano: 185–207). Harris, Edward M. 2015. ‘The legal foundations of economic growth in ancient Greece: The role of property records’. In The Ancient Greek Economy: Markets, Households and City-States, edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 116–46). Harris, Edward M. Forthcoming. ‘The dedication of phialai by metics and citizens. Or, appyling Ockham’s Razor to the interpretation of some Attic inscriptions’. Harris, Edward M. and Lewis, David M. 2015. ‘Introduction’. In The Ancient Greek Economy: Markets, Households and City-States, edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 1–37). Harris, William V. 1980. ‘Towards a study of the Roman slave trade’, Memoirs of the American Academy in Rome vol. 36: 117–40. Harris, William V. (ed.). 2013. Moses Finley and Politics (Leiden). Harrison, Alick R. W. 1968. The Law of Athens. The Family and Property (Oxford). Harvey, F. David. 1988. ‘Herodotus and the man-footed creature’. In Slavery and Other Forms of Unfree Labour, edited by Léonie Archer (London: 42–52). Harvey, F. David. 1990. ‘The sycophant and sycophancy: Vexatious redefinition?’ In Nomos: Essays in Athenian Law, Politics and Society, edited by Paul A. Cartledge, Paul Millett, and Stephen Todd (Cambridge: 103–21). Harvey, F. David. 2004. ‘The clandestine massacre of the helots (Thucydides 4.80)’. In Spartan Society, edited by Thomas Figueira (Swansea: 199–217). Harvey, F. David. 2007. ‘ “Help! I’m dying here.” A letter from a slave’, Zeitschrift für Papyrologie und Epigraphik vol. 163: 49–50. Haspels, C. H Emilie. 1936. Attic Black-Figured Lekythoi (Paris).

OUP CORRECTED PROOF – FINAL, 28/6/2018, SPi

324

Bibliography

Head, Ronan. 2010. ‘The Business Activities of Neo-Babylonian Private ‘Slaves’ and the Problem of Peculium’. Diss. Johns Hopkins. Heinen, Heinz. 1978. ‘Ägyptische und griechische Traditionen der Sklaverei im ptolemäischen Ägypten’. In Das ptolemäische Ägypten. Akten des internationalen Symposions 27.–29. September 1976 in Berlin, edited by Herwig Maehler and V. Michael Strocka (Mainz: 227–37). Heinen, Heinz. 2001. ‘Sklaverei im nördlichen Schwarzmeerraum: Zum Stand der Forschung’. In Fünfzig Jahre Forschungen zur antiken Sklaverei an der Mainzer Akademie 1950–2000, edited by Heinz Bellen and Heinz Heinen (Stuttgart: 487–503). Heinen, Heinz. 2010. ‘Ascesa e declino della ricerca sovietica sulla schiavitù: uno studio sul rapporto tra politica e scienza’, Rivista Storica Italiana vol. 122: 735–83. Henkelman, Wouter F. 2009. ‘Ethnic identity and labelling at Persepolis: The case of the Skudrians’. In Organisation de pouvoirs et contacts culturels dans les pays de l’empire achéménide, edited by Pierre Briant and Michel Chauveau (Paris: 271–329). Henkelman, Wouter F. 2013. ‘Administrative realities: The Persepolis archives and the archaeology of the Achaemenid heartland’. In The Oxford Handbook of Ancient Iran, edited by Daniel T. Potts (Oxford: 528–46). Henkelman, Wouter F. and Kleber, Kristin. 2007. ‘Babylonian workers in the Persian heartland: Palace building at Matannan in the reign of Cambyses’. In Persian Responses, edited by Christopher Tuplin (Swansea: 163–76). Herzogenrath-Amelung, Tristan. 2017. ‘Naval hoplites: Social status and combat reality of classical Greek epibatai’, Historia vol. 66: 45–64. Hezser, Catherine. 2005. Jewish Slavery in Antiquity (Oxford). Hezser, Catherine. 2016. ‘Blurred boundaries between slaves and free persons in ancient Judaism’, Association for Jewish Studies Perspectives: The Freedom Issue (Fall/Winter 2016: 44–5). Higman, B. W. (2001) ‘The invention of slave society’. In Slavery, Freedom and Gender: The Dynamics of Caribbean Slavery, edited by Brian L. Moore, B. W. Higman, Carl Campbell, and Patrick Bryan (Kingston: 57–75). Higman, B. W. 2011. ‘Demography and family structures’. In The Cambridge World History of Slavery, Vol. 3: AD 1420–AD 1804, edited by David Eltis and Stanley L. Engerman (Cambridge: 479–511). Hinz, Walther. 1949. Review of Cameron 1948. Zeitschrift für Assyriologie vol. 49: 347–53. Hodkinson, Stephen. 1986. ‘Land tenure and inheritance in classical Sparta’, Classical Quarterly 36: 378–406. Hodkinson, Stephen. 1988. ‘Animal husbandry in the Greek polis’. In Pastoral Economies in Classical Antiquity, edited by Charles R. Whittaker (Cambridge: 35–74). Hodkinson, Stephen. 1992. ‘Sharecropping and Sparta’s economic exploitation of the helots’. In ΦΙΛΟΛΑΚΟΝ. Lakonian Studies in Honour of Hector Catling, edited by Jan M. Sanders (London: 123–34). Hodkinson, Stephen. 1997. ‘The development of Spartan society and institutions in the archaic period’. In The Development of the Polis in Archaic Greece, edited by Lynette Mitchell and P. J. Rhodes (London: 83–102). Hodkinson, Stephen. 2000. Property and Wealth in Classical Sparta (Swansea).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

325

Hodkinson, Stephen. 2008. ‘Spartiates, helots and the direction of the agrarian economy: Toward an understanding of helotage in comparative perspective’. In Slave Systems, Ancient and Modern, edited by Enrico Dal Lago and Constantina Katsari (Cambridge: 285–320). Hodkinson, Stephen and Geary, Dick (eds.). 2012. Slaves and Religions in GraecoRoman Antiquity and Modern Brazil (Newcastle upon Tyne). Hodkinson, Stephen and Hall, Edith. 2011. ‘Appropriations of Spartan helotage in British antislavery debates of the 1790s’. In Ancient Slavery and Abolition: From Hobbes to Hollywood, edited by Richard Alston, Edith Hall, and Justine McConnell (Oxford: 65–102). Hoebel, E. Adamson. 1966. Anthropology: The Study of Man (New York). Holz, Shalom E. 2009. Neo-Babylonian Court Procedure (Leiden and Boston, MA). Honoré, Antony M. 1961. ‘Ownership’. In Oxford Essays in Jurisprudence, edited by Antony G. Guest (Oxford: 107–47). Honoré, Antony M. 1987. ‘Rights and the Rightless’. In Antony M. Honoré, Making Law Bind (Oxford: 241–55). Hooker, J. T. 1980. The Spartans (London). Hopkins, Keith. 1978. Conquerors and Slaves (Cambridge). Hopkins, Keith and Roscoe, P. J. 1978. ‘Between slavery and freedom: On freeing slaves at Delphi’. In Conquerors and Slaves, by Keith Hopkins (Cambridge: 133–71). Hopper, Matthew S. 2013. ‘Slaves of one master: Globalization and the African diaspora in Arabia in the age of Empire’. In Indian Ocean Slavery in the Age of Abolition, edited by Robert W. Harms, Bernard K. Freamon, and David W. Blight (New Haven, CT: 223–40). Horden, Peregrine and Purcell, Nicholas. 2000. The Corrupting Sea (Oxford). Hornblower, Simon. 2000. ‘Sticks, stones, and Spartans: The sociology of Spartan violence’. In War and Violence in Ancient Greece, edited by Hans van Wees (London: 57–82). Hoyos, Dexter. 2010. The Carthaginians (Abingdon and New York). Hunnings, Leanne. 2011. ‘Paradigms of execution: Managing slave death from Homer to Virginia’. In Reading Ancient Slavery, edited by Richard Alston, Edith Hall, and Laura Proffitt (London: 51–71). Hunt, Peter. 1998. Slaves, Warfare and Ideology in the Greek Historians (Cambridge). Hunt, Peter. 2011. ‘Slaves in Greek literary culture’. In The Cambridge World History of Slavery: Vol. I: The Ancient Mediterranean World, edited by Keith R. Bradley and Paul Cartledge (Cambridge: 22–47). Hunt, Peter. 2015. ‘Trojan slaves in classical Athens: Ethnic identity among Athenian slaves’. In Communities and Networks in the Ancient Greek World, edited by Claire Taylor and Kostas Vlassopoulos (Oxford: 128–54). Hunt, Peter. 2016. ‘Slaves or serfs? Patterson on the thetes and helots of ancient Greece’. In On Human Bondage: After Slavery and Social Death, edited by John Bodel and Walter Scheidel (Chichester and Malden, MA: 55–80). Hunter, Virginia. 1994. Policing Athens: Social Control in the Athenian Lawsuits, 420–320 B.C. (Princeton, NJ). Hunter, Virginia. 2000. ‘Introduction: Status Distinctions in Athenian Law’. In Law and Social Status in Classical Athens, edited by Virginia Hunter and Jonathan Edmondson (Oxford: 1–29).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

326

Bibliography

Iancu, Liviu M. 2016. ‘Greek and other Aegean mercenaries in the archaic age: Aristocrats, common people, or both?’, Studia Hercynia vol. 20.2: 9–29. Irwin, Robert. 2006. For Lust of Knowing: The Orientalists and their Enemies (London). Isager, Signe and Hansen, Mogens H. 1975. Aspects of Athenian Society in the Fourth Century B.C. (Odense). Ismard, Paulin. 2011. ‘Les Esclaves publics des cités grecques: Qu’est-ce qu’un statut personnel?’, EKEIED vol. 43: 27–41. Ismard, Paulin. 2015. La Démocratie contre les experts. Les esclaves publics en Grèce ancienne (Paris). Jackson, Bernard S. 1988. ‘Biblical laws of slavery: A comparative approach’. In Slavery and Other Forms of Unfree Labour, edited by Léonie Archer (London and New York: 86–101). Jacoby, Felix. 1923–58. Die Fragmente der griechischen Historiker (Berlin; Leiden). Jameson, Michael. 1977–8. ‘Agriculture and slavery in classical Athens’, Classical Journal vol. 73: 122–145. Jameson, Michael. 1992. ‘Agricultural labour in ancient Greece’. In Agriculture in Ancient Greece, edited by Berit Wells (Stockholm: 135–46). Jameson, Michael. 2002. ‘On Paul Cartledge, “The Political Economy of Greek Slavery”’. In Money, Labour and Land: Approaches to the Economies of Ancient Greece, edited by Paul A. Cartledge, Edward E. Cohen, and Lin Foxhall (London: 167–74). Janković, Bojana. 2010. ‘Uruk’. In Aspects of the Economic History of Babylonia in the First Millennium BC, by Michael Jursa (Münster: 418–37). Jeanmaire, Henri. 1939. Couroi et courètes: essai sur l’éducation spartiate et sur les rites d’adolescence dans l’antiquité hellénique (Lille). Jew, Daniel, Osborne, Robin, and Scott, Michael (eds.). 2016. Moses Finley: An Ancient Historian and his Impact (Cambridge). Jiménez, Enrique, Adalı, Selim F., and Radner, Karen. 2015. ‘Four 7th-century BCE Neo-Assyrian slave sale records from Marqasi (Kahramanmaraş) in the Erimtan Museum (Ankara) and elsewhere’, Altorientalischen Forschungen vol. 42.2: 153–72. Joannès, Francis. 1992. ‘Inventaire d’un cabaret’ Nouvelles Assyriologiques Brèves et Utilitaires (June): 48–50, no. 64. Jones, Arnold H. M. 1952. ‘The economic basis of Athenian democracy’, Past and Present vol. 1: 13–31. Jones, Arnold H. M. 1967. Sparta (Oxford). Jones, Christopher. 1987. ‘Stigma: Tattooing and branding in Graeco-Roman antiquity’, Journal of Roman Studies vol. 77: 139–55. Jones, Nicholas F. 2004. Rural Athens under the Democracy (Philadelphia, PA). Joosten, Jan. 2013. ‘Linguistic clues as to the date of the Book of Job: A mediating position’. In Interested Readers: Essays on the Hebrew Bible in Honor of David J. A. Clines, edited by James K. Aitken, Jeremy M. S. Clines, and Christl M. Maier (Atlanta, GA: 347–57). Jordan, David P. 1995. Transforming Paris: The Life and Labors of Baron Haussmann (New York). Joshel, Sandra. 2010. Slavery in the Roman World (Cambridge). Joshel, Sandra and Murnaghan, Sheila. 1998. ‘Introduction’. In Women and Slaves in Greco-Roman Culture: Differential Equations, edited by Sandra Joshel and Sheila Murnaghan (Abingdon and New York: 1–21).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

327

Jursa, Michael. 2005. Neo-Babylonian Legal and Administrative Archives: Typology, Contents, and Archives (Münster). Jursa, Michael. 2008. ‘Economic change and legal innovation on aspects of commercial interaction and land tenure in Babylonia in the first millennium BC’. In I diritti del mondo cuneiforme, edited by Mario Liverani and Clelia Mora (Pavia: 601–28). Jursa, Michael. 2010. Aspects of the Economic History of Babylonia in the First Millennium BC. With contributions by Johannes Hackl, Bojana Janković, Kristin Kleber, Elizabeth E. Payne, Caroline Waerzeggers, and Michaela Weszeli (Münster). Jursa, Michael, Paszkowiak, J., and Waerzeggers, Caroline. 2003–4. ‘Three court records’, Archiv für Orientforschung vol. 50: 255–68. Kahnemann, Daniel. 2011. Thinking, Fast and Slow (London). Kakavogianni, Olga and Anetakis, Michalis. 2012. ‘Les Agoras commerciales des dèmes antiques de la Mésogée et de la région du Laurion’. In Tout vendre, tout acheter. Structures et équipements des marchés antiques, edited by Véronique Chankowski and Pavlos Karvonis (Bordeaux: 185–99). Kamen, Deborah. 2005. ‘Conceptualizing Manumission in Ancient Greece’. Diss. Berkeley. Kamen, Deborah. 2009. ‘Servile invective in classical Athens’, Scripta Classica Israelica vol. 28: 43–56. Kamen, Deborah. 2010. ‘A corpus of inscriptions: Representing slave marks in antiquity’, Memoirs of the American Academy in Rome vol. 55: 95–110. Kamen, Deborah. 2012. ‘Manumission, social rebirth, and healing gods in ancient Greece’. In Slaves and Religions in Graeco-Roman Antiquity and Modern Brazil, edited by Stephen Hodkinson and Dick Geary (Newcastle upon Tyne: 174–94). Kamen, Deborah. 2013. Status in Classical Athens (Princeton, NJ). Kamen, Deborah. 2016. ‘Manumission and slave-allowances in classical Athens’, Historia vol. 65.4: 413–26. Kassel, Rudolf and Austin, Colin. 1983–2001. Poetae Comici Graeci. 7 vols. (Berlin). Kästner, Ursula. 1981. ‘Bezeichnungen für Sklaven’. In Soziale Typenbegriffe im alten Griechenland und ihr Fortleben in der Sprachen der Welt 3, edited by Elizabeth Welskopf (Berlin: 282–318). Kellogg, Danielle. 2013. Marathon Fighters and Men of Maple: Ancient Acharnai (Oxford). Kennell, Nigel. 2003. ‘Agreste genus: Helots in Hellenistic Laconia’. In Helots and their Masters in Laconia and Messenia: Histories, Ideologies, Structures, edited by Nino Luraghi and Susan Alcock (Cambridge, MA, and London: 81–105). Kennell, Nigel. 2010. Spartans: A New History (New York, Oxford, and Chichester). Kessler, Karlheinz. 1997. ‘ “Royal roads” and other questions of the Neo-Assyrian communication system’. In Assyria 1995: Proceedings of the 10th Anniversary Symposium of the Neo-Assyrian Text Corpus Project, edited by Simo Parpola and Robert M. Whiting (Helsinki: 129–36). Kiechle, Franz. 1963. Lakonien und Sparta (Munich). King, Philip J. 2009. ‘Slavery in antiquity’. In Exploring the Longue Durée: Essays in Honour of Lawrence E. Stager, edited by J. David Schloen (Winona Lake, IN: 243–50). King, Philip J. and Stager, Lawrence E. 2001. Life in Biblical Israel (Louisville, KY, and London).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

328

Bibliography

Kitto, Humphrey D. F. 1951. The Greeks (London). Klebaner, B. Joseph. 1955. ‘American manumission laws and the responsibility for supporting slaves’, The Virginia Magazine of History and Biography vol. 63.4: 443–53. Kleber, Kristin. 2008. Tempel und Palast: Die Beziehungen zwischen dem König und dem Eanna-Tempel im spätbabylonischen Uruk (Münster). Kleber, Kristin. 2010. ‘The Eanna archive in the years 14–23 Nbk and 2–11 Nbn’. In Aspects of the Economic History of Babylonia in the First Millennium BC, by Michael Jursa (Münster: 540–63). Kleber, Kristin. 2011. ‘Neither Slave nor Free: The Status of Dependents of Babylonian Temple Households’. In Slaves and Households in the Ancient Near East, edited by Laura Culbertson (Chicago: 101–12). Klees, Hans. 1975. Herren und Sklaven. Die Sklaverei im oikonomischen und politischen Schrifttum der Griechen in klassischer Zeit (Wiesbaden). Klees, Hans. 1998. Sklavenleben im klassischen Griechenland (Stuttgart). Kleingunther, Adolf. 1933. Πρῶτος εὑρετής: Untersuchungen zur Geschichte einer Fragestellung. Leipzig. Kleiwegt, Mark. (ed.). 2006. The Faces of Freedom: The Manumission and Emancipation of Slaves in Old World and New World Slavery (Leiden). Koerner, Reinhard. 1993. Inschriftliche Gesetzestexte der frühen griechischen Polis (Cologne, Weimar, and Vienna). Kolchin, Peter. 1993. American Slavery, 1619–1877 (New York). Kolobova, K. M. (n.d.) ‘Der Terminus Oiketes bei Thukydides’ Unpublished translation of Russian original, produced by the Forschungen zur Antiken Sklaverei at the Akademie der Wissenschaften und der Literatur, Mainz. Kock, Theodor. 1880. Comicorum Atticorum Fragmenta (Leipzig). Konstan, David. 2013. ‘Menander’s slaves: The banality of violence’. In Slaves and Slavery in Ancient Greek Comic Drama, edited by Benjamin Akrigg and Robert Tordoff (Cambridge and New York: 144–58). Korres, M. (ed.). 2009. Αττικής οδοί. Αρχαίοι δρόμοι της Αττικής (Athens). Kränzlein, Arnold. 1963. Eigentum und Besitz im griechischen Recht (Berlin). Kristensen, Karen R. 2004a. ‘Gortynian debt bondage: Some new considerations on IC IV 41 IV–VII, 47 and 72 I.56–II.2, X.25–32’, Zeitschrift für Papyrologie und Epigraphik vol. 149: 73–9. Kristensen, Karen R. 2004b. ‘Codification, tradition and innovation in the law code of Gortyn’, DIKE vol. 7: 135–68. Kron, Geof. 1996. ‘Landed and Commercial Wealth at Classical Athens, 500–300 BC’. Diss. Toronto. Kron, Geof. 2005. ‘Anthropometry, physical anthropology, and the reconstruction of ancient health, nutrition, and living standards’, Historia vol. 54.1: 68–83. Kron, Geof. 2008. ‘The much maligned peasant. Comparative perspectives on the productivity of the small farmer in classical antiquity’. In People, Land and Politics. Demographic Developments and the Transformation of Roman Italy, 300 BC–AD 14, edited by Luuk de Ligt and Simon Northwood (Leiden: 71–119). Kron, Geof. 2011. ‘The distribution of wealth at Athens in comparative perspective’, Zeitschrift für Papyrologie und Epigraphik vol. 179: 129–38.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

329

Kron, Geof. 2014. ‘Animal husbandry’. In The Oxford Handbook of Animals in Classical Thought and Life, edited by Gordon L. Campbell (Oxford: 109–35). Kron, Geof. 2017. ‘Roman livestock farming in southern Italy: The case against environmental determinism’. In Rural Communities in a Globalizing Economy: New Perspectives on the Economic Integration of Roman Italy, edited by Tymon de Haas and Gijs Tol (Leiden: 119–34). Krückmann, Oluf. 1933. Neubabylonische Rechts- und Verwaltungstexte (Leipzig). Kudlien, Fridolf. 1968. Die Sklaven in der griechischen Medizin der klassischen und hellenistischen Zeit (Wiesbaden). Kuhrt, Amélie. 1995. The Ancient Near East c. 3000–330 BC. Vol. II (London and New York). Kuhrt, Amélie. 2007. The Persian Empire: A Corpus of Sources from the Achaemenid Period (London and New York). Kuhrt, Amélie. 2014. ‘Even a dog in Babylon is free!’ In The Legacy of Arnaldo Momigliano, edited by Tim Cornell and Oswyn Murray (London: 77–88). Kvist, Kirsten. 2003. ‘Cretan grants of asylia: Violence and protection as interstate relations’, Classica et Mediaevalia vol. 54: 185–222. Kwasman, Theodore and Parpola, Simo (eds.). 1991. Legal Transactions of the Royal Court of Nineveh, Part I: Tiglath-Pileser III through Esarhaddon. State Archives of Assyria Volume VI (Helsinki). Kyrtatas, Dimitris. 2005. ‘The competition of slave and free labour in the classical Greek world’. In Esclavage antique et discriminations socio-culturelles. Actes du XXVIIIe colloque internationale de GIREA (Mytilène, 5–7 décembre 2003), edited by Vasilis I. Anastasiadis and Panagiotis N. Doukellis (Bern: 69–76). Kyrtatas, Dimitris. 2007. ‘The vocabulary of slavery’. In A History of Ancient Greek, from the Beginnings to Late Antiquity, edited by Anastassios F. Christidis (Cambridge: 1056–61). Kyrtatas, Dimitris. 2011. ‘Slavery and economy in the Greek world’. In The Cambridge World History of Slavery. Vol. I: The Ancient Mediterranean World, edited by Keith R. Bradley and Paul A. Cartledge (Cambridge: 91–111). Lagia, Anna. 2015. ‘Diet and the polis: An isotopic study of diet in Athens and Laurion during the Classical, Hellenistic, and Imperial Roman periods’. In Archaeodiet in the Greek World. Dietary Reconstruction from Stable Isotope Analysis, edited by Anastasia Papathanasiou, Michael P. Richards, and Sherry C. Fox (Athens: 119–45). Lakoff, George. 1987. Women, Fire and Dangerous Things: What Categories Reveal about the Mind (Chicago). Lakoff, George. 2008. ‘The neural theory of metaphor’. In The Cambridge Handbook of Metaphor and Thought, edited by Raymond W. Gibbs, Jr (Cambridge: 17–38). Lakoff, George and Johnson, Mark. 1980. Metaphors We Live By (Chicago). Lancel, Serge. 1995. Carthage: A History (Oxford). Lanfranchi, Giovanni B. and Parpola, Simo. 1990. The Correspondence of Sargon II, Part II: Letters from the Northern and Northeastern Provinces. State Archives of Assyria V (Helsinki). Langdon, Merle. 1991. ‘On the farm in classical Attica’, Classical Journal vol. 86.3: 209–13. Langdon, Merle. 2015. ‘Herders’ graffiti’. In Axon: Studies in Honour of Ronald S. Stroud, edited by Angelos Matthiou and Nicholaos Papazarkadas. Vol. 1 (Athens: 49–58).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

330

Bibliography

Larsen, Mogens T. 2017. ‘Between slavery and freedom’. In At the Dawn of History. Ancient Near Eastern Studies in Honour of J. N. Postgate, edited by Yagmur Heffron, Adam Stone, and Martin Worthington. Vol. 1 (Winona Lake, IN: 289–300). Lauffer, Siegfried. 1979. Die Bergwerkssklaven von Laureion. 2nd edn (Wiesbaden). Lavan, Myles. 2013. Slaves to Rome. Paradigms of Empire in Roman Culture (Cambridge). Lawrence, Thomas E. 1935. Seven Pillars of Wisdom (Cambridge). Leão, Delfim and Rhodes, P. J. 2015. The Laws of Solon. A New Edition with Introduction, Translation and Commentary (London and New York). Lee, Spike and Schwartz, Norbert. 2010. ‘Dirty hands and dirty mouths: Embodiment of the moral-purity metaphor is specific to the motor modality involved in moral transgression’, Psychological Science vol. 21: 1423–5. Leichty, Erle. 2011. The Royal Inscriptions of Esarhaddon, King of Assyria (680–669 BC) (Winona Lake, IN). Leigh Fermor, Patrick. 1958. Mani: Travels in the Southern Peloponnnese (London). Lemaire, André. 1977. Inscriptions hébraïques: tome I: les ostraca (Paris). Lemaire, André. 2003. ‘L’Esclave’ In El hombre fenicio: estudios e materiales, edited by José-Ángel Zamora (Rome: 219–22). Lemaire, André. 2015. ‘Esclavage pour dettes et autres formes de travail dépendant au Levant: tradition biblique et épigraphie (Xe–Ve s.)’ In La Main-d’œuvre agricole en Méditerranée archaïque. Statuts et dynamiques économiques, edited by Julien Zurbach (Paris: 67–84). Lemche, Niels P. 1976. ‘The “Hebrew slave”: Comments on the slave law Ex. XXI 2–11’, Vetus Testamentum vol. 25.2: 129–44. Lemche, Niels P. 1979. ‘Andurārum and mīšarum: Comments on the problem of social edicts and their application in the ancient Near East’, Journal of Near Eastern Studies vol. 38.1: 11–22. Lencman, J. A. 1966. Die Sklaverei im mykenischen und homerischen Griechenland (Wiesbaden). Lenski, Noel. 2011. ‘Captivity and slavery among the Saracens in late antiquity (ca. 250–630 CE)’, Antiquité Tardive vol. 19: 237–66. Lenski, Noel. 2014. ‘Captivity among the barbarians and its impact on the fate of the Roman Empire’. In The Cambridge Companion to the Age of Attila, edited by Michael Maas (Cambridge and New York: 230–46). Lenski, Noel. 2018a. ‘Framing the question: What is a slave society?’ In What is a Slave Society? The Practice of Slavery in Global Perspective, edited by Noel Lenski and Catherine Cameron (Cambridge: 15–57). Lenski, Noel. 2018b. ‘Ancient slaveries and modern ideology’. In What is a Slave Society? The Practice of Slavery in Global Perspective, edited by Noel Lenski and Catherine Cameron (Cambridge: 106–47). Lenski, Noel and Cameron, Catherine (eds.). 2018. What is a Slave Society? The Practice of Slavery in Global Perspective (Cambridge). Levine, Baruch A. 1963. ‘The Netînîm’, Journal of Biblical Literature vol. 82.2: 207–12. Levinson, Bernard. 2005. ‘The birth of the lemma: The restrictive reinterpretation of the Covenant Code’s manumission law by the Holiness Code (Leviticus 25:44–46)’, Journal of Biblical Literature vol. 124.4: 617–39.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

331

Levinson, Bernard. 2006a. ‘The “effected object” in contractual legal language: The semantics of “if you purchase a Hebrew slave” (Exod. XXI 2)’, Vetus Testamentum vol. 56.4: 485–504. Levinson, Bernard. 2006b. ‘The manumission of hermeneutics: The slave laws of the Pentateuch as a challenge to contemporary pentateuchal theory’. In Congress Volume Leiden 2004, edited by André Lemaire. Vetus Testamentum Supplements 109 (Leiden: 281–324). Levinson, Bernard. 2010. Legal Revision and Religious Renewal in Ancient Israel (Cambridge). Lévy, Edmond. 1974. ‘Les Esclaves chez Aristophane’. In Actes du colloque 1972 sur l’esclavage (Besançon: 29–46). Lévy, Edmond. 1997. ‘Libres et non-libres dans le Code de Gortyne’. In Esclavage, guerre, économie en Grèce ancienne: hommages à Yvon Garlan, edited by Pierre Brulé and Jacques Oulhen (Rennes: 25–41). Lewis, David Malcolm. 1985. ‘Persians in Herodotus’. In The Greek Historians. Literature and History, edited by Antony E. Raubitschek (Stanford, CA: 101–17). Lewis, David Martin. 2011. ‘Near Eastern slaves in classical Attica and the slave trade with Persian territories’, Classical Quarterly vol. 61.1: 91–113. Lewis, David Martin. 2013. ‘Slave marriages in the laws of Gortyn: A matter of rights?’, Historia vol. 62.4: 390–416. Lewis, David Martin. 2015a. ‘The market for slaves in the fifth- and fourth-century Aegean: Achaemenid Anatolia as a case study’. In The Ancient Greek Economy: Markets, Households and City-States, edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 316–36). Lewis, David Martin. 2015b. ‘Commodities in classical Athens: The evidence of Old Comedy’. In The Ancient Greek Economy: Markets, Households and City-States, edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 381–98). Lewis, David Martin. 2016. ‘Orlando Patterson, property and ancient slavery: The definitional problem revisited’. In On Human Bondage: After Slavery and Social Death, edited by John Bodel and Walter Scheidel (Chichester and Malden, MA): 31–54. Lewis, David Martin. 2017. ‘Making law grip: Inequality, injustice, and legal remedy in Solonian Attica and ancient Israel’. In Violence and Community: Law, Space and Identity in Ancient Eastern Mediterranean World, edited by Ioannis Xydopoulos, Kostas Vlassopoulos, and Eleni Tounta (Abingdon and New York: 28–49). Lewis, David Martin. Forthcoming a. ‘Slavery and manumission’. In The Oxford Handbook of Ancient Greek Law, edited by Mirko Canevaro and Edward M. Harris (Oxford). Lewis, David Martin. Forthcoming b. ‘The Homeric origins of helotage’. In From Homer to Solon: Continuity and Change in Archaic Greece, edited by Mirko Canevaro and Johannes Bernhardt (Leiden). Lewis, David Martin. Forthcoming c. ‘Behavioural economics and economic behaviour in classical Athens’. In Ancient Greek History and Contemporary Social Science, edited by Mirko Canevaro, Andrew Erskine, Benjamin Gray, and Josiah Ober (Edinburgh).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

332

Bibliography

Lewis, David Martin. Forthcoming d. ‘Notes on slave names, ethnicity, and identity in Classical and Hellenistic Greece’, Studia Źródłoznawcze. U Schyłku Starożytności. Lewis, David Martin. Forthcoming e. ‘Knowing where liability lies: Debt bondage and the liability of slaves in Gortynian law’. In The Discovery of the Fact, edited by Clifford Ando. Michigan. Lewis, Juan P. 2012. ‘What’s a Vicarius? Or How “True Meaning” Can Mislead You: Development and Typology of Subowned Slavery in Rome (212 BC–AD 235)’. Diss. Edinburgh. Liddell, Henry G., Scott, Robert, and Jones, Henry S. 1940. A Greek-English Lexicon. 9th edn (Oxford). Link, Stefan. 1994. Das griechische Kreta (Stuttgart). Link, Stefan. 2001. ‘Dolos und Woikeus im Recht von Gortyn’, DIKE vol. 4: 87–112. Link, Stefan. 2002. ‘100 Städte, 100 Verfassungen? Einheitlichkeit und Vielfalt in den griechischen Städten Kretas’, Cretan Studies vol. 7: 149–75. Link, Stefan. 2008. ‘Aristoteles, Ephoros und die “Kretische Verfassung” ’, Gymnasium vol. 115: 469–79. Link, Stefan. 2009. ‘Education and pederasty in Spartan and Cretan Society’. In Sparta: Comparative Approaches, edited by Stephen Hodkinson (Swansea: 89–112). Lipiński, Edward. 1976 ‘L’“Esclave hebreu” ’, Vetus Testamentum vol. 26.1: 120–4. Lipsius, Hermann. 1909. Zum Recht von Gortyns. Abhandlungen der sächs. Gesellschaft der Wissenschaft, phil.-hist. Kl. XXVII no. 11 (Leipzig). Llewellyn-Jones, Lloyd. 2011. ‘Domestic abuse and violence towards women in ancient Greece’. In Sociable Man: Essays on Ancient Greek Social Behaviour in Honour of Nick Fisher, edited by Stephen Lambert (Swansea: 231–66). Lloyd, Geoffrey E. R. 1966. Polarity and Analogy. Two Types of Argumentation in Early Greek Thought (Cambridge). Lloyd, Geoffrey E. R. 1990. Demystifying Mentalities (Cambridge). Lloyd-Jones, Hugh. 1991. Greek in a Cold Climate (Savage, MD). Lods, Adolphe. 1932. Israel, from its Beginnings to the Middle of the Eighth Century (London). Lohmann, Hans. 1992. ‘Agriculture and country life in classical Attica’. In Agriculture in Ancient Greece, edited by Berit Wells (Stockholm: 29–57). Loth, Vincent C. 1995. ‘Pioneers and Perkeniers: The Banda Islands in the 17th century’, Cakalele vol. 6: 13–35. Lotze, Detlef. 1959. ΜΕΤΑΞΥ ΕΛΕΥΘΕΡΩΝ ΚΑΙ ΔΟΥΛΩΝ. Studien zur Rechtsstellung unfreier Landbevölkerungen in Griechenland bis zum 4. Jahrhundert v.Chr. (Berlin). Lotze, Detlef. 1967. ‘Der gentilizisch-personale Grundzug der frühen Gemeinwesen als eine Voraussetzung der griechischen Sklaverei’, Eirene vol. 6: 5–15. Lovejoy, Paul. 1981. ‘Slavery in the Sokoto Caliphate’. In The Ideology of Slavery in Africa, edited by Paul Lovejoy (Beverly Hills, CA: 11–38). Lovejoy, Paul. 2011. Transformations in Slavery. A History of Slavery in Africa. 3rd edn (Cambridge). Lovejoy, Paul. 2014. ‘Pawnship, debt, and “freedom” in Atlantic Africa during the era of the slave trade: A re-assessment’, Journal of African History vol. 55: 1–24.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

333

Lucore, Sandra and Trümper, Monika (eds.). 2013. Greek Baths and Bathing Culture: New Discoveries and Approaches (Leuven). Luraghi, Nino. 2002. ‘Helotic slavery reconsidered’. In Sparta: Beyond the Mirage, edited by Anton Powell and Stephen Hodkinson (London: 229–50). Luraghi, Nino. 2003. ‘The imaginary conquest of the helots’. In Helots and their Masters in Laconia and Messenia: Histories, Ideologies, Structures, edited by Nino Luraghi and Susan Alcock (Cambridge, MA, and London: 109–41). Luraghi, Nino. 2009. ‘The helots: Comparative approaches, ancient and modern’. In Sparta: Comparative Approaches, edited by Stephen Hodkinson (Swansea: 261–304). Lytle, Ephraim. 2013. ‘Farmers into sailors: Ship maintenance, Greek agriculture, and the Athenian monopoly on Kean ruddle (IG II² 1128)’, Greek, Roman and Byzantine Studies vol. 53: 520–50. McCormick, Michael. 2001. Origins of the European Economy: Communications and Commerce AD 300–900 (Cambridge). MacDowell, Douglas M. 1978. The Law in Classical Athens (London). MacDowell, Douglas M. 1986. Spartan Law (Edinburgh). MacDowell, Douglas M. 1993. ‘Foreign birth and Athenian citizenship in Aristophanes’. In Tragedy, Comedy and the Polis, edited by Alan Sommerstein, Stephen Halliwell, John Henderson, and Bernhard Zimmerman (Bari: 359–71). McHugh, Maeve. 2017. The Ancient Greek Farmstead (Oxford). McKeown, Niall. 2007. The Invention of Ancient Slavery? (London). McMahon, April M. S. 1994. Understanding Language Change (Cambridge). Mactoux, Madéline M. 1980. Douleia: esclavage et pratiques discursives dans l’Athènes classique (Paris). Mactoux, Madéline M. 2008. ‘Regards sur la proclamation de l’affranchissement au théâtre à Athènes’ In La Fin du statut servile? Affranchissement, libération, abolition. XXXe colloque du GIREA, Besançon 2005, edited by Antonio Gonzales (Besançon: 437–51). Maffi, Alberto. 1997a. Il diritto di famiglia nel Codice di Gortina (Milan). Maffi, Alberto. 1997b. ‘Processo di libertà e rivendicazione in proprietà dello schiavo a Gortina e ad Atene’. In Symposion 1995: Vorträge zur griechischen und hellenistischen Rechtsgeschichte, edited by Gerhard Thür and Julie VelissaropoulosKarakostas (Cologne: 17–25). Magdalene, F. Rachel, Wells, Bruce, and Wunsch, Cornelia. 2008. ‘Pre-trial negotiations: the case of the run-away slave in Dar. 53’, Iraq vol. 70: 205–13. Magdalene, F. Rachel and Cornelia Wunsch. 2011. ‘Slavery between Judah and Babylon: The exilic experience’. In Slaves and Households in the Near East, edited by Laura Culbertson (Chicago: 113–34). Malinowski, Brolinslaw. 1926. Crime and Custom in Savage Society (London). Manning, Joseph and Oliver, Graham J. 2017. Review of Ober 2015. Journal of Economic History vol. 77.2: 599–601. Martinez, Jean-Luc, Baralis, Alexandre, Mathieux, Néguine, Stoyanov, Totko, and Tonkova, Milena (eds.). 2015. L’Épopée des rois thraces. Des guerres médiques aux invasions celtes, 479–278 av. J.-C. (Paris).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

334

Bibliography

Matilla Vicente, Eduardo. 1977. ‘Surgimiento y desarrollo de la esclavidad cartaginesa’, Hispania antiqua vol. 7: 99–103. Mattila, Raija. 2000. The King’s Magnates: A Study of the Highest Officials of the NeoAssyrian Empire (Helsinki). Mattila, Raija. 2002. Legal Transactions of the Royal Court of Nineveh, Part II: Assurbanipal through Sin-šarru-iškun. State Archives of Assyria Volume XIV (Helsinki). Mayor, Adrienne, Colarusso, John, and Saunders, David. 2014. ‘Making sense of nonsense inscriptions associated with Amazons and Scythians on Athenian vases’, Hesperia vol. 83: 447–93. Mele, Alfonso. 1968. Società e lavoro nei poemi omerici (Naples). Melville, Sarah C. 2010. ‘Kings of Tabal: Politics, competition, and conflict in a contested periphery’. In Rebellions and Peripheries in the Mesopotamian World, edited by Seth Richardson (Winona Lake, IN: 87–109). Mendelsohn, Isaac. 1949. Slavery in the Ancient Near East. A Comparative Study of Slavery in Babylonia, Assyria, Syria and Palestine from the Middle of the Third Millennium to the End of the First Millennium (New York). Meyer, Eduard. 1910. Kleine Schriften zur Geschichtstheorie und zur wirtschaftlichen und politischen Geschichte des Altertums (Halle). Meyer, Elizabeth A. 2010. Metics and the Athenian Phialai Inscriptions: A Study in Athenian Epigraphy and Law (Wiesbaden). Michell, Humfrey. 1952. Sparta (Cambridge). Mielziner, Moses. 1859. Die Verhältnisse der Sklaven bei den alten Hebräern, nach biblischen und talmudischen Quellen dargestellt (Copenhagen). Miles, Richard. 2010. Carthage Must Be Destroyed: The Rise and Fall of an Ancient Civilisation (London). Millett, Paul. 1984. ‘Hesiod and his World’, Proceedings of the Cambridge Philological Society vol. 30: 84–115. Millett, Paul. 1989. ‘Patronage and its avoidance in classical Athens’ In Patronage in Ancient Society, edited by Andrew Wallace-Hadrill (London: 15–48). Mirhady, David. 1996. ‘Torture and rhetoric in Athens’, Journal of Hellenic Studies vol. 116: 119–31. Mirhady, David. 2000. ‘The Athenian rationale for torture’. In Law and Social Status in Classical Athens, edited by Virginia Hunter and Jonathan Edmondson (Oxford: 53–74). Missiou, Anna. 1993. ‘Δοῦλος τοῦ Βασιλέως: The politics of translation’, Classical Quarterly vol. 43.2: 377–91. Momigliano, Arnaldo. 1987. ‘Moses Finley and slavery: A personal note’. In Classical Slavery, edited by Moses I. Finley (London: 1–8). Montanari, Franco. 2014. The Brill Dictionary of Ancient Greek (Leiden). Moore, Ellen W. 1939. Neo-Babylonian Documents in the University of Michigan Collection (Ann Arbor, MI). Morakis, Andreas. 2015. ‘The Gamoroi and the history of archaic Syracuse: A new examination’, Studi di antichità vol. 13: 33–50. Moreno, Alfonso. 2007. Feeding the Democracy. The Athenian Grain Supply in the Fifth and Fourth Centuries BC (Oxford). Moreno, Alfonso and Lape, Susan. 2014. ‘Comedy and the social historian’. In The Cambridge Companion to Greek Comedy, edited by Martin Revermann (Cambridge: 336–70).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

335

Moreno García, Juan Carlos. 2015. ‘L’Évolution des statuts de la main-d’œuvre rurale en Égypte de la fin du Nouvel Empire à l’époque saïte (c.1150–525 a.C.)’. In La Main-d’œuvre agricole en Méditerranée archaïque: statuts et dynamiques économiques, edited by Julien Zurbach (Paris: 15–48). Morris, Ellen. 2014. ‘Mitanni enslaved: Prisoners of war, pride, and productivity in a new imperial regime’. In Creativity and Innovation in the Reign of Hatshepsut, edited by José M. Galán, Betsy M. Bryan, and Peter F. Dorman (Chicago: 361–79). Morris, Ian. 1986. ‘The use and abuse of Homer’, Classical Antiquity vol. 5: 81–138. Morris, Ian. 1987. Burial and Ancient Society (Cambridge). Morris, Ian. 2002. ‘Hard surfaces’. In Money, Labour and Land: Approaches to the Economies of Ancient Greece, edited by Paul A. Cartledge, Edward E. Cohen, and Lin Foxhall (London: 8–43). Morris, Ian, Saller, Richard, and Scheidel, Walter. 2007. ‘Introduction’. In The Cambridge Economic History of the Greco-Roman World, edited by Ian Morris, Richard Saller, and Walter Scheidel (Cambridge: 1–12). Morrow, Glen. 1939. Plato’s Law of Slavery in its Relation to Greek Law (Urbana, IL). Mossé, Claude. 1969. The Ancient World at Work. Translated by Janet Lloyd (London). Moyal, Yigal and Faust, Avraham. 2015. ‘Jerusalem’s hinterland in the eighth–seventh centuries BCE: Towns, villages, farmsteads, and royal estates’, Palestine Exploration Quarterly vol. 147.4: 283–98. Muhs, Brian. 2016. The Ancient Egyptian Economy, 3000–30 BCE (Cambridge and New York). Müller, Gerfrid G. W. 2004. ‘Zur Entwicklung von Preisen und Wirtschaft in Assyrien im 7. Jh. v. Chr’. In Von Sumer nach Ebla und zurück: Festschrift Giovanni Pettinato, edited by Hartmut Waetzold (Heidelberg: 185–210). Müller, Karl W. L. (1841–70) Fragmenta Historicorum Graecorum (Paris). Murray, Oswyn. 1993. Early Greece. 2nd edn (London). Na’aman, Nadav. 1986. ‘Habiru and Hebrews: The transfer of a social term to the literary sphere’, Journal of Near Eastern Studies vol. 45.4: 271–88. Na’aman, Nadav. 2007. ‘When and how did Jerusalem become a great city? The rise of Jerusalem as Judah’s premier city in the eighth–seventh centuries BCE’, Bulletin of the American Schools of Oriental Research vol. 347: 21–56. Nadel, Benjamin. 1976. ‘Slavery and related forms of labor on the north shore of the Euxine in antiquity’. In Actes du colloque 1973 sur l’esclavage. Besançon 2–3 mai 1973 (Besançon: 197–233). Nafissi, Massimo. 2009. ‘Sparta’. In A Companion to Archaic Greece, edited by Kurt Raaflaub and Hans van Wees (Malden, MA, Oxford, and Chichester: 117–37). Nafissi, Massimo. 2015. ‘Krypteiai Spartane’. In Los espacios de la esclavitud y la dependencía desde la antigüedad, edited by Alejandro Beltrán, Inés Sastre, and Miriam Valdés (Besançon: 201–29). Nafissi, Mohammed. 2005. Ancient Athens and Modern Ideology: Value, Theory and Evidence in Historical Sciences: Max Weber, Karl Polanyi and Moses Finley (London). Naiden, Fred. 2006. Ancient Supplication (Oxford). Naiden, Fred and Talbert, Richard (eds.). 2014. Moses Finley in America: The Making of an Ancient Historian. Special issue of the American Journal of Philology (vol. 135.2) (Baltimore, MD).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

336

Bibliography

Nauck, August. 1889. Tragicorum Graecorum Fragmenta. 2nd edn (Leipzig). Ndoye, Malick. 2010. Groupes sociaux et idéologie du travail dans les mondes homérique et hésiodique (Besançon). Nicholas, Barry. 1962. An Introduction to Roman Law (Oxford). Nieboer, Herman. 1900. Slavery as an Industrial System (The Hague). Ober, Josiah. 2010. ‘Wealthy Hellas’, Transactions of the American Philological Association vol. 140: 241–86. Ober, Josiah. 2015. The Rise and Fall of Classical Greece (Princeton, NJ, and Oxford). O’Connor, Stephen. 2013. ‘The daily grain consumption of classical Greek soldiers and sailors’, Chiron vol. 43: 327–56. Oded, Bustenay. 1979. Mass Deportations and Deportees in the Neo-Assyrian Empire (Wiesbaden). Oelsner, Joachim. 2014. ‘Spätbabylonische Sklavenpreise’. In Studies in Economic and Social History of the Ancient Near East in Memory of Péter Vargyas, edited by Zoltán Csabai (Budapest: 683–99). Oliva, Pavel. 1971. Sparta and her Social Problems (Prague). Osborne, Robin. 1985. Demos: The Discovery of Classical Attika (Cambridge). Osborne, Robin. 1988. ‘Social and economic implications of the leasing of land in Classical and Hellenistic Greece’, Chiron vol. 18: 279–323. Osborne, Robin. 1991. ‘Pride and prejudice, sense and subsistence: Exchange and society in the Greek city’. In City and Country in the Ancient World, edited by John Rich and Andrew Wallace-Hadrill (London and New York: 119–46). Osborne, Robin. 1995. ‘The economics and politics of slavery at Athens’. In The Greek World, edited by Anton Powell (London: 27–43). Osborne, Robin. 2000. ‘Religion, imperial politics, and the offering of freedom to slaves’. In Law and Social Status in Classical Athens, edited by Virginia Hunter and Jonathan Edmondson (Oxford: 75–92). Osborne, Robin. 2007. ‘Archaic Greece’. In The Cambridge Economic History of the Greco-Roman World, edited by Walter Scheidel, Ian Morris, and Richard Saller (Cambridge: 277–301). Otto, Eckart. 2013. ‘The history of the legal-religious hermeneutics of the Book of Deuteronomy from the Assyrian to the Hellenistic period’. In Law and Religion in the Eastern Mediterranean, edited by Anselm Hagedorn (Oxford: 211–50). Palais, James B. 1984. ‘Slavery and slave society in the Koryo period’, Journal of Korean Studies vol. 5: 173–90. Papazoglu, Fanula. 1995. ‘Sur la condition des hilotes affranchis’, Historia vol. 44.3: 370–5. Papazoglu, Fanula. 1997. LAOI et PAROIKOI: Recherches sur la structure de la société hellénistique (Belgrade). Paradiso, Annalisa. 2004. ‘The logic of terror: Thucydides, Spartan duplicity, and an improbable massacre’. In Spartan Society, edited by Thomas Figueira (Swansea: 179–98). Paradiso, Annalisa. 2008. ‘Politiques de l’affranchissement chez Thucydide’. In La Fin du statut servile? Affranchissement, libération, abolition. XXXe colloque du GIREA, Besançon 2005, edited by Antonio Gonzales (Besançon: 65–76). Paradiso, Annalisa. 2013. ‘Nannies “bought” in Laconia: Slave or hired? A note to Plutarch, Lycurgus 16,5’, Eirene vol. 49: 48–53.

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

337

Parpola, Simo (ed.). 1993. Letters from Assyrian and Babylonian Scholars. State Archives of Assyria Volume X (Helsinki). Parpola, Simo. 2007. ‘The Neo-Assyrian ruling class’. In Studien zu Ritual und Sozialgeschichte im Alten Orient/Studies on Ritual and Society in the Ancient Near East. Tartuer Symposion 1998–2004, edited by Thomas Kämmerer (Berlin and New York: 257–74). Patterson, K. David. 1988. ‘Epidemics, Famines, and Population in the Cape Verde Islands, 1580–1900’, International Journal of African Historical Studies vol. 21.2: 291–313. Patterson, Orlando. 1982. Slavery and Social Death (Cambridge, MA). Patterson, Orlando. 1991. Freedom in the Making of Western Culture (New York). Patterson, Orlando. 2008. ‘Slavery, gender, and work in the pre-modern world and early Greece: A cross-cultural analysis’. In Slave Systems, Ancient and Modern, edited by Enrico Dal Lago and Constantina Katsari (Cambridge: 32–69). Patterson, Orlando. 2012. ‘Trafficking, gender and slavery: Past and present’. In The Legal Understanding of Slavery: From the Historical to the Contemporary, edited by Jean Allain (Oxford: 322–59). Patterson, Orlando. 2016. ‘Revisiting slavery, property, and social death’. In On Human Bondage: After Slavery and Social Death, edited by John Bodel and Walter Scheidel (Chichester and Malden, MA: 265–96). Pavlovskaja, Aleksandra. 1979. ‘Les Recherches des historiens soviétiques sur l’esclavage en Grèce et dans le monde hellénistique’. In Actes du colloque sur l’esclavage. VIIIe colloque du GIREA, Nieborów 1975, edited by Iza Bieżuńska-Małowist and Jerzy Kolendo (Warsaw: 23–33). Pečírka, Jan. 1964. ‘Die sowjetischen Diskussionen über die asiatische Produktionsweise und über die Sklavenhalterformation’, Eirene vol. 3: 147–69. Penner, James E. 2012. ‘The concept of property and the concept of slavery’. In The Legal Understanding of Slavery: From the Historical to the Contemporary, edited by Jean Allain (Oxford: 242–52). Perlman, Paula. 1992. ‘One hundred-citied Crete and the “Cretan politeia” ’, Classical Philology vol. 87: 193–205. Perlman, Paula. 2004a. ‘Crete’. In An Inventory of Archaic and Classical Poleis, edited by Mogens H. Hansen and Thomas H. Nielsen (Oxford: 1144–95). Perlman, Paula. 2004b. ‘Tinker, tailor, soldier, sailor: The economies of archaic Eleutherna, Crete’, Classical Antiquity vol. 23.1: 95–137. Petit, Paul. 1970. ‘L’Esclavage antique dans l’historiographie soviétique’. In Actes du colloque d’histoire sociale 1970. Ier colloque du GIREA (Besançon: 9–27). Picard, Gilbert and Picard, Colette. 1961. Life in Carthage (Norwich). Pierrot, Antoine. 2015. ‘Who were the Eupatrids in archaic Athens?’. In ‘Aristocracy’ in Antiquity: Redefining Greek and Roman Elites, edited by Nick Fisher and Hans van Wees (Swansea: 147–68). Pipili, Maria. 2000. ‘Wearing an Other hat: Workmen in town and country’. In Not the Classical Ideal. Athens and the Construction of the Other in Greek Art, edited by Beth Cohen (Leiden: 153–79). Porten, Bezalel. 1996. The Elephantine Papyri in English. Three Millennia of CrossCultural Continuity and Exchange (Leiden).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

338

Bibliography

Porten, Bezalel and Yardeni, Ada. 1987–99. Textbook of Aramaic Documents from Ancient Egypt. 4 vols. (Jerusalem). Pospisil, Leopold. 1971. Anthropology of Law: A Comparative Theory (New York). Postgate, J. Nicholas. 1974. ‘Some remarks on conditions in the Assyrian countryside’, Journal of the Economic and Social History of the Orient vol. 17.3: 225–43. Postgate, J. Nicholas. 1976. Fifty Neo-Assyrian Legal Documents (Warminster). Postgate, J. Nicholas. 1979. ‘The economic structure of the Assyrian empire’. In Power and Propaganda: A Symposium on Ancient Empires, edited by Mogens T. Larsen (Copenhagen: 193–221). Postgate, J. Nicholas. 1987. ‘Employer, employee and employment in the NeoAssyrian Empire’. In Labor in the Ancient Near East, edited by Marvin A. Powell (New Haven, CT: 257–70). Postgate, J. Nicholas. 1989. ‘Ownership and exploitation of land in Assyria in the 1st Millennium B.C.’. In Reflets des deux fleuves: volume de mélanges offerts à André Finet, edited by Marc Lebeau and Philippe Talon (Leuven: 141–52). Postgate, J. Nicholas. 2007. The Land of Assur and the Yoke of Assur. Studies on Assyria: 1971–2005 (Oxford). Pritchard, David. 2012. ‘Aristophanes and de Ste. Croix: The value of Old Comedy as evidence for Athenian popular culture’, Antichthon vol. 46: 14–51. Pritchett, W. Kendrick and Pippin, Anne. 1956. ‘The Attic Stelai, part II’, Hesperia vol. 25: 178–317. Probert, Philomen. 2015. Early Greek Relative Clauses (Oxford). Psoma, Selene. 2015. ‘Choosing and changing weight standards in the Greek world during the archaic and classical periods’. In The Ancient Greek Economy: Markets, Households and City-States, edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 90–115). Raaflaub, Kurt. 2004. The Discovery of Freedom in Ancient Greece (Chicago and London). Rabinowitz, Adam. 2014. ‘Drinkers, hosts, or fighters? Masculine identities in preclassical Crete’. In Cultural Practices and Material Culture in Archaic and Classical Crete, edited by Gunnar Seelentag and Oliver Pilz (Berlin and Boston, MA: 91–120). Radner, Karen. 1997. Die neuassyrischen Privatrechtsurkunden als Quelle für Mensch und Umwelt (Helsinki). Radner, Karen. 2001. ‘The Neo-Assyrian period’. In Security for Debt in Ancient Near Eastern Law, edited by Richard Jasnow and Raymond Westbrook (Leiden: 265–88). Radner, Karen. 2002. Die neuassyrischen Texte aus Tall Šēḫ Ḥ amad (Berlin). Radner, Karen. 2007. ‘Hired labour in the Neo-Assyrian Empire’, State Archives of Assyria Bulletin vol. 16: 185–226. Radner, Karen. 2008. Review of Galil 2007. Zeitschrift für Assyriologie vol. 98: 295–6. Radner, Karen. 2015a. ‘A Neo-Assyrian slave sale contract of 725 BC from the Peshdar Plain and the location of the palace herald’s province’, Zeitschrift für Assyriologie vol. 105.2: 192–7. Radner, Karen. 2015b. ‘Hired labour in the Neo-Assyrian Empire’. In Labor in the Ancient World, edited by Piotr Steinkeller and Michael Hudson (Dresden: 329–43). Radner, Karen. 2017a. ‘Neo-Assyrian tablets’ In Assyrian Archival Texts in the Schøyen Collection and Other Documents from North Mesopotamia and Syria. Manuscripts

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

339

in the Schøyen Collection, edited by Andrew George, Thomas Hertel, Jaume LlopRaduà, Karen Radner, and Wilfred H. van Soldt (Bethesda, MD: 77–94). Radner, Karen. 2017b. ‘Economy, society, and daily life in the Neo-Assyrian period’. In A Companion to Assyria, edited by Eckart Frahm (Malden, MA: 209–28). Raepsaet, Georges. 2002. Attelages et techniques de transport dans le monde grécoromain (Brussels). Ragen, Asher. 2006. ‘The Neo-Babylonian Širku: A Social History’. Diss. Harvard. Rainey, Anson. 1967. ‘The Samaria ostraca in the light of fresh evidence’, Palestine Exploration Quarterly vol. 99: 32–41. Rainey, Anson. 1970. ‘Compulsory labour gangs in ancient Israel’, Israel Exploration Journal vol. 20: 191–202. Rainey, Anson. 2003. ‘Aspects of life in ancient Israel’. In Life and Culture in the Ancient Near East, edited by Richard E. Averbeck, Mark W. Chavalas, and David B. Weisberg (Bethesda, MD: 253–68). Randall, Richard H. 1953. ‘The Erechtheum workmen’, American Journal of Arcaeology vol. 57.3: 199–210. Rankine, Patrice. 2011. ‘Odysseus as slave: The ritual of domination and social death in Homeric society’. In Reading Ancient Slavery, edited by Richard Alston, Edith Hall, and Laura Proffit (London and New York: 34–50). Rauh, Nicholas. 1998. ‘Who were the Cilician pirates?’. In Res Maritimae: Cyprus and the Eastern Mediterranean from Prehistory to Late Antiquity, edited by Stuart Swiny et al. (Atlanta, GA: 263–83). Rhodes, P. J. 2004. ‘Keeping to the point’. In The Law and the Courts in Ancient Greece, edited by Edward M. Harris and Lene Rubinstein (London: 137–58). Rhodes, P. J. and Osborne, Robin. 2003. Greek Historical Inscriptions 404–323 BC (Oxford). Richardson, Seth. 2016. ‘Getting confident: The Assyrian development of elite recognition ethics’. In Cosmopolitanism and Empire, edited by Myles Lavan, Richard E. Payne, and John Weisweiler (Oxford: 29–64). Richelle, Matthieu. 2016. ‘Elusive scrolls: Could any Hebrew literature have been written prior to the eighth century BCE?’, Vetus Testamentum vol. 66: 1–39. Ridder, Jacob J. de. 2017. ‘Slavery in Old Assyrian documents’. In Kültepe International Meetings, Vol. II: Movement, Resources, Interaction, edited by Fikri Kulakoğlu and Gojko Barjamovic (Antwerp: 49–61). Rihll, Tracey E. 1991. ‘EKTHMOPOI: Partners in crime?’, Journal of Hellenic Studies vol. 111: 101–27. Rihll, Tracey E. 1996. ‘The origins and establishment of Greek slavery’. In Serfdom and Slavery: Studies in Legal Bondage, edited by Michael L. Bush (London: 89–111). Rihll, Tracey E. 2010. ‘Skilled slaves and the economy: The silver mines of the Laurion’ In Antike Sklaverei: Rückblick und Ausblick. Neue Beiträge zur Forschungsgeschichte und zur Erschließung der archäologischen Zeugnisse, edited by Heinz Heinen (Stuttgart: 203–20). Rihll, Tracey E. 2011. ‘Classical Athens’ In The Cambridge World History of Slavery: Vol. I: The Ancient Mediterranean World, edited by Keith R. Bradley and Paul A. Cartledge (Cambridge: 48–73). Robertson, Bruce. 2008. ‘The slave names of IG I³ 1032 and the Athenian ideology of slavery’. In Epigraphy and the Greek Historian, edited by Craig Cooper (Toronto: 79–116).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

340

Bibliography

Rocca, Francesca. Forthcoming. ‘Hera Eleutheria, the Water of Freedom and the Slaves’. In Acts of the International Conference on the History of Classical Antiquity. Interdisciplinary Approaches, University of Coimbra, 20th–22nd May 2015. Rollinger, Robert. 2009. ‘Near Eastern perspectives on the Greeks.’ In The Oxford Handbook of Hellenic Studies, edited by George Boys-Stones, Barbara Graziosi, and Phiroze Vasunia (Oxford: 32–47). Roobaert, Arlette. 1977. ‘Le Danger hilote?’ Ktema vol. 2: 141–55. Rose, Peter. 2012. Class in Archaic Greece (Cambridge and New York). Rose, Valentin. 1886. Aristotelis Qui Ferebantur Librorum Fragmenta (Leipzig). Rosivach, Vincent. 1993. ‘Agricultural slavery in the Northern Colonies and in classical Athens: Some comparisons’, Comparative Studies in Society and History vol. 35: 551–67. Rosivach, Vincent. 1999. ‘Enslaving barbaroi and the Athenian ideology of slavery’, Historia vol. 48.2: 129–57. Ross, Brandon D. 2011. ‘Krypteia: A form of ancient guerilla warfare’, Grand Valley Journal of History vol. 1.2: 1–10. Roth, Martha T. 1991. ‘The dowries of the women of the Itti-Marduk-balat ụ family’, Journal of the American Oriental Society vol. 111.1: 19–37. Roth, Martha T. 1995. Law Collections from Mesopotamia and Asia Minor (Atlanta, GA). Rotman, Youval. 2004. Les Esclaves et l’esclavage: de la Méditerranée antique à la Méditerranée médievale VIe–XIe siècles (Paris). Roy, James. 1994. Review of Fisher 1993. Classical Review vol. 45.1: 190. Roy, James. 1996. ‘The countryside in classical Greek drama, and isolated farms in dramatic landscapes’. In Human Landscapes in Classical Antiquity: Environment and Culture, edited by Graham Shipley and John Salmon (London and New York: 98–118). Roy, James. 2005. Review of Jones 2004. Hermathena vol. 179: 208–11. Roy, James. 2012. ‘Le Bérger, un personnage en marge de la société grecque classique’. In Dipendenza ed emarginazione nel mondo antico e moderno. Atti del XXXIII Convegno Internazionale G.I.R.E.A, edited by Francesca R. Merola (Rome: 219–25). Ruffing, Kai and Drexhage, Hans-Joachim. 2008. ‘Antike Sklavenpreise’. In Antike Lebenswelten: Konstanz, Wandel, Wirkungsmacht: Festschrift für Ingomar Weiler zum 70. Geburtstag, edited by Peter Mauritsch et al. (Wiesbaden: 32151). Runciman, Steven. 1954. A History of the Crusades. Volume III: The Kingdom of Acre (Cambridge). Saba, Sara. 2012. The Astynomoi Law from Pergamon. A New Commentary (Mainz). Sahlins, Marshall. 1972. Stone Age Economics (New York). Said, Edward. 1978. Orientalism (New York and London). Sallares, Robert. 1991. The Ecology of the Ancient Greek World (New York). Samuel, Alan. 1965. ‘The role of paramonê clauses in ancient documents’, Journal of Juristic Papyrology vol. 15: 221–311. San Nicolò, Mariano and Ungnad, Arthur. 1929–35. Neubabylonische Rechts- und Verwaltungsurkunden übersetzt und erläutert (Leipzig). Sargent, Rachel. 1924. The Size of the Slave Population at Athens during the Fifth and Fourth Centuries before Christ (Urbana, IL).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

341

Scheidel, Walter. 2002. ‘The Hireling and the slave: A transatlantic perspective’. In Money, Labour and Land: Approaches to the Economies of Ancient Greece, edited by Paul A. Cartledge, Edward E. Cohen, and Lin Foxhall (London and New York: 175–84). Scheidel, Walter. 2005. ‘Real slave prices and the relative cost of slave labour in the Greco-Roman world’, Ancient Society vol. 35: 1–17. Scheidel, Walter. 2008. ‘The comparative economics of slavery in the Greco-Roman world’. In Slave Systems, Ancient and Modern, edited by Enrico Dal Lago and Constantina Katsari (Cambridge: 105–26). Scheidel, Walter. 2011. ‘The Roman slave supply’. In The Cambridge World History of Slavery, Vol. 1: The Ancient Mediterranean World, edited by Keith R. Bradley and Paul A. Cartledge (Cambridge: 287–310). Scheidel, Walter. 2017. ‘Slavery and forced labor in early China and the Roman world’. In Eurasian Empires in Antiquity and the Early Middle Ages: Contacts and Exchange between the Graeco-Roman World, Inner Asia and China, edited by Hyun Jim Kim, Frederik J. Vervaet, and Selim F. Adalı (Cambridge: 133–50). Schmidt, Martin. 2006. ‘Die Welt des Eumaios’. In Geschichte und Fiktion in der homerischen Odyssee, edited by Andreas Luther (Munich: 117–38). Schmitz, Winfried. 2011. ‘Der Verkauf einer Sklavenfamilie’, Zeitschrift für Papyrologie und Epigraphik vol. 179: 54–6. Schmitz, Winfried. 2012. ‘Sklavenfamilien im archaischen und klassischen Griechenland’. In Sklavenkinder. Schicksale zwischen Zuneigung und Ausbeutung in der Antike und im interkulturellen Vergleich, edited by Heinz Heinen (Stuttgart: 63–102). Schmitz, Winfried. 2014. ‘Sklavenaufseher der Heloten?’, Historia vol. 63.3: 293–300. Scholl, Reinhold. 1990. Corpus der ptolemäischen Sklaventexte (Stuttgart). Scholten, Joseph. 2000. The Politics of Plunder: Aitolians and their Koinon in the Early Hellenistic Era, 279–217 B.C. (Berkeley, Los Angeles, and London). Schumacher, Leonard. 2001. Sklaverei in der Antike: Alltag und Schiksal der Unfreien (Munich). Scott, John and Marshall, Gordon (eds.). 2005. Oxford Dictionary of Sociology. 3rd edn (Oxford). Sears, Matthew. 2013. Athens, Thrace, and the Shaping of Athenian Leadership (Cambridge). Seelentag, Gunnar. 2015. Das archaische Kreta. Institutionalisierung im frühen Griechenland (Berlin). Seelentag, Gunnar. 2017. ‘Der Xenios Kosmos und die Freigelassenen von Latosion. Eine archaische Inschrift aus dem kretischen Gortyn (IC 4.78)’, Hermes vol. 145.3: 275–87. Seelentag, Gunnar and Pilz, Oliver (eds.). 2014. Cultural Practices and Material Culture in Archaic and Classical Crete (Berlin and Boston, MA). Sekunda, Nicholas. 1985. ‘Achaemenid colonization in Lydia’, Revue des Études Anciennes vol. 87: 7–29. Sekunda, Nicholas. 1988. ‘Persian settlement in Hellespontine Phrygia’. In Achaemenid History III: Method and Theory, edited by Amélie Kuhrt and Heleen Sancisi-Weerdenburg (Leiden: 175–96).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

342

Bibliography

Sekunda, Nicholas. 1991. ‘Achaemenid settlement in Caria, Lycia and Greater Phrygia’. In Achaemenid History VI. Asia Minor and Egypt: Old Cultures in a New Empire, edited by Heleen Sancisi-Weerdenburg and Amélie Kuhrt (Leiden: 83–143). Seidl, Erwin. 1968. Ägyptische Rechtsgeschichte der Saiten- und Perserzeit (Glückstadt). Seri, Andrea. 2013. The House of Prisoners: Slavery and State in Uruk during the Revolt against Samsu-iluna (Boston, MA). Shipton, Kirsty. 1997. ‘The private banks in fourth century B.C. Athens: A reappraisal’, Classical Quarterly vol. 47: 396–422. Silver, Morris. 2006. ‘Slave versus free hired workers in ancient Greece’, Historia vol. 55.3: 257–63. Simon, Zsolt. 2012. ‘Where is the land of Sura of the hieroglyphic Luwian inscription Karkamiš A4b and why were Cappadocians called Syrians by the Greeks?’, Altorientalischen Forschungen vol. 39.1: 167–80. Simon, Zsolt. 2017. ‘The northern border of Tabal’. In Hittology Today: Studies on Hittite and Neo-Hittite Anatolia in Honor of Emmanuel Laroche’s 100th Birthday, edited by Alice Mouton (Istanbul: 201–11). Sinclair, R. K. 1988. Democracy and Participation in Athens (Cambridge). Snell, Daniel C. 2001. Flight and Freedom in the Ancient Near East (Leiden). Snell, Daniel C. 2011. ‘Slavery in the ancient Near East’. In The Cambridge World History of Slavery. Vol. I: The Ancient Mediterranean World, edited by Keith R. Bradley and Paul A. Cartledge (Cambridge: 4–21). Snodgrass, Antony. 1980. Archaic Greece: The Age of Experiment (London). Sobak, Robert. 2015. ‘Sokrates among the shoemakers’, Hesperia vol. 84: 669–712. Sokolowski, Franciszek. 1962. Lois sacrées des cités grecques: supplément (Paris). Sokolowski, Franciszek. 1969. Lois sacrées des cités grecques (Paris). Sommerstein, Alan. 1980. ‘Notes on Aristophanes’ Knights’, Classical Quarterly vol. 30.1: 46–56. Spittler, Gerd. 2012. ‘Anthropologie/Ethnologie’. In Handwörterbuch der antiken Sklaverei, edited by Heinz Heinen. CD-ROM-Lieferung I–IV (Stuttgart). Stager, Lawrence. 1983. ‘The finest oil in Samaria’, Journal of Semitic Studies vol. 28: 241–5. Starr, Chester G. 1958. ‘An overdose of slavery’, Journal of Economic History vol. 18.1: 17–32. Starr, Chester G. 1965. ‘The credibility of early Spartan history’, Historia vol. 14: 257–72. Stewart, Roberta. 2012. Plautus and Roman Slavery (Malden, MA, Oxford, and Chichester). Stolper, Matthew. 1984. ‘The Neo-Babylonian text from the Persepolis fortification’, Journal of Near Eastern Studies vol. 43.4: 299–310. Stolper, Matthew. 1985. Entrepreneurs and Empire. The Murašu Archive, the Murašu Firm, and Persian Rule in Babylonia (Istanbul). Stolper, Matthew. 1989. ‘The registration and taxation of slave sales in late Achaemenid Babylonia’, Zeitschrift für Assyriologie vol. 79: 80–101. Stolper, Matthew. 2014. ‘Case in point: The Perseopolis Fortification Archive’. In Archaeologies of Text: Archaeology, Technology, and Ethics, edited by Matthew T. Rutz and Morag M. Kersel (Oxford and Philadelphia, PA: 14–30). Storey, Ian C. 2003. Eupolis: Poet of Old Comedy (Oxford). Strassmeier, Johann N. 1889a. Inschriften von Nabuchodonosor, König von Babylon (604–561 v.Chr.) (Leipzig).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

343

Strassmeier, Johann N. 1889b. Inschriften von Nabonidus, König von Babylon (555–538 v.Chr.) (Leipzig). Strassmeier, Johann N. 1890a. Inschriften von Cambyses, König von Babylon (529–521 v.Chr.) (Leipzig). Strassmeier, Johann N. 1890b. Inschriften von Cyrus, König von Babylon (538–529 v.Chr.) (Leipzig). Strassmeier, Johann N. 1897. Inschriften von Darius, König von Babylon (521–485 v.Chr.) (Leipzig). Stroud, Ronald. 1974. ‘An Athenian law on silver coinage’, Hesperia vol. 43: 157–93. Sulzberger, Mayer. 1923. The Status of Labor in Ancient Israel (Philadelphia, PA). Swain, Simon. 1996. Hellenism and Empire. Language, Classicism and Power in the Greek World, AD 50–250 (Oxford). Swoboda, Heinrich. 1905. ‘Beiträge zur griechischen Rechtsgeschichte 2. Über die altgriechische Schuldknechtschaft’, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Romanistische Abteilung vol. 26: 190–280. Tačeva, M. 1996. ‘Entre la liberté et l’ésclavage en Thrace ancienne’. In Captius i esclaus a l’antiguitat i al món modern. Actes del XIX colloqui internacional del GIREA. Organizat pel departement de ciències historiques i teoria de les arts, Universitat de les illes Balears. Palma de Mallorca, 2–5 octobre 1991, edited by Maria Luisa Sánchez León and Gonçal López Nadal (Naples: 21–5). Talbert, Richard. 1989) ‘The role of the helots in the class struggle at Sparta’, Historia vol. 38: 22–40. Tamiolaki, Mélina. 2010. Liberté et esclavage chez les historiens grecs classiques (Paris). Tedeschi, Gennaro. 1986. ‘Il canto di Hybrias il cretese, un esempio di poesia conviviale’, Quaderni di filologia classica vol. 5: 53–74. Tenney, Jonathan S. 2011. Life at the Bottom of Babylonian Society. Servile Laborers at Nippur in the 14th and 13th Centuries B.C. (Leiden and Boston, MA). Testart, Alain, Jeunesse, Christian, Baray, Luc, and Boulestin, Bruno. 2010. ‘Les Esclaves des tombes néolithiques’, Pour la Science vol. 396: 74–80. Thalheim, Theodor. 1913. Lysiae Orationes. Leipzig. Thalmann, William G. 1998a. The Swineherd and the Bow: Representations of Class in the Odyssey (Ithaca, NY, and London). Thalmann, William G. 1998b. ‘Female slaves in the Odyssey’. In Women and Slaves in Greco-Roman Culture: Differential Equations, edited by Sandra Joshel and Sheila Murnaghan (London: 22–34). Thommen, Lukas. 2014. Die Wirtschaft Spartas (Stuttgart). Thompson, Dorothy. 2011. ‘Slavery in the Hellenistic world’. In The Cambridge World History of Slavery, Volume I: The Ancient Mediterranean World, edited by Keith R. Bradley and Paul A. Cartledge (Cambridge: 194–213). Thompson, Edward P. 1963. The Making of the English Working Class (London). Thompson, Thomas L. 1992. Early History of the Israelite People: From the Written and Archaeological Sources (Leiden). Thompson, Wesley E. 1982. ‘The Athenian entrepreneur’, L’Antiquité Classique vol. 51: 53–85. Thonemann, Peter. 2013. ‘Phrygia: An anarchist history, 950 BC–AD 100’. In Roman Phrygia: Culture and Society, edited by Peter Thonemann (Cambridge: 1–40).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

344

Bibliography

Thorner, Daniel. 1966. ‘Marx on India and the Asiatic mode of production’, Contributions to Indian Sociology vol. 9: 33–66. Thorner, Daniel. 1971. ‘Peasant economy as a category in economic history’. In Peasants and Peasant Societies, edited by Teodor Shanin (Harmondsworth: 202–18). Thür, Gerhard. 1977. Beweisführung vor den Schwurgerichtshöfen Athens: Die Proklesis zur Basanos (Vienna). Thür, Gerhard. 1996. ‘Reply to DC Mirhady: Torture and rhetoric in Athens’, Journal of Hellenic Studies vol. 116: 132–4. Todd, Stephen C. 1993. The Shape of Athenian Law (Oxford). Todd, Stephen C. 2007. A Commentary on Lysias, Speeches 1–11 (Oxford). Tolini, Gauthier. (n.d.) ‘Between Babylon and Kiš. The economic activities of Isḫunnatu, a slave woman of the Egibi family (sixth century BC)’. Unpublished paper presented at The Economic Role of Women in the Public Sphere in Mesopotamia, Chuo University, Tokyo, 24 and 25 June 2013. Tompkins, Daniel. 2014a. ‘What happened in Stockholm?’, Hyperboreus vol. 20: 436–52. Tompkins, Daniel. 2014b. ‘La Formation de Moses Finley d’après les documents américains, 1932–1955’, Anabases vol. 19: 111–29. Tordoff, Robert. 2013. ‘Introduction: Slaves and slavery in ancient Greek comedy’. In Slaves and Slavery in Ancient Greek Comedy, edited by Benjamin Akrigg and Robert Tordoff (Cambridge: 1–62). Toynbee, Arnold. 1969. Some Problems of Greek History (Oxford). Trevelyan, George M. 1952. History of England. 3rd edn (London and Colchester). Trümper, Monika et al. 2013. ‘Catalog of Greek public baths’. In Greek Baths and Bathing Culture: New Discoveries and Approaches, edited by Sandra Lucore and Monika Trümper (Leuven: 265–333). Tsakirgis, Barbara. 2015. ‘Whole cloth: Exploring the question of self-sufficiency through the evidence for textile manufacture and purchase in Greek houses’. In The Ancient Greek Economy: Markets, Households and City-States, edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 166–86). Tsiafakis, Despoina. 2015. ‘Thracian tattoos’. In Bodies in Transition: Dissolving the Boundaries of Embodied Knowledge, edited by D. Boschung, A. Shapiro, and F. Waschek (Paderborn: 89–117). Tuplin, Christopher. 2013a. ‘Arshama: Prince and satrap’. In The Arshama Letters from the Bodleian Library, vol. 1: Introduction, edited by Lindsay Allen, John Ma, Christopher Tuplin, and David G. K. Taylor, http://arshama.classics.ox.ac.uk, pp. 5–44. Tuplin, Christopher. 2013b. The Arshama Letters from the Bodleian Library, vol. 3: Commentary, http://arshama.classics.ox.ac.uk. Turley, David. 2000. Slavery (Malden, MA, and Oxford). Tzochev, Chavdar. 2015a. ‘Markets, amphora trade and wine industry: The case of Thasos’. In The Ancient Greek Economy: Markets, Households and City-States, edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 230–53).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

345

Tzochev, Chavdar. 2015b. ‘Trade’. In A Companion to Ancient Thrace, edited by Julia Valeva, Denver Graninger, and Emil Nankov (Malden, MA and Oxford: 412–25). Van Alfen, Peter. 2015. ‘Aegean-Levantine trade, 600–300 BCE. Commodities, consumers, and the problem of autarkeia’. In The Ancient Greek Economy: Markets, Households and City-States, edited by Edward M. Harris, David M. Lewis, and Mark Woolmer (Cambridge and New York: 277–98). Van de Mieroop, Marc. 1997. ‘Why did they write on clay?’, Klio vol. 79: 7–18. Van der Brugge, Caroline and Kleber, Kristin. 2016. ‘The empire of trade and the empires of force: Tyre in the Neo-Assyrian and Neo-Babylonian periods’. In Dynamics of Production in the Ancient Near East, edited by Juan Carlos Moreno García (Oxford and Philadelphia, PA: 187–222). van der Ploeg, Johannes P. M. 1971. ‘Slavery in the Old Testament’, Vetus Testamentum Supplementum vol. 22: 72–87. van der Spek, Robartus J. 1990. Review of Dandamaev 1984. Mnemosyne vol. 43: 247–52. Van der Toorn, Karel. 2007. Scribal Culture and the Making of the Hebrew Bible (Cambridge, MA). Van Dongen, Erik. 2014. ‘The extent and interactions of the Phrygian kingdom’. In From Source to History: Studies on Ancient Near Eastern Worlds and Beyond Dedicated to Giovanni Battista Lanfranchi on the Occasion of his 65th Birthday, edited by Salvatore Gaspa, Alessandro Greco, Daniele M. Bonacossi, Simonetta Ponchia, and Robert Rollinger (Münster: 697–711). Van Driel, Govert. 1970. ‘Land and people in Assyria’, Bibliotheca Orientalis vol. 27: 168–75. Van Driel, Govert. 2002. Elusive Silver: In Search of a Role for a Market in an Agrarian Environment: Aspects of Mesopotamia’s Society (Leiden). Van Koppen, Franz. 2004. ‘The geography of the slave trade and northern Babylonia in the Late Old Babylonian Period’. In Mesopotamian Dark Age Revisited, edited by Hermann Hunger and Regine Pruzsinszky (Vienna: 9–33). Van Seters, John. 1992. Prologue to History: The Yahwist as Historian in Genesis (New Haven, CT). Van Wees, Hans. 1992. Status Warriors: War, Violence and Society in Homer and History (Amsterdam). Van Wees, Hans. 2002. ‘Homer and early Greece’, Colby Quarterly vol. 38.1: 94–117. Van Wees, Hans. 2003. ‘Conquerors and serfs: Wars of conquest and forced labour in archaic Greece’. In Helots and their Masters in Laconia and Messenia: Histories, Ideologies, Structures, edited by Nino Luraghi and Susan Alcock (Cambridge, MA, and London: 33–80). Van Wees, Hans. 2009. ‘The economy’. In A Companion to Archaic Greece, edited by Kurt Raaflaub and Hans Van Wees (Malden, MA, and Oxford: 444–67). Van Wees, Hans. 2011. ‘Demetrius and Draco: Athens’ property classes and population in and before 317 BC’, Journal of Hellenic Studies vol. 131: 95–114. Van Wees, Hans. 2013a. Ships and Silver, Taxes and Tribute: A Fiscal History of Archaic Athens (London and New York).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

346

Bibliography

Van Wees, Hans. 2013b. ‘Farmers and hoplites: Models of historical development’. In Men of Bronze: Hoplite Warfare in Ancient Greece, edited by Donald Kagan and Gregory F. Viggiano (Princeton, NJ, and Oxford: 222–55). Van Wees, Hans and Fisher, Nick. 2015. ‘The trouble with “aristocracy” ’. In ‘Aristocracy’ in Antiquity: Redefining Greek and Roman Elites, edited by Nick Fisher and Hans van Wees (Swansea: 1–57). Velkov, Velizar. 1964. ‘Zur Frage der Sklaverei auf der Balkanhalbinsel während der Antike’, Etudes Balkaniques vol. 1: 125–38. Velkov, Velizar. 1967. ‘РАБЫ-ФРАКИЙЦЫ В АНТИЧНЫХ ПОЛИСАХ ГРЕЦИИ VI–II вв. до н. э’ (‘Thracian slaves in the ancient Greek cities, 6th–2nd centuries BC’), Vestnik Drevneii Istorii vol. 4: 70–80. Vidale, Massimo. 2002. L’idea di un lavoro lieve. Il lavoro artigianale nelle immagini della ceramica greca tra VI e IV secolo a.C. (Padua). Vidal-Naquet, Pierre. 1986. The Black Hunter. Forms of Thought and Forms of Society in the Greek World (Baltimore, MD, and London). Villanueva-Puig, Marie-Christine. 1992. Images de la vie quotidienne en Grèce dans l’antiquité (Paris). Vinogradoff, Paul. 1922. Outlines of Historical Jurisprudence Vol. II: The Jurisprudence of the Greek City (Oxford). Vlassopoulos, Kostas. 2007a. Unthinking the Greek Polis. Ancient Greek History beyond Eurocentrism (Cambridge). Vlassopoulos, Kostas. 2007b. ‘Free spaces: Identity, experience and democracy in classical Athens’, Classical Quarterly vol. 57.1: 33–52. Vlassopoulos, Kostas. 2009. ‘Slavery, freedom and citizenship: Beyond a legalistic approach’, European Review of History vol. 16.3: 347–63. Vlassopoulos, Kostas. 2010. ‘Athenian slave names and Athenian social history’, Zeitschrift für Papyrologie und Epigraphik vol. 175: 113–44. Vlassopoulos, Kostas. 2011a. ‘Two images of ancient slavery: The “living tool” and the “koinônia” ’. In Sklaverei und Zwangsarbeit zwischen Akzeptanz und Widerstand, edited by Elizabeth Hermann-Otto (Hildesheim, Zurich, and New York: 467–77). Vlassopoulos, Kostas. 2011b. ‘Regional perspectives and the study of Greek history’, Incidenza dell’antico vol. 9: 9–31. Vlassopoulos, Kostas. 2011c. ‘Greek slavery: From domination to property and back again’, Journal of Hellenic Studies vol. 131: 115–30. Vlassopoulos, Kostas. 2012a. Review of Bradley and Cartledge 2011. Mnemosyne vol. 65: 877–81. Vlassopoulos, Kostas. 2012b. Review of Tamiolaki 2010. The Classical Review vol. 62.2: 555–7. Vlassopoulos, Kostas. 2013. Greeks and Barbarians (Cambridge). Vlassopoulos, Kostas. 2015. ‘Plotting strategies, networks, and communities in classical Athens: The evidence of slave names’. In Communities and Networks in Ancient Greece, edited by Claire Taylor and Kostas Vlassopoulos (Oxford: 101–27). Vlassopoulos, Kostas. 2016a. ‘Finley’s slavery’. In M. I. Finley: An Ancient Historian and his Impact, edited by Daniel Jew, Robin Osborne, and Michael Scott (Cambridge: 76–99).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

347

Vlassopoulos, Kostas. 2016b. ‘Does slavery have a history? The consequences of a global approach’, Journal of Global Slavery vol. 1: 5–27. Vogt, Joseph. 1974. Sklaverei und Humanität im klassischen Griechentum (Wiesbaden). Volkmann, Hans. 1990. Die Massenversklavungen der Einwohner eroberter Städte in der hellenistisch-römischen Zeit (Stuttgart). Von Dassow, Eva. 2011. ‘Freedom in Ancient Near Eastern Societies’. In The Oxford Handbook of Cuneiform Culture, edited by Karen Radner and Eleanor Robson (Oxford: 205–24). Waerzeggers, Caroline. 2003–4. ‘The Babylonian revolts against Xerxes and the “end of archives” ’, Altorientalische Forschungen vol. 50: 150–73. Waerzeggers, Caroline. 2010. The Ezida Temple of Borsippa: Priesthood, Cult, Archives (Leiden). Waerzeggers, Caroline. 2011. ‘The Babylonian priesthood in the long sixth century BC’, Bulletin of the Institute of Classical Studies vol. 54.2: 59–70. Wagner, Carlos G. 2000. ‘Elites, parentesco y dependencia en Tartessos’. In Las edades de la dependencia durante la Antigüedad, edited by María del Mar Myro, Juan Miguel Casillas, Jaime Alvar, and Domingo Plácido (Madrid: 321–47). Wallace, Robert. 2010. ‘Tecmessa’s legacy: Valuing outsiders in Athens’ democracy’. In Valuing Others in Classical Antiquity, edited by Ralph Rosen and Ineke Sluiter (Leiden: 137–54). Wallon, Henri. 1847. Histoire de l’esclavage dans l’antiquité. 3 vols. (2nd edn, 1879) (Paris). Weiler, Ingomar. 2004. ‘Sklaverei in der homerischen und altorientalischen Welt. Vergleichende Betrachtungen’. In Commerce and Monetary Systems in the Ancient World, edited by in Robert Rollinger and Christoph Ulf (Stuttgart: 270–91). Weippert, Manfred. 1973–4. ‘Die Kämpfe des assyrischen Königs Assurbanipal gegen die Araber. Redaktionskritische Untersuchung des Berichtes in Prisma A’, Welt des Orients vol. 7.1: 39–85. Welles, C. Bradford. 1965. Review of Kränzlein 1963. Gnomon vol. 37: 63–7. Welwei, Karl-Wilhelm. 1974. Unfreie im antiken Kriegsdienst. Teil I (Wiesbaden). Welwei, Karl-Wilhelm. 1977. Unfreie im antiken Kriegsdienst. Teil II (Wiesbaden). Welwei, Karl-Wilhelm. 2008. ‘Ursprung, Verbreitung und Formen der Unfreiheit abhängiger Landbewohner im antiken Griechenland’. In Sklaverei, Knechtschaft, Zwangsarbeit: Untersuchungen zur Sozial-, Rechts- und Kulturgeschichte, edited by Elizabeth Hermann-Otto (Hildsheim, Zurich, and New York: 1–52). West, Emily. 2004. Chains of Love: Slave Couples in Antebellum South Carolina (Urbana and Chicago). West, Martin L. 1989–92. Iambi et Elegi Graeci ante Alexandrum cantati. 2 vols. (Oxford). Westbrook, Raymond. 1991. Property and the Family in Biblical Law (Sheffield). Westbrook, Raymond. 1995. ‘Slave and master in ancient Near Eastern law’, Chicago Kent Law Review vol. 70.4: 1631–76. Westbrook, Raymond. 1998. ‘The female slave’. In Gender and Law in the Hebrew Bible and the Ancient Near East, edited by Victor H. Matthews, Bernard Levinson, and Tikva Frymer-Kensky (London: 214–38). Westbrook, Raymond and Jasnow, Richard (eds.). 2001. Security for Debt in Ancient Near Eastern Law (Leiden).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

348

Bibliography

Westbrook, Raymond and Wells, Bruce. 2009. Everyday Law in Biblical Israel (Louisville, KY). Westermann, William L. 1935. ‘Sklaverei’. In RE Suppl. VI: 894–1068. Westermann, William L. 1945. ‘Between slavery and freedom’, American Historical Review vol. 50: 213–27. Westermann, William L. 1955. The Slave Systems of Greek and Roman Antiquity (Philadelphia, PA). Whitby, Michael. 1994. ‘Two shadows: Images of Spartans and helots’. In The Shadow of Sparta, edited by Anton Powell and Stephen Hodkinson (London and New York: 87–126). Whittaker, Charles R. 1978. ‘Land and labour in north Africa’, Klio vol. 60: 331–62. Whittaker, Charles R. 1980. ‘Rural labour in three Roman provinces’. In Non-Slave Labour in the Greco-Roman World, edited by Peter Garnsey (Cambridge: 73–99). Whybray, Roger N. 1990. Wealth and Poverty in the Book of Proverbs (Sheffield). Wickert-Micknat, Gisela. 1983. Unfreiheit im Zeitalter der homerischen Epen (Wiesbaden). Wiedemann, Thomas. 1987. Slavery. Greece and Rome New Surveys in the Classics no. 19 (Oxford). Wiedemann, Thomas. 2000. ‘Fifty years of research on ancient slavery: The Mainz Academy project’, Slavery and Abolition vol. 21.3: 152–8. Wiedemann, Thomas and Gardner, Jane (eds.). 2002. Representing the Body of the Slave (London). Wiggermann, Frans A. M. 2000. ‘Agriculture in the Northern Balikh Valley: The case of Middle Assyrian tell Sabi Abyad’ In Rainfall and Agriculture in Northern Mesopotamia, edited by Remco Jas (Leiden: 171–231). Wilkinson, Tony, Wilkinson, Eleanor, Altaweel, Mark, and Ur, Jason. 2005. ‘Landscape and settlement in the Neo-Assyrian Empire’, Bulletin of the American Schools of Oriental Research vol. 340: 23–55. Willetts, Ronald F. 1954. ‘Freedmen at Gortyna’ Classical Quarterly vol. 4.3: 216–19. Willetts, Ronald F. 1967. The Law Code of Gortyn (Berlin). Willetts, Ronald F. 1977. The Civilisation of Ancient Crete (Berkeley and Los Angeles). Winn, Philip. 2010. ‘Slavery and cultural creativity in the Banda Islands’, Journal of Southeast Asian Studies vol. 41.3: 365–89. Wittgenstein, Ludwig. 1963. Philosophical Investigations. Translated by Gertrude E. M. Anscombe (Oxford). Wolters, Al. 1985. ‘Ṣôpiyyâ (Prov. 31:27) as hymnic participle and play on sophia’, Journal of Biblical Literature vol. 104: 577–87. Wood, Ellen M. 1988. Peasant-Citizen and Slave: The Foundations of Athenian Democracy (London). Wright, Benjamin G. 1997. ‘Ebed/doulos: Terms and social status in the meaning of Hebrew biblical and Hellenistic Roman culture’ Semeia vol. 83–4:83–111. Wright, Benjamin G. 2001. ‘δοῦλος and παῖς as translations of ‫עבד‬: Lexical equivalences and conceptual transformations’. In Proceedings of the IXth Congress of the IOSCS, Cambridge, England, edited by Bernard Taylor (Atlanta, GA: 263–77). Wunsch, Cornelia. 1993. Die Urkunden des babylonischen Geschäftsmannes IddinMarduk. Band I (Groningen).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

Bibliography

349

Wunsch, Cornelia. 1999. ‘The Egibi family’s real estate in Babylon (6th c BC)’. In Urbanisation and Land Ownership in the Ancient Near East, edited by Michael Hudson and Baruch Levine. Vol. 2 (Cambridge, MA: 391–419). Wunsch, Cornelia. 2000. Das Egibi-Archiv I: Die Felder und Gärten (Groningen). Wunsch, Cornelia. 2003. Urkunden zum Ehe-, Vermögens- und Erbrecht aus verschiedenen babylonischen Archiven (Dresden). Wunsch, Cornelia. 2007. ‘The Egibi family’. In The Babylonian World, edited by Gwendolyn Leick (New York and London: 232–43). Wunsch, Cornelia. 2010. ‘Neo-Babylonian entrepreneurs’. In The Invention of Enterprise: Entrepreneurship from Ancient Mesopotamia to Modern Times, edited by David S. Landes, Joel Mokyr, and William J. Baumol (Princeton, NJ: 40–61). Wunsch, Cornelia. and Magdalene, F. Rachel. 2012. ‘A slave is not supposed to wear such a garment!’, Kaskal vol. 9: 99–120. Wunsch, Cornelia. and Magdalene, F. Rachel. 2014. ‘Freedom and dependency: NeoBabylonian manumission documents with oblation and service obligation’. In Extraction and Control: Studies in Honor of Matthew W. Stolper, edited by Wouter F. Henkelman, Charles E. Jones, Michael Kozuh, and Christopher Woods (Chicago: 337–46). Yadin, Yigael. 1959. ‘Recipients or owners: A note on the Samaria ostraca’, Israel Exploration Journal vol. 9: 185. Yaldır, Aylin K. 2011. ‘Imported trade amphoras in Daskyleion from the seventh and sixth centuries BC and the Hellespontine-Phrygia route’, World Archaeology vol. 43.3: 364–79. Yamauchi, Edwin. 1966. ‘Slaves of God’, Bulletin of the Evangelical Theological Society vol. 9.1: 31–49. Yardeni, Ada. 1994. ‘Maritime trade and royal accountancy in an erased customs account from 475 B.C.E. on the Ahiqar scroll from Elephantine’, Bulletin of the American Schools of Oriental Research vol. 293: 67–78. Yuge, Toru and Doi, Masaoki (eds.). 1988. Forms of Control and Subordination in Antiquity (Leiden). Zaccagnini, Carlo. 1983. ‘Patterns of mobility among ancient Near Eastern craftsmen’, Journal of Near Eastern Studies ivol. 42.4: 245–64. Zadok, Ran. 2009–10. ‘The archive of Šulmu-šarri from Dūr-Katlimmu’, State Archives of Assyria Bulletin vol. 18: 47–90. Zadok, Ran. 2010. ‘Anatolians in Neo-Assyrian documents’. In Pax Hethitica: Studies on the Hittites and their Neighbours in Honour of Itamar Singer, edited by Yoram Cohen, Amir Gilan, and Jared L. Miller (Wiesbaden: 411–39). Zanovello, Sara. 2016. ‘From Slave to Free: A Legal Perspective on Greek Manumission’. Diss. Padua and Edinburgh. Zel’in, K. K. 1975. ‘Principes de classification morphologique des formes de dépendance’, Recherches Internationales à la Lumière du Marxisme vol. 84.3: 45–77. Zelnick-Abramovitz, Rachel. 2005. Not Wholly Free. The Concept of Manumission and the Status of Manumitted Slaves in the Ancient Greek World (Leiden). Zelnick-Abramovitz, Rachel. 2012. ‘Slaves and role reversal in ancient Greek cults’. In Slaves and Religions in Graeco-Roman Antiquity and Modern Brazil, edited by Stephen Hodkinson and Dick Geary (Newcastle upon Tyne: 96–132).

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

350

Bibliography

Zimmerman Munn, Mary Lou. 2003. ‘Corinthian trade with the Punic west in the classical period’. In Corinth: The Centenary, 1896–1996, edited by Charles Williams and Nancy Bookidis (Princeton, NJ: 195–217). Zurbach, Julien. 2013. ‘La Formation des cités grecques: Statuts, classes et systèmes fonciers’, Annales HSS vol. 68.4: 957–98. Zurbach, Julien. 2014. ‘Entre libres et esclaves dans l’Athènes classique’. In Les Affaires de Monsieur Andreau. Economie et société du monde romain, edited by Catherine Apicella, Marie-Laurence Haack, and François Lerouxel (Bordeaux: 273–85). Zurbach, Julien. (ed.). 2015. La Main-d’œuvre agricole en Méditerranée archaïque. Statuts et dynamiques économiques (Paris). Zurbach, Julien. 2016. ‘Aegean economies from Bronze Age to Iron Age: Some lines of development, 13th–7th centuries BC’. In Dynamics of Production in the Ancient Near East, edited by Juan Carlos Moreno García (Oxford and Philadelphia, PA: 357–68).

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Index locorum Classical Sources Achaeus of Eretria fr.32 TGF 301, 302 n. 29 Aeneas Tacticus 40.3 299 Aeschines 1.15 42 1.17 42, 298 1.54 297 1.97 43, 46, 244, 297 1.99 298 1.105 173 1.138–9 124 1.139 45 n. 69 2.127 298 3.41 298 3.41–4 44 3.44 297 Agatharchides FGrHist 86 F 17 101 Alexis fr.167 K-A 168 Anacreon fr.9 West 62 Andocides 1.28 140 1.38 43 1.96–8 45 n. 72 Antiochus of Syracuse FGrHist 555 F13 145 n. 68 Antiphanes fr.69 K-A 190 fr. 177 K-A 175 fr. 202 K-A 47 n. 79 Antiphon 1.20 42 n. 54, 47 n. 79 1.30 298 5.47 42 5.69 47 n. 79 6.4 42 Appian Punica 1.3 262 n. 11 2.9 261 9.59 265 15 262 n. 11

93 263 n. 17 117 263 Archemachus of Euboea FGrHist 424 F1 138 Archilochus Elegy 2 West 157 Aristophanes Acharnians 28–36 184 33 184 33–6 189 71–2 184 243 185 259 185 271–6 45 273 41 406 184 874–90 189 1003 185 Birds 1461–4 41 Clouds 5–7 304 n. 38 7 185 43–50 185 70–2 185 104–6 185 991 177 n. 47 Ecclesiazusae 817–22 185, 190 867 185 868 185 1113 185 Frogs 364 138 693–4 48, 263 n. 17 1046 300 n. 19 1407–10 300 n. 19 1508–14 41 Knights 464 191 n. 116 847–57 136 n. 41 947–8 174 1060–2 177 n. 47 1225 132 n. 26 1312 43 1397–1401 177 n. 47 Lysistrata 18 174

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

352 Aristophanes (cont.) Peace 190 185 563 190 573 41 919 185 1138 41 1249 297 Thesmophoriazousai 224 43 340 300 n. 19 407 300 n. 19 565–6 300 n. 19 Wasps 169–71 190 435 41 441–3 40 n. 45 712 173 n. 28 768 44 n. 62 Wealth 26 184 29 184 276 41 513 191 n. 116 552–4 168 843 172 1105 184 Fragments fr. 59 K-A 177 n. 47 fr. 95 K-A 41 n. 49 fr. 111 K-A 189 fr. 197 K-A 40 fr. 305 K-A 189 fr. 314 K-A 41 n. 49 fr. 352 K-A 173 n. 28 fr. 402 K-A 189 n. 105 fr. 596 K-A 44 n. 62, 300 n. 19 fr. 600 K-A 41 n. 49 fr. 871 K-A 41 Aristotle De generatione animalium 744b20 299 Politics 1252b12 172 n. 25 1253b32 25, 54, 59 1255a 44 n. 62 1263a35–7 130 1264a17–24 149 n. 6 1264a20–2 124 1264a32–6 143 1269a34–b12 159 1269a37–b5 143 1270a15–b6 164 1271a15–b6 127 1271a26–37 142, 162 n. 56 1271b 121 1271b40–72a2 143

Index locorum 1271b41–1272a1 148 n. 3 1272a3 142 1272a13–21 149, 162 n. 57 1272a18 161 n. 52 1272b15–22 149 1272b24–1273b25 259 n. 1, 264 1277b33–1278a13 63 1279b8 302 1330a25–33 144 1330a32–3 46 Rhetoric 1361a12–16 173 1361a21 33 1365a25–6 190 n. 109 1367a27–8 32, 63 1380a14–24 41 Fragments fr. 86 Rose 150 n. 10 fr. 513 Rose 41 fr.538 Rose 131 fr.586 Rose 143, 148 n. 3 fr.611.10 Rose 131 [Aristotle] Athenaion Politeia 2.2 274 n. 19 4.4 274 n. 19 6.1 274 n. 19 9.1 274 n. 19 50.2 43 52.1 40, 46, 274 n. 20 52.2 40 57.3 46, 297, 298 59.5 47 n. 78 De mirabilibus auscultatibus 88 261 Oeconomica 1344a25–6 174 1344a33–4 263 1344a35 221 n. 86 1344b 46 1350a 140 De virtutibus et vitiis libellus 1251b 63 Arrian Anabasis 1.25 248 n. 7 Asius f. 14 West 107 n. 2 Athenaeus Deipnosophistai 3.97d–e 302 6.84.263d–e 144 n. 67 6.85.263F 150 n. 10 6.93.267b–e 301 6.93.267d 301 6.93.267e 301–3

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Index locorum 6.93.272c–d 295 6.263 147 n. 2 6.265b–c 120 n. 56 15.695f–69a 157 Augustine Epistulae 10*.2 284 Carystius of Pergamum FHG IV 358 149 n. 6 C.Ptol.Sklav. 1.13–14 139 Cratinus fr.18 K-A 223 n. 1 fr.171 K-A 43 Critias 88B F2 D-K 145 n. 68 88B F31 D-K 119 n. 50 88B F37 D-K 125, 136 n. 41 Ctesias fr.4 ap. Athen. Deip. 14.44.639c 298 Ctesicles FGrHist 245 F1 301 Curtius 5.5.5–24 257 Democritus of Abdera 68 F 50 D-K 64 68 F 214 D-K 64 68 F 227 D-K 176 68 F 251 D-K 64 Demosthenes 9.3 298 17.3 298 18.51 173 n. 28 18.129 41 18.258–9 298 19.209 170 n. 15 20.131–3 299 21.46 42 21.48 42 21.149 300 n. 19 21.158 169 24.124 298 27.9 48, 170 n. 14, 171, 227 27.30 43 28.12 43 29.11 46 29.38 298 30.27 48, 298 33.8–9 48 34.37 138 35.50–1 138

37 179 37.4 171, 227 37.28–9 297 37.40 298 37.41 298 37.51 47 n. 78 45.28 46 45.33 41 n. 49 45.35 298 45.76 298 45.86 298 47.52–3 46 47.70 46 49.52 298 49.55–7 153 52.22 298 53.20 43 55.31–3 47 n. 78 57 67, 169 n. 8 57.34 68 57.35 69, 169, 273 57.45 69, 173 n. 28 58.19 40 58.21 40 58.8–9 138 [Demosthenes] 34 300 34.5 300 34.10 300 45.75–6 299 n. 17 45.83 299 n. 17 45.86 172 47.61 40 48.12 46 53.6 44, 298 53.16 45 53.19 44 n. 62 53.20 43, 47, 48 53.21 173 n. 28 53.27 48 59 67 59.9 46 59.18–19 170 n. 15 59.20 43 59.21 43 59.23–5 68 59.26 68 59.28 68 59.29 68 59.30 68 59.32 68 59.42 298 59.84 68 Dinarchus 1.23 40, 41 n. 49, 274 n. 20

353

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

354

Index locorum

Dio Cassius 44.39 302 n. 28 Dio Chrysostom Discourses 14.3–4 32 15.24 26 Diodorus Siculus 1.46.4 257 5.13.5 261 5.17.2 261 n. 8 5.17.3 261 13.58 266 13.97.1 48, 263 n. 17 17.27.28 257 n. 57 17.69.2–9 257 20.8.2–4 264 20.8.4 263 20.13.1 262 n. 11 20.40.5–6 259 n. 1 20.100.1–4 140 Diogenes Laertius 5.13 46 5.14 44 n. 63 5.54 46 5.62–2 46 5.73 46 6.39 41 n. 46 6.40 177 n. 47 6.47 177 n. 47 Dionysius Halicarnessensis Antiquitates Romanae 20.13.2 137 De Isocrate 1 P. 534 177 De Lysia 32 186 Dosiadas FGrHist 458 F2 163 Ephorus FGrHist 70 F 29 147 n. 2, 149 n. 6, 150 n. 10 FGrHist 70 F 117 128 FGrHist 70 F 149 121 Epicrates fr.5 K-A 111 Epictetus 4.1.34 32 Eupolis fr.163 K-A 175 fr.199 K-A 259 n. 1 Euripides Cyclops 654 223 n. 1 Florentinus Digest 1.5.4.1 25

Harpocration s.v. δίκη ἀποστασίου 72 Hellanicus FGrHist 323a F 25 48, 263 n. 17 Hellenica Oxyrhynchia 2.1.1 248 n. 7 Hermippus fr.63 K-A 250, 259 n. 1 Herodotus 1.137 40 1.156 256 1.2 122 n. 62 1.65 121 2.113 298 2.167 142 n. 63 3.48 41 n. 46 4.203–4 256 5.3.1 285 5.15 256 5.17 256 5.31 99 6.112 145 n. 68 6.20 257 6.101 257 7.27–8 249 7.165 259 n. 1 8.4 303 8.41 303 8.105 41 n. 46 8.106 303 n. 34 8.142 303 Hesiod Works and Days 37–41 117 46 117 306–34 117 308 117 341 117, 118, 219 399–400 118 405 172 n. 25 405–6 117 426 191 n. 116 435–6 117 441–7 117, 118 451 118 451–4 117 469–71 118 502–3 118 571–3 117, 118 571–81 117 589 117 597–9 118 597–600 117 602–3 118 641–5 117 663–93 117

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Index locorum 765–8 118 786–7 117 Hesychius s.v. ἀμφίδουλος 44 n. 62 s.v. Σκυθικός 300 n. 19 Homer Iliad 1.29–31 112 1.275–6 110 4.35 125 6.289–92 120, 145 n. 68 6.455 60, 113 6.462–3 113 6.528 61 n. 12 7.467–75 112 12.310–21 157 16.830–2 61 16.831 113 19.192–5 110 19.243–6 110 19.245 112 19.33 64 n. 22 19.85–9 110 19.330–3 114 n. 36 19.339–3 112 21.33–44 112 21.441–7 214 21.441–55 116 23.509–13 112 23.700–5 112 Odyssey 1.144–9 115 1.397–8 157 1.427–38 112, 116 n. 44 2.239–40 115 2.290 115 4.244–6 112 4.245 296 n. 6 4.49 114 n. 36 4.735 112 4.756 114 7.225 64 n. 22 9.200–7 114 n. 36 10.84–5 116 11.431 114 n. 36 11.489–91 118 n. 49 14.4 296 n. 6 14.20–5 161 14.26–8 115, 161 14.37–44 111 14.55–71 111 14.61–7 113 14.63 296 n. 6 14.100–4 115, 161 14.102 115, 116 14.105–6 161 14.199–359 122 n. 62

14.340 61, 113 14.449–52 113, 216 15.402–84 116 n. 44 15.417 120, 145 n. 68 15.425–9 112 15.483–4 112 16.17 111 16.31 111 17.212 116 n. 44 17.297–9 115 17.300 115 17.320–3 61, 113 18.322 116 n. 44 18.356–64 116 19.525–9 64 n. 22 20.5–30 113 20.14.80–1 115 20.16.51–2 115 20.105–8 115 20.147–54 116 20.162–3 115 20.173–5 115 20.185–7 115 21.213–16 113 21.244 116 n. 44 21.350–2 116 22.35–41 113 22.159 116 n. 44 22.421–2 114 22.440–4 112 22.462–73 112 22.462–77 112 24.209–25 116 n. 44 24.222–5 115 24.226–7 115 24.244–7 115 24.245–7 115 24.340–2 116 24.389 120, 145 n. 68 Homeric Hymn to Demeter 119–44 107 n. 2 Homeric Hymn to Dionysus 6–57 107 n. 2 Hipponax fr. 26–26a West 107 n. 2 fr.27 West 107 n. 2 fr.40 West 107 n. 2 fr.115 West 107 n. 2 Hybrias skolion 157 Hyperides 1A fr.4.1 43, 45 n. 67 3 46, 170 n. 15 3.11 47 3.15 44, 298 3.16 299

355

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

356

Index locorum

Hyperides (cont.) 3.22 43, 47, 297, 298 3.23–8 44 3.24 42 3.26 189 3.5 41 3.9 46 4.3 43 fr.120 42 Isaeus 6.16 298 6.19 177 6.33 177 6.49 299 8.9–10 298 8.10 298 8.35 43 8.41 47 n. 79 Isocrates 1.21 298, 299 n. 15 4.123 298 4.150 298, 299 n. 15 6.88 298 6.96 125 7.26 299 n. 15 7.32–3 175 12.97 298 12.181 42 12.211–12 137 n. 44 17.49 298 18.52 297 18.52–4 46 John Chrysostom, ap. Migne, Patrologia Graeca 60.418–19 124 62.693 124 Justin Martyr 11.14.11–12 257 18.3.1–19 253 n. 30 21.4.6 265 Livy 21.45.7 263 Lycurgus Against Leocrates 29–32 298 29 298 30 298 32 298 34 298 58 177 n. 49 65 46, 298 Fragments fr.11.1 46 n. 74, 297, 298

Lysias 1.12 41 1.18–22 41 4.16 298 5.3–5 140 5.5 172 7 174 7.10 174 7.16 140 7.16–17 153 7.19 297 10.9 300 n. 21 10.19 47 n. 76 12.19 171, 227 13.18 44 n. 62 13.65 46 23 67 23.2 67 23.3–6 67 23.7 67 23.9 67 23.10 67 23.14 67 24.6 172, 181 fr.1 Thalheim 48, 177, 298 Menander Aspis 238–45 41 n. 49 Dyscolus 141–3 45 n. 67 Epitrepontes 378–80 43 469 44 Heros 1–3 41 n. 49 20 73 n. 46 27–30 174 Kolax 38 171 n. 19 Samia 380–2 44 Fragments fr.411 K-A 44 n. 62 fr.891 K-A 285 Myron of Priene FGrHist 106 F2 131 Oppian Halieutica 3.370 302 n. 28 Pausanias 3.20.6 127 n. 8 Pherecrates fr.10 K-A 44 n. 62 fr.212 Kock 302 n. 28

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Index locorum Philo Special Laws 2.79–85 206 n. 28 Phrynicus fr.14 K-A 173 n. 28 fr.47 K-A 41 Plato Cratylus 384d 41 410a 170 n. 10 Euthydemus 285c 223 n. 1 301e–302a 33 Euthyphro 4c5 298 Gorgias 483a–b 42 Laches 187b 223 n. 1 Laws 633b9–c4 133 n. 31 637a 137 674a 263 704–5 137 720c7 298 763a 304, 305 n. 39 776b–d 144 776c–d 143 776d–e 40 776d3 298 777a 40, 304 777a4 298 777a6 298 777c3 297 777d4 298 777e6 298 794b7 298 853e 304, 305 n. 39 868a 42 916a 44, 225 n. 13 Meno 82a–b 44 n. 62, 170 n. 15 Protagoras 314c 41 n. 46 Republic 9.537e–538a 300 n. 19 9.578d9 298 9.578d–e 227 9.578d–579a 304 n. 38 9.578e 53, 171 9.578e4 298 9.578e–579a 153 Symposium 183a 64 Theatetus 173c3 299 n. 15

357

Plato Comicus fr.61 K-A 170 n. 10 Plautus Cassina 71 262 Plutarch Agis 5.1–2 127 Alcibiades 123b–c 250 n. 20 Alexander 22 253 37.1 257 n. 57 Amatorius 751b 124 Convivium Septem Sapientium 152d 124 Eumenes 8.5 249 Lycurgus 8.3 127 16.1 127 28.1–3 134 n. 32 28.1–4 131 28.1–5 134 n. 35 28.5 125, 131 n. 25, 135 Lysander 4.1–2 121 12.1 142 Nicias 3 44 n. 64 Pericles 24.6 177 Solon 1.6 123 23.2 274 n. 19 23.6 = fr.63 Leão and Rhodes 137 n. 46 24.1 = fr.65 Leão and Rhodes 138 Themistocles 29 250 n. 20 [Plutarch] Moralia 836e 177 Pollux 3.76 44 n. 62 3.78–9 41 3.80 302 n. 28 3.83 72 n. 42, 132 n. 26, 143, 293 3.125 44, 170 n. 15 10.14 285 Polybius 1.29.7 265 1.71.1 260 1.72.1–7 260

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

358

Index locorum

Polybius (cont.) 3.24.4–14 262 n. 11 15.18.1 265 38.8.4 266 38.8.11 266 Poseidonius FGrHist 87 F8 138 scholia Aristophanes’ Ecclesiazusae 253 178 Aristophanes’ Knights 739, 740 177 Demosthenes 2.13 170 n. 15 Semonides of Amorgos fr.7 West 176 Solon fr.9 West 62 fr.13 West 117 fr.36 West 28, 61, 274 n. 19 Sosicrates FGrHist 461 F4 150 n. 10 Stephanus Byzantinus s.v. Λακεδαίμων 131 n. 24 Strabo 3.2.10 266 8.5.4 127 14.5.2 283 Strattis fr.3 K-A 177 fr.36 K-A 248 n. 7 Tacitus Annales 14.42 47 n. 79 Teleclides fr.8 K-A 173 n. 28 Theognidea 301–2 107 n. 2 535–8 107 n. 2 1211–16 107 n. 2, 123 Theophrastus Characteres 4.6 173 n. 28 4.15 190 9.8 177 13.5 137 n. 46 19.2 300 n. 19 21.3 279 n. 35 21.5 170 n. 15 24 169 30.9 43 30.15 43 30.17 43 Historia plantarum 5.7.6 191 n. 116

Theopompus of Chios FGrHist 115 F 13 125 FGrHist 115 F 39 101 n. 25 FGrHist 115 F 40 100, 101 n. 25 FGrHist 115 F 114 253 FGrHist 115 F 122 101, 120 n. 56, 139, 145 FGrHist 115 F 129 101 n. 25 FGrHist 115 F 130 101 n. 25 FGrHist 115 F 253 44 Thucydides 1.5 275 n. 24 1.5–6 100 n. 24 1.10.3 119 1.98 62 1.128.1 132 n. 26 1.138.5 250 n. 20 2.4.2 303 n. 33 2.97.5 285 3.17 172 3.73 153 3.73.1 298 4.26 135 4.80 135 4.85 62 5.34 135 5.82.6 303 n. 33 6.15 259 n. 1 7.75.5 172 8.28.3–4 62 n. 17, 83 8.40 139 8.40.2 99, 145, 194, 297 Timaeus FGrHist 566 F11 298 Varro De re rustica 1.1.10 264 Xenophon Agesilaus 1.18–21 229 n. 24 Anabasis 1.2.1 248 n. 7 2.3.17 257 2.4.27 257 2.5.13 248 n. 7 3.2.23 248 n. 7 4.6.14–15 137 n. 44 4.8.4–8 169 4.8.14 125 6.3.3 249 6.6.38 249 n. 9 7.3.43–7.4.2 284 7.3.48 284

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Index locorum 7.4.2 284 7.6.26 249 n. 9 7.7.53 100 7.8.7–22 17, 248 7.8.12 248, 249 7.8.19 248, 249 Apologia Socrates 29 178 Cynegeticus 2.3 170 n. 15 2.4 259 n. 1 5.34 137 n. 46 Cyropaedia 4.2.2 303 4.2.37 304 4.3.1–2 304 8.2.5 178 n. 52 De equitandi ratione 2.2 43, 244 Hellenica 1.2.4 249 n. 9 1.6.24 48, 263 n. 17 2.1.1 173 n. 28 3.1.13 248 n. 7 3.2.2 249 n. 9 3.2.26 249 n. 9 3.3.6 125 4.1.15–16 248 4.6.6 249 n. 9 5.3.7 40 6.2.6 249 n. 9 6.2.37 173 n. 28 6.5.28–9 135 Hiero 1.28 42 6.10 173 n. 28 Lacedaemonion Politeia 2.6–9 137 n. 44 2.10 137 4.6 137 5.3 163 6.1–2 137 6.3 130 7.1–2 142 n. 63 12.4 136 n. 41 Memorabilia 1.3.9–11 64 1.5.6 64 2.1.16 41 2.3.3 172, 181 2.4.5 249 n. 10 2.5.2 44, 170, 298 2.5.5 170 n. 15 2.7.6 178 2.8 63, 169 3.5.26 248 n. 7

3.11.1–4 173 4.2.22 63, 249 n. 10 Oeconomicus 3.4 40 3.10 170 n. 15 4.2–4 63 4.3 142 n. 63 5.3 175 n. 35 5.15–16 174 n. 29 7.32–4 176 7.35 298 n. 10 7.36 176 7.41 170 n. 15, 244 9.5 44 n. 62 9.15 174 10.12 41 13.9–11 174 n. 29 13–23 64 16–18 175 n. 35 19 175 n. 35 41 176 Symposium 4.4. 187 n. 97 4.36 284 De Vectigalibus (Poroi) 4.4. 249 n. 10 4.14 43, 171, 179 4.15 249 n. 10 4.17 249 n. 10 4.21 249 n. 10 4.23 249 n. 10 4.23–4 261 4.24 249 n. 10 4.25 249 n. 10 4.26 249 n. 10 4.30 249 n. 10 4.35 249 n. 10 4.36 249 n. 10 4.39 249 n. 10 [Xenophon]/Old Oligarch Athenaion Politeia 1.10 235 1.10–12 124, 131 1.11 43, 45, 69, 153 1.17 43 1.19 172 n. 23 2.8 170 n. 10 2.10 177 59.5 47 Hebrew Bible Amos 2:6 204 8:4–6 204 Deuteronomy 1:39 202

359

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

360 Deuteronomy (cont.) 5:12–15 214 12:18 214 15:3 205 15:12 210 n. 45 15:12–15 206 n. 28 15:12–18 208 15:13–15 209 15:17 202 n. 17 15:18 209 16:11 214 20:12–14 202 21:10–14 209 n. 42 23:20–1 205 24:7 202 24:14–15 214 Exodus 12:43–4 202 20:2–17 214 20:17 214 21:1 203, 210 n. 45 21:1–6 205 21:1–11 204, 207 n. 37 21:2 206 n. 28 21:6 202 n. 17 21:7–11 206 21:20–1 207 n. 34 21:26–7 207 Ezekiel 27:13 199 n. 1, 286 Genesis 12:16 202, 214 16:1–15 214 17:12–13 12 n. 24, 202, 214 17:23 214 20:14 202, 215 24:1 214 24:35 202 26:12 215 29:24 202 29:24–9 215 30:43 215 32:6 215 37:26–9 202 42:10 85

Index locorum 19:13–21 221 42:10–16 221 Judges 6:27 215 I Kings 5:7–8 217 II Kings 4:1–7 204 5 218 5:2 202 5:26 218 18:19–35 231 18:31–2 231 Leviticus 25:8–55 210 25:10–12 220 25:29–31 209 n. 43 25:39 202 n. 17 25:39–40 210 25:39–46 209 25:40 206 n. 28 25:44–6 202 n. 17, 205 25:46 210 Micah 2:1–2 212 Nehemiah 5:1–13 204, 220, 237 n. 9, 275 5:8 204 Numbers 14:31 202 21:29 202 Proverbs 22:7 204 31:10–31 218 31:13 219 31:15 219 31:16 219 31:19 219 31:20 219 31:21 219 31:22 219 31:23 219 31:24 219

Isaiah 5:8 212 36:4–20 231 44:5 206

Qoholeth (Ecclesiastes) 2:1–12 221 2:4–7 221 I Samuel 8:14 218 n. 79

Jeremiah 22:13 214 34:8–22 204, 210, 219, 220, 275 Job 1:1–3 221 1:14–18 221

II Samuel 9:1–13 216, 218 n. 79 9:2 216 9:3 216 9:9 217 n. 70 9:9–11 218

OUP CORRECTED PROOF – FINAL, 26/6/2018, SPi

Index locorum 9:10–11 216 16:1 217 n. 71 16:1–4 216 n. 70 19:25–31 216 n. 70 Near Eastern Sources BE8123 52 Camb. 143 49 Camb. 321 236 Camb. 329 236 n. 5 Camb. 330 243 n. 43 Camb. 365 51 CIS I 269 262 Cyr. 161 239 Cyr. 332 51 Cyr. 339 51 Dar. 53 49 n.81 Dar. 379 238, 289 n. 73 Dar. 429 239 Dar. 509 50, 243 n. 42 LNB 6 48, 49 Mich. 26 256 n. 51 Nbk. 409 50 Nbn. 679 50 Nbn. 682 50 Nbn. 693 51 Nbn. 697 51 Nbn. 1113 49, 86, 87 nn.16 and 17 NRVU 27 52 PF 1363 255 PF 1368 255 PF 1557 255 PF 2055 255 PFa 18 255 PFa 30 255 PNA 2: 771b 288 n. 70 SAA V 10 224 n. 4 SAA V 149 285 n. 59, 289 n. 71 SAA V 150 139 n. 52, 289 n. 71 SAA VI 1–9 227 SAA VI 4 227 SAA VI 2 227 SAA VI 6 227 SAA VI 7 225, 227 SAA VI 34 227 SAA VI 34–56 227 SAA VI 37 227 SAA VI 40 227 SAA VI 41 227 SAA VI 50 227, 232 n. 46 SAA VI 52 227

SAA VI 57 227 SAA VI 58 227 SAA VI 65 232 n. 46 SAA VI 83 225 n. 13 SAA VI 90 232 n. 46 SAA VI 91 232 n. 46 SAA VI 96 225 n. 13 SAA VI 103 225 n. 13 SAA VI 112 232 n. 46 SAA VI 123 232 n. 46 SAA VI 129 232 n. 46 SAA VI 134 225 n. 13 SAA VI 153 232 n. 46 SAA VI 163 232 n. 46 SAA VI 169 232 n. 46 SAA VI 269 232 n. 46 SAA VI 315 232 n. 46 SAA IX 14 224 n. 4 SAA X 191 223 n. 1 SAA XI 201–20 232 SAA XI 207 233 n. 51 SAA XIV 1 232 n. 46 SAA XIV 2 232 n. 46 SAA XIV 8 225 n. 13 SAA XIV 22 232 n. 46 SAA XIV 36 232 n. 46 SAA XIV 85 226 SAA XIV 91 288 n. 70 SAA XIV 93 226 SAA XIV 97 226 SAA XIV 101 226 SAA XIV 108 226 SAA XIV 195 225 n. 13 SAA XIV 198 232 n. 46 SAA XIV 207 232 n. 46 SAA XIV 229 232 n. 46 SAA XIV 243 225 n. 13 SAA XIV 244 225 n. 13 SAA XIV 254 232 n. 46 SAA XIV 265 232 n. 46 XIV 345 232 n. 46 SAA XIV 355 232 n. 46 SAA XIV 401 232 n. 46 SAAB V 5 226 SAAB V 13 226 SAAB V 44 226 SAAB V 46 226 SAAB V 52 226, 227 n. 21 SAAB V55 226 SAAB IX 71 226 TMH 2/3 121 52 YOS 3 165 52 YOS 6 143 51 YOS 7 189 52

361

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

General Index Abdera 284 n. 53 Abimelech of Gerar 214 Abraham 214 Achaeus of Eretria 301, 302 Adams, Colin 191 Adeimantus 43 Aegina 9, 99, 194 Aeneas Tacticus 299 Aeschines 70, 168 n. 2, 173, 297, 300 Aeschines of Sphettus 177 Africa 97, 282–3, 284 Agathocles 262 n. 11, 264 Aḫata 87, 88 Ahat-abiša 226 Aḫija-likin 51 Ahmad, Ali 85, 227 Aḫušunu 52, 87 Ajartu 51 Akbar 225, 227 Akkadian 49, 50, 53, 85, 199 n. 1 Alcias 174 Alexander the Great 257 Alexis 139 n. 50 Alkê 299 Allain, Jean 9, 29 n. 15 Amarna letters 205 n. 27 Ambrakis 44 n. 63 Amorgos 62 n. 17 Amos, Book of 204 amphipolis see under slavery (and servitude), terminology Amtija 50 Anatolia 7, 114 n. 35, 247–51, 276, 280, 281, 286 Anatolian slaves 9, 167, 252, 288 basilikoi laoi 251 slave ownership 253 Andania 194–5 andrapoda see under slavery (and servitude), terminology Andreau, Jean 19 Andrewes, Anthony 135 Anthemion 177 apeleutheros see under slavery (and servitude), terminology Aperghis, Gerassimos 20, 255, 257 nn. 57 and 58 aphamiotai see under slavery (and servitude), terminology Apollodorus 43, 47, 299 n. 17 Appian 261, 262, 263 Aramaic 49

Arbela 232 Arcadia 148 Archibald, Zosia 100 n. 24 Ardiaioi 100 Arethusius 43, 47 Arginusae, Battle of 48, 263 Argos 148 Gymnetes 144 Aristarchus 43, 235 n. 2 Aristoclides 62 Aristophanes 184–5, 189–90, 192, 297 Aristophanes of Byzantium 144 n. 67 Aristotle definition of freedom 58, 73 definition of slavery 59, 64 n. 22, 70 on Cretan slavery 148, 149, 158, 159, 162 on feeding slaves 299 on helots 126, 127, 130, 131, 133 n. 31, 136, 144, 145, 148 on the Penestai and Mariandynoi 138 will 44 n. 63 Aršama, satrap 7, 251–2, 255, 288 Asidates 248, 249 Aspis 265 Assur 224 n. 4, 231 Aššurbanipal 228, 231 Assyria 6, 7, 85, 139, 199 n. 1, 223–34, 236, 288, 289 bound tenancy 232, 273 debt bondage 226 deportees to Assyria 231–2, 256, 290 economic role of slavery 223, 234 elites 224, 226, 227, 229, 230, 234 metaphorical meaning of slavery 225 sale of slaves 223–4, 225 slave numbers 227, 234, 238–9 slave ownership 226–30, 288 slave prices 228–30, 239, 288 slave status 225–6 Atarham 289 n. 71 Athenaeus 301–2 Athenogenes 46, 47 Athens 83, 98, 120–1 andrapoda 304 n. 38 apeleutheros 73, 77 apophora 43, 178 n. 51, 235 ‘aristocracy’ 122 n. 67 atimoi 159 banking 179 bathhouses 177 children of slaves 43

OUP CORRECTED PROOF – FINAL, 28/6/2018, SPi

General Index citizenship 36, 63, 68, 69, 83, 143, 168, 169, 186–7, 193, 273 classical period 39–48, 168–95 comedies as source material 182, 183–5 debt bondage 10, 89, 107 demosioi (public slaves) 273 desirable slave traits 170 n. 15 douleia/doulos 53, 59, 65, 66, 78 doulos and oiketes 53, 153, 297 eleutheria/eleutheros 58, 59, 63, 66, 78 elites 143, 167, 168, 172–80, 187 exeleutheros 77 freedom 70, 71, 73, 78, 89–90 hiring out slaves 43, 45 killing of slaves 42, 46 law on hubris 42 manumission of slaves 44, 46, 83–4, 140, 141 metaphorical meaning of slavery 62–4, 65–6, 291 metics 36, 67, 69, 83–4, 167, 175 n. 34, 176–7 moneylending 179–80 non-Greek slaves 169, 292 paramone 72–3 pledging of slaves 47–8 property ownership 36 punishment of slaves 40–1, 44–5, 141, 194 relative property rights 31 sale of slaves 44 sexual abuse of slaves 41–2 shrine of the Eumenides 43 silver 121, 178 slave numbers 95, 103, 119, 171, 301 slave ownership 25, 167, 168–72 slave ownership (elite) 172–80 slave ownership (sub-elite) 180–93 slave owners’ rights 39–42, 43–8, 54 slave owners’ liabilities 46–7 slave prices 170–1, 182 n. 78, 228, 239, 271, 288 slave rights 42–3 slaves earning income 43 slaves owning slaves 44, 113 slave society 98, 103, 143 slave sources 169–70, 250, 275–6 slave status 68–70, 144 slave trade 9, 167, 274, 284 subjugation of allies 62 tattooing of slaves 40 tax farming 179 theft of slaves 40, 46 therapontes 304 n. 38 Theseion 43 torture of slaves (basanos) 45 trials 46, 67–70, 76 see also Attica Attica (classical period) 168–95, 292 citizens 182

363

economic role of slavery 121, 173–80, 189, 192–3 farmers and farming 117 n. 48, 186–91 farm produce 175 honey 175 laws banning export of agricultural products 138 leasing slaves 174, 179, 274 master, as a term of address 85 non-Greek slaves 120, 144, 250, 280, 292 no-slaving zone 275–6 oiketes 7 phialai inscriptions 44 n. 66, 175 n. 34 silver mines 178 Solon’s reforms 28, 107, 108 n. 7, 274 slave numbers 179, 281 n. 45 slave ownership 31 n. 21, 167, 168–93 slaves owning slaves 50 slave status 69 slave trade 121, 160, 250 stelai as source material 183 n. 78, 250 zeugitai 188 n. 102 see also Athens; Crete; Greece (archaic); Sparta Atys 249 Augustine 284 Azoria 162 n. 55 Bābāya family 239 Babylon 240, 241 Babylonia 6–7, 86–9, 102, 235–45, 288, 289 apprenticeships 244 Babylonian slaves in Persia 256 debt bondage 87–8, 207 n. 37 economic role of slavery 243–5 elites 237, 239 freedom 88, 89, 91 Judahite exile 220 mandattu 50, 235 manumission of slaves 51 manumission with obligations 235 pledging of slaves 51, 52 sale of slaves 48, 50–2 slave branding 49 slave numbers 103 slave ownership 48–9, 238–9, 288 slave prices 44, 288 slaves 91, 274, 289 slaves owning slaves 50, 235 slave status 87, 88, 89 temple servants 87–8, 237 n. 11 workhouses 49 Bacchiads 122 n. 67 Badian, Ernst 28 n. 10 Bakchis 44 Baker, Heather 20, 85, 224 n. 7, 225 n. 11, 239 n. 21, 241 Balat ̣u-ereš 225 Bales, Kevin 9

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

364

General Index

Banat-ina-Esagila 87, 88 Banda Islands 97 Barca 256 Bariki-ili 49, 86–7 basilikoi laoi see under slavery (and servitude), terminology Bazuzu 51 Beaulieu, Paul-Alain 245 Bechtold, B. 264 Beldam Painter 41 n. 45 Bel-upaḫḫir 50 Benedict, Burton 97 Benson, Bruce 37 Beringer, Walter 110, 111 Besa 253 Bible, Hebrew 200, 212, 213 Bielman, Anne 62 n. 17 Birks, Peter 38 Borowsky, Ron 66–7 Borsippa temple of Ezida 236 n. 6 Bourdieu, Pierre 76 Bradley, Keith 19 Brasidas 135 Braund, David 250, 286 Bresson, Alain 188 n. 99, 239, 279, 282, 301 Briant, Pierre 20, 255 n. 42, 257 Burašu 51 Burchfiel, K. J. 110 Caesar, Julius 284, 302 n. 28 Callimachus 46 Callistratus 144 n. 67, 174 Cambridge World History of Slavery 2–3, 19 Cambyses 257 Cameron, George 93 n. 1, 255 Canevaro, Mirko 42, 184 Cantarella, Eva 110 Cappadocia 255, 288 Carr, David 201 n. 9, 215 Carthage 3, 7, 98, 259–66, 287, 289 n. 74 economic role of slavery 263–6 elites 264 manumission of slaves 262 metaphorical meaning of slavery 263 Punic word for slave 263 slave marriage 262 slave numbers 265 slave ownership 264, 266 slave population 14 slave society 259 slave status 260–3 slave trade 261–2 trade 264 Cartledge, Paul 19, 28 n. 12, 31, 91 n. 35, 127, 128, 132, 133, 135, 140, 194 Cassius Dionysius 261 n. 9, 263 Ceccarelli, Paola 257 n. 56 Cephalus 171, 177

Cephalus of Collytus 178 Cerdon 43, 47 Chairedemos 172 Childe, Vere Gordon 2 n. 2 China 272 n. 9 Chios 9, 99, 139, 145, 194, 297 Chrysippus of Athens 301 chrysonetai see under slavery (and servitude), terminology Cilicia 248, 252, 288 Cleon 177 Cohen, Edward 42 nn. 52, 55, and 56, 300 n. 20 Corcyra 297 n. 9 Corinth 9, 99, 122 n. 67, 194 Corsica 261 Cratinus 46 Crete 6, 9, 62 n. 17, 121, 124, 143, 147–65, 172, 290, 293 andreia (mess groups) 158, 159, 160, 162–3 apetairoi 88, 158 aphamiotai 150 n. 10 children of mixed-status couples 155 chrysonetai 150 n. 10 citizenship 75, 158–9, 164 debt bondage 10, 73–5, 154 dolos and woikeus 150–3 economic role of slavery 157–9 eleutheros 75, 76 gender balance 161 inscriptions 147 katakeimenos/katathemenos 73–4, 154 klarotai 143, 147 n. 2, 148 n. 3, 150 n. 10 lack of slave rebellions 148–9, 159 manumission 155 mnoia (public slaves) 150 n. 10, 157 mnoitai 143 oiketeia 150 n. 10 perioikoi 148 n. 3, 150 n. 10 punishment of slaves 152 slave residential patterns 160–2 sale of slaves 155 slave families 160 slave marriage 157, 160 slave numbers 160 slave ownership 153, 155–6 slave possession of money 154, 163 slave society 164–5 slaves seeking sanctuary 155–6 slave status 73–5, 83, 88, 153–5 state intervention in slavery 156 Crieland, Jan Paul 109 Critias 125, 131 Critobolus 64 Cruz-Uribe, Eugene 28 n. 10, 253 Ctesias 298 Ctesicles 301 Cydonia 149 n. 6, 150 n. 10 Cyrebus 178 Cyrus the Great 256, 257 n. 56

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

General Index Dal Lago, Enrico 7 Dandamaev, Muhammad A. 20, 27, 48 n. 80, 55, 103, 237, 238, 239, 243, 255, 256, 257, 289 n. 73 Dardanians 101 Darius I. 256 Daskyleion 286 David, king 216 Davies, J. K. 152, 158, 162, 163, 164, 194, 247 Davies, Norman 3–4 Davies, P. A. 76 n. 58 Decimus Junius Silanus 264 Decree of Demophantus 45 n. 70 Delos 194, 195 Delphi 71, 276 n. 25, 279 n. 34, 280, 281 Demeas of Collytus 178 Demetrius 174 Democritus 301, 302 demosioi see under slavery (and servitude), terminology Demosthenes 171, 177, 187, 298, 299 de Ste. Croix, Geoffrey E. M. 17, 72, 74, 127, 128, 129 n. 18, 132, 181, 182 n. 78, 184, 249 Descat, Raymond 19, 101, 108 n. 6, 120 n. 56, 295, 301, 304 Deuteronomy, Book of 202 n. 17, 204 n. 25, 205, 208, 209, 210, 211, 214 De Vaux, Roland 15 n. 39, 20, 103 Dio Cassius 302 n. 28 Diodorus Siculus 257, 263, 264–5 Diogenes the Cynic 177 n. 47 Dion 300 Dmitriev, Sviatoslav 42 n. 56, 295 n. 3 Docter, Roald 264 dolos see under slavery (and servitude), terminology Douglass, Frederick 134 doulos see under slavery (and servitude), terminology Dover Kenneth 39 n. 41 Ducat, Jean 101 n. 25, 122, 128, 129, 134 n. 35, 136, 138 n. 48, 140, 157 Duplouy, Alain 122 Dur-Katlimmu 227 Dur-Šarrukin 231 Durkheim, Emile 31 Dutch East India Company (VOC) 97 Eagleman, David 66 n. 27 Ecclesiastes, Book of see Qoheleth, Book of Eco, Umberto 60 Eden, Jeff 276 n. 29 Egibi family 238, 240–1, 243 Egypt 3, 7, 51, 84, 251–4, 280, 287 Anatolian slaves in Egypt 252 branding/tattooing slaves 251, 252, 253 Egyptian slaves in Persia 255, 256, 257 slave prices 271 word for slave 14, 84

365

Einstein, Albert 99 Eion 62 Elephantine papyri 252, 253 Eleutherna 150 n. 10 Elisha 218 Eltis, David 28 n. 10 Engerman, Stanley 28 n. 10 Enlil-ḫatin 241 Enlil-šum-iddin 241 Ephorus 128–9, 130, 145 Epicrates 46 Epirotes 261 n. 9 Erechtheum building inscriptions 181 Esagil-ramat 51 Esarhaddon 223 n. 1, 232 Esau 215 Etymologicum Genuinum 302 n. 28 Etymologicum Magnum 302 n. 28 Etymologicum Symeonis 302 n. 28 Euboulides 67, 68, 69, 70 Euctemon 177 Eumenes of Cardia 249 Eupatridae 122 n. 67 Euphorion 302 n. 28 Eupolis 132 n. 26 Euripides 300 n. 19 Eurybiades 303 Eutherus 63, 72 Euthydemus 63 Euxitheus 68, 69, 70 exeleutheros see under slavery (and servitude), terminology Exodus, Book of 202 n. 17, 203, 204 n. 25, 206, 207, 210 Decalogue 214 Ezekiel 199, 250, 252, 286 Fachard, Sylvian 190 n. 112 Fales, F. M. 229, 230, 233, 234 Falkenstein, Adam 255 Fars 7, 254–8, 290 kurtaš 255 slave numbers 255 Faust, Avraham 104, 209 n. 43, 212, 213 Fermor, Patrick Leigh 170 n. 9 Février, James-Germain 20, 262 Figueira, Thomas 31 n. 21 Finkelman, Paul 29 n. 15 Finley, Moses I. and Westmann 13 definition of slavery 10 n. 19 distinction between Greek and Near Eastern worlds 1–2, 14, 81–2 influence 3, 21 on Aristophanes of Byzantium 144 n. 67 on Attic rural economy 189 on Carthaginian slavery 260, 261 on freedom 57, 72, 73, 85, 88

OUP CORRECTED PROOF – FINAL, 28/6/2018, SPi

366

General Index

Finley, Moses I. (cont.) on gender imbalance in Homeric slavery 114 n. 37 on Greek slavery 14–15, 77, 89–91, 120, 270 on helots 128, 144 on Homeric poems 107, 108, 109 on the katakeimenos (debt bondsman) 73, 74 on the Mainz Academy 18 on Near Eastern slavery 102 on property 110 n. 16 on the rise in slavery due to end of debt bondage 107–8 on slavery in Homeric period 5, 14, 65, 107–8, 110 n. 16 on slaves as outsiders 27 slave societies/slaveholding societies 8, 93, 94, 95, 96, 99–100, 101, 269, 292 spectrum of statuses 57, 70, 72, 73, 75–6, 83, 87 n. 18, 101, 291 Finnegan, Ruth 111 Fisher, Nick 31 n. 21, 112, 123 n. 70, 181–2 Flaig, Egon 98, 259 n. 2, 262 n. 10, 266 Flower, Michael 126 Fogo 97 Foxhall, Lin 31 n. 21, 186–8 freedom 32, 57–79, 86, 291 Fynn-Paul, Jeffrey 274, 275, 282 Gaga 87 Gagarin, Michael 156 n. 32 Galil, Gershon 20, 232, 233 Gallant, Thomas 192 n. 120 Gallego, Julián 188 n. 102 Gamarth 265 n. 25 Garlan, Yvon 17, 27 n. 9, 114 n. 37, 127, 181 Garnsey, Peter 66 n. 28 Gehazi 218 Gehrke, Hans-Joachim 164 Genesis, Book of 214–15 Genovese, Eugene 132 Gideon 215 Goody, Jack 37, 85 Gordion 285 Gortyn 10, 73–5, 83, 88, 147, 150, 153, 296 n. 6 economy 157–9 Great Code of Gortyn 83, 150–7, 158, 164 Little Code of Gortyn 158 Granovetter, Mark 53 n. 111 Greece (archaic) 107–24 Dark Ages 107 debt bondage 89, 90, 107 douleia/doulos 60, 64, 65 economic role of slavery 82, 114–19 eleutheria/eleutheros 60, 64, 65 elites 108, 116, 119, 120, 122, 292 freedom 60–2, 65, 90 gender balance of slaves 114 n. 37

Greek slaves 199, 288 hired labour (thetes/pelatai) 89, 90, 107 n. 4, 116, 117, 118, 119, 272 metaphorical meaning of slavery 61–2 non-Greek slaves 89, 107, 120 punishment of slaves 45 n. 69, 112 sea raiding 120, 122 silver 120 n. 56, 121 slave–master relations 109–14 slave numbers 119 slaves as property 61, 64 n. 22 slave status 61, 109–14, 122–4 Greece (classical) 8 manumission 71 monoglot slave populations 101 n. 25, 144, 145, 160, 165 ownership, legal conception of 30–1 paramone 51, 71, 72–3, 235 property ownership 38 punishment of slaves 45 n. 69, 194–5 slaves as property 25, 54, 59, 83 slave status 72–6, 77–8 therapontes 183 n. 79 see also Attica (classical); Athens; Crete; Sparta; Thessaly Groupe International de Recherche sur l’Esclavage dans l’Antiquité (GIREA) 18–19 Gschnitzer, Fritz 296 n. 6 Gsell, Stefan 259 n. 2 Guía, Miriam 188 n. 102 Gundlach, Rolf 18 Guzanu 50, 244 Gymnetes see under Argos Hahn, István 270 Haifa 199 Halif, Tell 213 n. 56 Hall, Jonathan 90 n. 30 Halle Focus Group 37 Hammurabi 48, 208 n. 37 Hann, Chris 37, 38 n. 39, 53, 54 n. 112 Ḫ anna 52 Hannibal 262 n. 11, 265 Hanno the Great 265 Hansen, M. H. 188, 193 Hargu 289 n. 71 Harpalus 44 Harper, Kyle 28 n. 12, 98, 103 n. 30, 137 n. 47, 229, 243 n. 44, 270, 272, 284, 295, 299, 302 Harran census 232–3 Harris, Edward M. 28 n. 12, 85–6, 98, 107 n. 5, 108, 111, 112, 114 n. 37, 156 n. 32, 189 Harrison, Alick 30 n. 18, 47 n. 77 Harvey, F. David 135 Ḫ ašdija 51 Hasdrubal Gisco 261 Hasdrubal the Boetharch 266

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

General Index Haussmann, Baron 35–6, 141 Head, Ronan 87 n. 18, 244 n. 53 Heinen, Heinz 17 n. 46, 18 Hellas 248 helots see under Sparta Henkelman, Wouter 254 Heraclea Pontica Mariandynoi 138, 143, 144, 165, 293 Herodotus 62 n. 17, 256, 285, 303, 305 n. 39 Hesiod 107, 108 n. 9, 120 Works and Days 109, 117–18 Hezser, Catherine 20, 214, 215, 253 n. 35 Ḫ ibta 51 Hickey, Robin 29 n. 15 Hierapytna 150 n. 10 Hinz, Walther 255 Hipponicus 300 n. 19 Hodkinson, Stephen 129 n. 19, 130, 138, 142, 175, 188 n. 100, 189 n. 106, 297 Homer 60, 61, 64, 65, 90, 107, 108, 110–14, 119, 120 dmoes/dmoai 110, 112 as historical source 109 Iliad 110 Odysseus’ estate 114–17 Odyssey 44, 109, 111 thetes 116 xenoi 115, 116 Honoré, Antony M. 29 n. 15, 34–6, 39, 54, 140 Hopkins, Keith 94, 95, 97, 99–100, 269, 281 Hopper, Matthew 98 Hornblower, Simon 131 n. 24 Hoyos, Dexter 264 n. 24 Hume, David 301 Hunter, Virginia 40 n. 44 Hymettus 175 Hypereides 299 Iasus 83 Iddina 86, 87 Iddin-Marduk 244 n. 53 Iddin-Nabû 241 Illyria prospelatai 100, 101 n. 25 Ina-Esagil-ramât 241 Ina-ṣilli-abulli 87–8 Institute for the Study of Slavery (ISOS) 19 Ion of Chios 301 Iqiša 51 Iran see Persian Empire Isaac 214, 215 Isaeus 38 Isaiah, Book of 212, 231 Isaios 299 Islam 275 Ismard, Paulin 90 n. 30, 273 Isocrates 177 Israel 3, 6, 102, 103, 104, 199–222, 275, 289 n. 74

367

concubines 206–7, 209 ear-piercing of slaves 205, 206, 208 economic role of slavery 211–21 elites 212, 214, 219, 221 Greek slaves 199 ḫ abiru 205 n. 27 Hebrew slaves, treatment of 204–11 Hebrew word for master 85 Hebrew word for slave 14, 84, 199 n. 1, 202, 263 hired labour 202, 208, 209, 210, 213, 214, 222 indentured servants/bondsmen 203–4, 206, 220 jubilee 209, 210 kingdom of Israel 200, 211 manumission 203 slave ownership 214–17 slave status 201–2 slave terminology 202–4 Ištar-mukin-apli 51 Itti-Marduk-balat ̣u 50, 87, 238, 240, 289 n. 73 Itti-Uraš-pāniya 244 Jacob 214, 215 Jameson, Michael 181, 182, 192, 296 n. 8 Jeanmaire, Henri 129 n. 17 Jedaniah 253 Jeremiah, Book of 210, 214, 219, 220, 275 Jeroboam II, king 217, 218 Jerusalem 231 Temple of Solomon 208 Job, Book of 221 Johnson, Mark 66 Johnston, Alexandre 91 n. 35 Jones, A. H. M. 181 Jones, Christopher 40 n. 43 Josiah 208 Judah, kingdom of 6, 104, 200, 202, 210, 211, 212, 219, 220; see also Israel Judges, Book of 215 Jursa, Michael 103 n. 32, 237 nn. 10 and 12, 239, 242 n. 39, 243–4, 245 Justin Martyr 253, 265 Kalba 50 Kalhu 223 n. 2, 224 n. 4, 227, 231, 232 Kallikyrioi see under Syracuse 143 Kalzu 232 Kamen, Deborah 40 n. 43, 76, 83, 84, 299 Katsari, Constantina 7 Kaudos 162 Kellogg, Danielle 190 n. 112 Kennell, Nigel 129 n. 19 Keos 138 Kephisodorus 289 n. 73 Kina 52 King, Philip 215 n. 65 Kings, Books of 215, 217, 218, 231

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

368

General Index

Kiš 243 n. 43 Kleber, Kristin 254 Kolobova, K. M. 297 n. 9 Korea 98 n. 17 Korynephoroi see under Sicyon Kränzlein, Arnold 38 n. 40 Kron, Geof 184 kurtaš see under slavery (and servitude), terminology Kyrtatas, Dimitris 272 n. 8 Laconia see Sparta Lakoff, George 66 Lampis 300 Lancel, S. 264 Langdon, Merle 121 n. 59 Lato 150 n. 10 Lauffer, Siegfried 179 Laurion 121, 180, 250, 272 Lavan, Myles 66 n. 28 Lavreotiki 178 Lawrence, T. E. 98 n. 16 League of Nations 26 Lemaire, André 20, 217–18 Lenski, Noel 90 n. 29, 93 n. 1, 97 n. 10, 98, 101, 259 n. 2, 292 Leocrates 177 n. 49 Levinson, Bernard 204 n. 25, 205, 210 n. 45 Leviticus, Book of 202 n. 17, 210, 220 Holiness Code 205 nn. 25 and 28, 209, 210 Lévy, Edmond 185 Lewis, David Martin 113 n. 30, 255 n. 48 Libys 121 n. 59 Liddell and Scott’s A Greek-English Lexicon 305 Lindos 278 Link, S. 151 Lipit-Ištar 48 Lipsius, Hermann 151 Lloyd, Geoffrey 76 n. 55 Lotze, Detlef 127, 144 Lovejoy, Paul 97, 282 Lukonin, Vladimir 256 n. 51 Luraghi, Nino 129 n. 17, 132 n. 26, 145, 159 n. 47 Lycians 255 Lycurgus 126, 130, 298 Lykidas 299 Lysias 171, 297, 298 Lysicles 177 Lyttos 150 n. 10, 163 MacDowell, Douglas 129 n. 16 McHugh, Maeve 188 n. 98 McKeown, Niall 18 n. 51 McMahon, April 65, 296 Mactoux, Madéline 299 n. 15 Madanu-bel-uṣur 50, 235 Magdalene, Rachel 20, 88

Mago 260, 261 n. 9, 263, 265, 266 Mahseiah 253 Mainz Academy 17 Malinowski, Bronislaw 37 Marduk-iqišanni 243 n. 43 Marduk-nāṣir-apli 241, 244 Mariandynoi see under Heraclea Pontica Marx, Karl 13, 15–16, 31, 94 n. 4 Massinissa 262 n. 11 Masson, Michael 66–7 Mathus 260 Matilla Vicente, Eduardo 20 Mauritius 97 Megalopolis 264 Menander 184 n. 84, 285 Mendelsohn, Isaac 15 n. 39, 19, 212–13, 223 Menon 178 Mephibosheth 216 Meshech 199, 250 Mesopotamia see Assyria; Babylonia Messenia 133, 134, 136 n. 42, 142, 148 Meyer, Eduard 5, 12–13, 44 n. 66 Mibtahiah 253 Micah, Book of 212 Midas 46–7 Mielziner, Moses 201 n. 13 Milesians 257 Mirhady, David 45 n. 70 Missiou, Anna 85 mnoia see under slavery (and servitude), terminology Momigliano, Arnaldo 3 n. 7 Morris, Ian 32, 108 n. 7, 109, 110, 111 Morrow, Glen 129 n. 15 Müller, G. G. W. 229, 230 Murašû archive 236, 244 n. 53 Murašû family 238, 240, 241–2, 243, 245 Murray, Oswyn 108 n. 9, 122 n. 64 Mušezib-Šamaš 51 Mušallim-Issar 225, 227 Myron of Priene 131, 136 Mysia 248 Naaman 218 Na’aman, Nadav 205 n. 27 Nabopolassar 236 Nabu-aḫḫē-iddin 50, 240 Nabû-ēt ̣ir 235 Nabu-mušetiq-uddi 51 Nabu-šumu-iškun 227 Nabu-šum-ukin 52 Nabu-tuklatua 227 Nakhtḥor 251, 252 Nana-reṣua 52 Nana-silim 51 Nappāḫu family 238, 240, 241 Nausycides of Cholargus 178 Naxos 62, 99 Ndoye, Malick 112, 113 n. 34, 114 n. 37

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

General Index Neaera 67, 68 Near East 2, 288–90 cuneiform sign for slave 14 economic role of slavery 102, 104 freedom 89 metaphorical meaning of slavery 291 relative unimportance of slaves in Near East 102, 223, 249 slaves 81–2, 84–5, 102, 103, 104 slave societies 104 see also Assyria; Babylonia; Israel Nebuchadnezzar 220 Nehemiah, Book of 204, 220–1 Nergal-reṣua 235 Nicholas, Barry 40 n. 44 Nicomedes 67 Nidinti-Bēl 241 Nieboer, Herman 270 Nineveh 223, 231, 232, 236 Nippur 240, 241 Nupta 86, 87 Nur-Šamaš 51 Nūr-Sîn 50, 244 n. 53 Ober, Josiah 287 n. 65 Oded, B. 231, 232 Odrysians 284 n. 53 oiketes see under slavery (and servitude), terminology Oman 98 Oppian 302 n. 28 Osborne, Robin 45 n. 70, 175, 179, 183 n. 79, 190 n. 112, 193 n. 123 Pancleon 67–8, 69 Paphlagonia 248 Paradiso, Annalisa 135 Pasion 171, 177, 179 Passow, Franz 305 n. 39 Patterson, Orlando 28–9, 30, 32, 37, 70, 96, 97, 113 n. 33, 114 n. 37, 117 n. 46 Pausanias 64, 128, 129, 130 Penestai see under Thessaly Pergamum 195 Perlman, P. 149 Persepolis 254 Persepolis Fortification Tablets 7, 2 54, 257 Persepolis Treasury Tablets 254 Perses 117 Persian Empire 7, 236, 247–58, 289 ethnicity of slaves 255–6 kurtaš 255 see also Anatolia; Egypt; Fars Petosiri 253 Phainippos 176 Phano 68 Phormion 179, 299 Phormisius 187, 188

369

Phrygia 169, 199, 248, 250, 280, 284, 285–6, 288 Piraeus 172 Pisidia 248 Pittalacus 297, 300 Plataea 297 n. 9 Plato 126, 138, 144, 145, 297, 298, 304 Plutarch 127, 129, 134–5, 253 Polemarchus 171 Pollux 126, 143–4, 145–6, 285 Polybius 260, 265, 266 Pompey 284 Porten, Bezalel 253 Pospisil, Leopold 37 Postgate, Nicholas 85, 224 n. 8, 227, 233 property ownership 30–9 propolos see under slavery (and servitude), terminology prospelatai see under Illyria Protarchos 278, 289 n. 73 Proteas 174 Proverbs, Book of 204, 218–19 Pudija 52 Pylos 136 n. 42 Pylos tablets 114 Pythias 172 Pythionike 44 Qoheleth (Ecclesiastes), Book of 221 Raaflaub, Kurt 31, 59, 61 nn. 12 and 13, 82, 89, 91, 110, 114 n. 37 Radner, Karen 20, 224 n. 4, 229, 230 Rainey, Anson 217, 218 n. 79 Rankine, Patrice 26 n. 4 Rheneia inscription 278 Rhodes 140, 280 Rihll, Tracey 101, 108 n. 6 Rimanni-Bel 51 Rimut-Ninurta 241 Rocca, Francesca 61 n. 12 Rome property law/ownership 30, 31, 38 servus vicarius 113 slave numbers 103 slave prices 171, 272 n. 9 slaves 3, 281, 283–4 slave society 98, 103 Roscoe, P. J. 281 Rose, Peter 186 n. 94 Rosivach, Vincent 248 Roth, Ulrike 18 n. 51 Sahlins, Marshall 38 n. 38 Said, Edward 15 n. 40 Sallares, R. 181 Samaria ostraca 217, 218 Samšek 251 Samuel, Books of 215, 216 Ša-Nabu-taqum 52

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

370

General Index

Santiago (Cape Verde) 97 Saracens 98 Sardis 256 Sargent, R. 95 Sargon II 232 Saul, king 216 Scheidel, Walter 270, 271, 272 n. 9, 276 n. 30 Schmidt, Martin 112 Schumacher, Leonard 19 Scipio 262 n. 11, 265 Scipio Aemilianus 266 Scyrus 62 Šēḫ Ḥ amad, Tall 224 n. 4 Sekunda, N. 248 Selinus 265 Šellebi 241 Sennacherib 227, 231 Seuthes 100 Seuthes II 284 Seychelles 97 Sicyon Korynephoroi 144 Silver, Morris 169 n. 8 Silwan 206 n. 32 Sinclair, R. 181 Sinope 44 Sippar 48 temple of Ebabbar 236 Sirach 221 n. 86 slavery economic role: Assyria 223, 234; Attica 173–80, 189, 192–3; Babylonia 243–5; Carthage 263–6; Crete 157–9; Greece (archaic) 82, 114–19; Israel 211–21; Near East 102, 104; Sparta 142–3 generating elite wealth 90, 95, 100, 108, 173, 269, 293 ideal environment for slavery 287 legal status 25, 27, 32, 61, 290–1, 294; Assyria 225–6; Attica 68–70; Babylonia 87, 88, 89; Carthage 260–3; Crete 73–5, 83, 88, 153–5; Greece (archaic) 61, 109–14, 122–4; Greece (classical) 72–6, 77–8; Israel 201–2 manumission: Attica 44, 46, 83–4, 140, 141; Babylonia 51, 235; Carthage 262; Israel 203; Sparta 128, 129, 135, 137, 139, 140 metaphorical meaning of slavery 296, 299; Assyria 225; Attica 62–4, 65–6, 291; Carthage 263; Greece (archaic) 61–2; Israel 203, 216 monoglot slave populations 101 n. 25, 144, 145, 160, 165, 297 punishment of slaves: Attica 40–1, 44–5, 141, 194; Crete 152; Greece (archaic) 45 n. 69, 112; Sparta 131, 137; Thessaly 194, 195 sale of slaves: Assyria 223–4, 225; Attica 44;

Babylonia 48, 50–2; Crete 155; Sparta 128, 129, 130, 137, 140 size of slave population a threat to social order 103 slave concubines/female slaves 206–7, 209 slave marriages and families 43, 155, 157, 160, 262, 278–9, 289 n. 73 slave numbers: Assyria 227, 234, 238–9; Attica 95, 103, 119, 171, 179, 281 n. 45, 301; Babylonia 103; Carthage 265; Crete 160; Greece (archaic) 119; Persian Empire 255; Rome 103; Sparta 103, 119, 133, 143, 160 slave ownership: Anatolia 253; Assyria 226–30, 288; Attica 25, 167, 168–93; Babylonia 48–9, 238–9, 288; Carthage 264, 266; Crete 153, 155–6; Israel 214–17; Sparta 127–9, 130–1 slave ownership marks (branding, etc.) 40, 49, 205, 206, 208, 251, 252, 253 slave prices 271, 281–2, 287; Assyria 228–30, 239, 288; Attica 170–1, 182 n. 78, 228, 239, 271, 288; Babylonia 44, 288; Egypt 271; Rome 171, 272 n. 9 slave rebellions 132–3, 148–9, 159, 297 slave rights 42–3 slavery versus other labour options 271–4, 294 slave societies 93–104; Attica 98, 103, 143; Carthage 259; Crete 164–5, 269, 292; Rome 98, 103; Sparta 143 slave trade 167, 250, 251, 252, 261–2, 274, 277–82, 284–6 slave transport costs 276–7, 288–9 slaving/no-slaving zones 274–5 sources of slaves 274–6 slavery, definition 25–9 absence of term 35, 272; Assyria 225; Athens 46; Babylonia 51, 52; Crete 156 liability to execution 35; Athens 47–8; Babylonia 52 prohibition of harmful use 35; Athens 46–7; Babylonia 52; Crete 156 right to capital 35; Assyria 225; Athens 44; Babylonia 50; Crete 155; Sparta 140 right to income 35; Athens 43; Babylonia 50; Crete 154 right to manage 34; Athens 43; Babylonia 50 right to possess 34; Athens 39–40; Babylonia 48; Crete 153 right to security 35; Assyria 225; Athens 39–40, 44–5; Babylonia 49; Crete 155; Sparta 141 right to use 34; Athens 40–2; Babylonia 49; Sparta 140 transmissibility 35; Athens 46; Babylonia 51; Crete 156

OUP CORRECTED PROOF – FINAL, 28/6/2018, SPi

General Index slavery (and servitude), forms chattel 9, 125 debt bondage 9, 10, 46 n. 75, 72, 73–5, 89, 90, 107, 154, 203 n. 22, 226 helotic 11, 121, 143–6, 147, 165, 292 indentured servants 203–4, 206 manumitted slaves with paramone obligations 72–3, 235 serfs 11, 125, 127, 128, 147, 151, 164 n. 62, 165, 293 temple servants 87–8, 213, 237 n. 11 slavery (and servitude), terminology 84–6 amphipolos 301 andrapoda 304 n. 38 apeleutheros 73, 77 aphamiotai 150 n. 10 Assyrian word for slave 85 basilikoi laoi 251, 257 chrysonetai 150 n. 10 demosioi (public slaves) 273 dolos and woikeus 150–3 doulos and oiketes 53, 153, 182, 183, 296–305 exeleutheros 77 Hebrew word for slave 14, 84, 199 n. 1, 202, 263 kurtaš 255 mnoia (public slaves) 150 n. 10, 157 propolos 301 Punic word for slave 263 therapontes 183 n. 79, 304 n. 38 see also Crete, klarotai; Sparta, helots; Thessaly, Penestai slavery, uses agriculture 114–17, 142, 148 n. 3, 165, 169, 173–6, 180, 181–93, 218, 223, 264, 266, 272 apprentices (crafts) 43, 244 banking 169, 179, 180 business agents 244–5 civic tasks 273 ergasteria (sweatshops) 177–8 mining and quarrying 178–9, 272 prostitution 42, 50 soldiers 135, 136 textile manufacture 161, 176, 219 workshops 169, 178, 180, 181 Snell, Daniel C. 3 n. 9 Socrates 63, 64, 173 Solomon, king 221 Solon 28, 61–2, 89, 101, 107, 117 n. 48, 121, 123, 138, 274 Sophocles 177 Sosin, Joshua 72 n. 40 Sparta 6, 9 citizenship 142–3, 163, 173 communal use of private property 130, 137 economic role of helots 142–3 helot massacre 135–6 helot numbers 103, 119, 133, 143, 160

371

helot ownership 127–9, 130–1 helot rebellions/threat of 133, 148, 159 helots 6, 11, 31, 44, 72, 91, 99, 100, 101, 121, 125–46, 148, 290, 292 helots as soldiers 135, 136 hypomeiones 159 krypteia 131, 133–4 land ownership 127, 129 long hair for free men 70 manumission of helots 128, 129, 135, 137, 139, 140 origins of helots 145 phiditia (mess groups) 142, 162 punishment of helots 131, 137 sale of helots 128, 129, 130, 137, 140 slave society 143 state intervention in slavery 130, 132, 133–4, 136, 138, 140, 141 state ownership of slaves 127–8 xenelasia 149 Spendius 260 Sphakteria 136 n. 42 Spittler, Gerd 28 n. 10 Stager, Lawrence 215 n. 65 Stalin, Josef 16 Starr, Chester 104 Stephanus 68 Stolper, Matthew 254 Strabo 128, 129, 130, 283 Straton, king 253 Struve, Vasily 16, 27 Šulaja 240 Šulmu-šarri 227 Šumma-ilani 227 Susa 257 Swain, Simon 302 Swoboda, Heinrich 144 n. 67 Syracuse Kallikyrioi 143, 148 n. 3, 293 Syria 9, 280, 281 Tabal, kingdom of 199, 250 Tabbanea 51 Tabi 253 Teos 141 Thalmann, William G. 64 n. 22, 108, 111, 112, 113 n. 27 Themistocles 250 n. 20 Theocrines 40 Theophrastus 63, 169, 190 Theopompus of Chios 126, 145, 253 therapontes see under slavery (and servitude), terminology Thessaly 6, 9, 101, 119, 122 Penestai 103, 138, 143, 145, 148, 165, 293 punishment of slaves 194, 195 slave rebellions 159 Thommen, Lukas 129 n. 17 Thompson, Dorothy 55 n. 115

OUP CORRECTED PROOF – FINAL, 25/6/2018, SPi

372

General Index

Thompson, E. P. 168 n. 3 Thonemann, Peter 285 Thorner, Daniel 192 n. 120 Thrace 9, 100, 169, 250, 255, 276, 280, 281, 284–5 Thucydides 62, 100, 135, 136 Tiglath-pileser III 227 Timaeus of Tauromenium 261 Timarchus 70, 173, 297 Tissaphernes 83, 257 n. 56 Todd, Stephen 32–3 Tompkins, Daniel 16 Torah ‘slave’ laws 6, 10, 11, 201, 202 n. 17, 203, 204–11, 222 Covenant Code 204, 208, 209, 210 Tordoff, Robert 192 n. 120 Trevelyan, George 119 n. 51 Tubal see Tabal, kingdom of Tyre 199, 250, 251, 253, 288 Ubru-Aššur 226 Ulpian 302, 303 Uruk temple of Eanna 236 US slavery 27, 93, 94, 96, 137 n. 47, 140, 165, 170 n. 15, 171, 239 n. 23 Van der Spek, Robartus J. 20, 87 n. 18 Van Koppen, Franz 274 Van Wees, Hans 95, 98, 108, 109, 115, 117, 123 n. 70, 295, 300, 301, 304 Vidale, Massimo 112 n. 26 Vinogradoff, Sir Paul 31, 38 Virafša 252 Vlassopoulos, Kostas 26 n. 6, 58, 62 n. 17, 64 n. 22, 65, 86, 91 n. 34, 94 n. 5, 102 n. 28, 265 n. 26 VOC see Dutch East India Company Vogt, Joseph 13, 18

Wadi Daliyeh papyri 52, 253 Wallon, Henri 11–12 Weiler, Ingomar 18 n. 54 Welles, C. Bradford 38 n. 39 Wells, Bruce 210 Westbrook, Raymond 206, 210 Westermann, William L. 7, 13–14, 82, 84, 301 Whitby, Michael 135 Whittaker, C. R. 260, 261 Whybray, Roger 213 n. 55 Wickert-Micknat, Gisela 114 n. 37 Wiedemann, Thomas 18 n. 51, 19, 111 n. 18 Winn, Philip 97 Wittgenstein, Ludwig 54, 60 n. 10 woikeus see under slavery (and servitude), terminology Wolters, Al 218 n. 80 Wood, Ellen M. 30, 38, 174 n. 29, 181, 182–3, 185, 192, 295, 304 Wunsch, Cornelia 20, 88 Xenophon 49, 130, 131, 173–4, 257, 261, 284, 303–4 Xerxes 236, 249 Yadin, Yigael 217 Yawan 199, 250 Zababa-iddin 86, 87 Zabdî 226 Zedekiah, king 219–20 Zel’in, K. K. 27 Zelnick-Abramovitz, Rachel 32, 54, 57, 59, 70, 71, 72, 147 n. 2 Ziba 216 Zurbach, Julien 19 n. 59, 108 n. 6, 112 n. 24, 114 n. 35, 123, 138 n. 48. 165 n. 63