Social Inequality in Early Medieval Europe: Local Societies and Beyond (Haut Moyen Age) (Haut Moyen Age, 39) 9782503585659, 2503585655

Citation preview

Social inequality in Early Medieval Europe

Collection Haut Moyen Âge dirigée par Régine Le Jan 39

Social Inequality in Early Medieval Europe Local Societies and Beyond

Edited by Juan Antonio Quirós Castillo

F

© 2020, Brepols Publishers n.v., Turnhout, Belgium. 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, electronic, mechanical, photocopying, recording, or otherwise without the prior permission of the publisher. D/2020/0095/57 ISBN 978-2-503-58565-9 E-ISBN 978-2-503-58566-6 DOI 10.1484/M.HAMA-EB.5.118103 ISSN 1783-8711 E-ISSN 2294-8473 Printed in the EU on acid-free paper.

Contents

Preface 7 Equal and Unequal Societies in Early Medieval Europe. An Introduction Juan Antonio Quirós Castillo 11 1 State Formation and Socio-Political Complexity Towards an Archaeology of State Formation in North-Western Iberia Julio Escalona

33

Maize, Mounds, and Cosmos. Durable Inequality in the Mississippian World (ad 1000–1250) Robin A. Beck

55

2 Economic Specialization, Elite Demand and Social Inequality Pottery as Inequality? Systems of Production and Distribution in North Italian Societies During the Early Middle Ages Sauro Gelichi

75

Social Complexity in Peripheral Areas in the Light of Pottery Production Between Sixth and Tenth Centuries (Alava, Basque Country, Spain) Francesca Grassi

99

Indications of an Estate Economy from a Renewed Analysis of Sites of Rural Settlements Edith Peytremann

131

Social Complexity in Local Communities During the Tenth Century in Central-Northern Portugal. Negotiation and Opposition Catarina Tente

149

6

co n te n ts

The Primitivism of the Early Medieval Peasant in Italy? Richard Hodges

165

3 ‘Small World’ and Social Inequalities Land, Oxen, and Brooches. Local Societies, Inequality, and Large Estates in the Early Medieval Ardennes (c. 850–c. 900) Jean-Pierre Devroey & Nicolas Schroeder

177

Fiscal Lands, Rural Communities and the Abbey of Nonantola. Social Inequality in Ninth-Century Emilia (Italy) Igor Santos Salazar

203

Rural Communities and Landscapes in Northern Italy (Ninth–Twelfth Centuries ad) Fabio Saggioro

227

Unequal Small Worlds. Social Dynamics in Tenth-Century Leonese Villages Iñaki Martín Viso 255 Collective Action and Local Leaderships in Early Medieval North-Western Iberia. Ninth-Eleventh Centuries Álvaro Carvajal Castro

281

Village Formation, Social Memories and the Archaeology of Rural Communities in North-Western Iberia Juan Antonio Quirós Castillo

301

The Archaeology of the Peasant Mode of Production. Peasant-Based Societies in Central and Northern Iberia During the Early Middle Ages Carlos Tejerizo-García

331

Conclusion Chris Wickham

355

Preface

The origin of this volume lies in the international symposium held on 22 and 23 September 2016 in Vitoria-Gasteiz (Spain) under the title ‘Archaeology of Social Inequality in Early Medieval Europe: A Tribute to Chris Wickham’. This initiative was preceded by other meetings and events dedicated to celebrating not only Chris’ intellectual career, but also his humanity and personal qualities. In March 2014, it was held in Oxford the 6th International Seminar of the ‘Centro Interuniversitario per la Storia e l’Archeologia dell’Alto Medioevo’,1 of which the proceedings were dedicated to Chris on the occasion of his 65th birthday.2 A few months later, in January 2015, a meeting was held in Rome in his honor in which mainly Italian and British researchers participated,3 which gave rise to a volume recently published in Oxford.4 The following year a further symposium was organized in Oxford by his former colleagues and students in which mainly British scholars participated.5 Why then organize another event a few months later? Birmingham, Oxford and Rome-the United Kingdom and Italy-are the main places where Chris has worked during his long career-although they are by no means the only regions upon which he has worked. How then does the distant Basque Country fit with the preceding schedule of celebrations in which many European researchers participated? Chris Wickham’s publications and broad intellectual horizons are unconventional in European medievalism. He is a historian of medieval Italy, and in fact most of his publications have been devoted to that country. However, his works that have achieved greatest influence are those made in recent years from a comparative perspective, analyzing social processes and dynamics throughout Europe and the Mediterranean in the early to high Middle Ages: he has not hesitated to engage with other periods and regions when deploying the full potential of comparative social history. He has also devoted notable efforts to explicitly theoretical and methodological aspects,

1 http://saame.it/vi-seminario-internazionale/. 2 V. West-Harling, ed., Three Empires, Three Cities: Identity, Material Culture and Legitimacy in Venice, Ravenna and Rome, 750–1000. Volume offered to Chris Wickham as a gift for his 65th birthday, Brepols, Turnhout, 2015. 3 ‘Italy and Medieval Europe. A Fest in Honour of Chris Wickham’, British School at Rome 20th–23th January 2015, http://www.bsr.ac.uk/italy-and-medieval-europe-a-fest-in-honour-of-chris-wickham. 4 R. Balzaretti, J. Barrow, P. Skinner, eds, Italy and Early Medieval Europe. Papers for Chris Wickham, Oxford, 2018. 5 ‘Framing Chris Wickham’ (Birmingham and Oxford, 1976–2016: 27th of June 2016: St John’s College, Oxford).

8

p r e face

such as the relationships between history and archaeology, social anthropology, social history and historical materialism.6 It is precisely this latter aspect that facilitated the connection between his career and southwestern Europe and, more generally, with the historiography of the Middle Ages in the Spanish language, both in Spain and in Latin America. Social history and historical materialism in particular played an important role in the 60s and 70s in many historiographical traditions in southern Europe as well as in the emergent archaeology of historical societies. But this influence has been particularly noticeable in places like Spain, where intellectual renewal and the overcoming of the Francoist historical discipline has been remedied through social history.7 This influence has not been homogeneous across all historiographic fields, but it has certainly been relevant in medieval history — especially in early medieval history — and in the emergent discipline of Hispanic medieval archaeology.8 Although several European scholars had already been working on the Hispanic Middle Ages from these perspectives.9 it was the integration into the European Union in the 80s that opened the doors to a second stage of renewal of the social history of the Middle Ages in Spain by incorporating concepts, theories and models that were being debated in other European traditions. Since then, Chris Wickham’s work — along with that of other authors — has been translated into Spanish, including his principal publications over the last ten years. In addition, his involvement in Hispanic academic life through his participation in meetings and collective publications in Spain and South America has been growing.10 What is more, the rise of Anglo-Saxon medievalism and Chris Wickham’s work among specialists of the Hispanic Middle Ages in recent times has been proportional to the decline of the influence of French authors, who had had provided intellectual direction in previous decades.11 All this has been possible because, unlike other European traditions, social history remains one of the main



6 A selection of Chris Wickham’s major publications can be found in https://en.wikipedia.org/wiki/ Christopher_Wickham. 7 F. J. Moreno Martín, ed., El franquismo y la apropiación del pasado. El uso de la historia, de la arqueología y de la historia del arte para la legitimación de la dictadura, Madrid, 2016. 8 A. Barbero, M. Vigil, La formación del feudalismo en la Península Ibérica, Barcelona, 1978; R. Pastor, Resistencias y luchas campesinas en la época del crecimiento y consolidación de la formación feudal: Castilla y León, siglos X–XIII, Madrid, 1980; M. J. Hidalgo, D. Pérez, M. J. Gervás, eds, “Romanización” y “Reconquista” en la Península Ibérica: Nuevas perspectivas, Salamanca, 1998; M. Barceló, (ed.), Arqueología Medieval. En las afueras del “medievalismo”, Barcelona, 1998. 9 In particular French authors such as P. Guichard, Tribus arabes et berbères en Al-Andalus, Doctoral thesis read at the Université Lyon 2 (which was translated into Spanish in 1976, and then published in France with the title Structures sociales “orientales” et “occidentales” dans l’Espagne musulmane, Paris–La Haye, 1977); P. Bonnassie, La Catalogne du milieu du xe à la fin du xie siècle. Croissance et mutation d’une société, Toulouse, 1975–1976 (translated into Catalan in 1979 and in Spanish in a brief version in 1988). 10 Some of them are gathered in https://dialnet.unirioja.es/servlet/autor?codigo=9482. 11 P. Cressier and P. Moret, “La Casa de Velázquez y la Arqueología: algunos apuntes históricos”, in M. B. Deamos and J. Beltrán Fortes, eds, Las instituciones en el origen y desarrollo de la Arqueología en España, Sevilla, pp. 343–60.

pre face

intellectual axes of medievalism in Spanish even today.12 This background explains why a meeting was organised in Vitoria-Gasteiz, dedicated to Chris Wickham, focusing on the analysis of social inequality and why there were not only Hispanic authors, but also historians and archaeologists. Not all of the papers given at the conference are published here: several had either already been promised elsewhere or were not available for different reasons. Among the works that could not be included are those of Sonia Gutiérrez (illustrator, however, of the cartoon which accompanies this introduction), Andrew Reynolds, Dries Tys, Alfonso Vigil-Escalera and Anne Nissen Jaubert.13 However, other authors who, for different reasons, could not attend the meeting, have contributed to the volume, incorporating additional points of view and approaches. There is, therefore, not a direct correspondence between the structure of the conference and the ensuing monograph. While the symposium of Vitoria focused mainly on the archaeology of social inequalities, the contributions of several specialists in written sources (Álvaro Carvajal, Igor Santos, Jean-Pierre Devroey, Nicolas Schroeder) and archaeological records (Fabio Saggioro, Richard Hodges, Edith Peytremann) has resulted in a much more vibrant and articulated work in terms of geography, theory and methodologies. As in the case of the volume edited by V. West-Harling already quoted, also on this occasion Chris himself concludes the book with incisive and stimulating conclusions in which he rethinks those that he raised in Vitoria-Gasteiz. I would like to conclude these notes by thanking Regine Le Jan for accepting this volume in the prestigious Haut Moyen Âge series and all the authors for their contributions and patience. Juan Antonio Quirós Castillo University of the Basque Country

12 However, there are some authors who do not doubt to characterise this situation as anachronic (see J. Aurell, “Le médiévisme” espagnol au xxe siècle: de l’isolationnisme à la modernisation”, Cahiers de civilisation médiévale 48 (2005), pp. 201–18; J. Aurell, La escritura de la memoria. De los positivismos a los postmodernismos, Valencia, 2017). 13 The conference programme can be seen in https://socialcomplexitysite.wordpress.com/ programme/. Besides, the meeting was video recorded, so both the lectures and debates are available in https://ehutb.ehu.eus/series/58c672bef82b2bac2c8b456b.

9

10

p r e face

“Excuse me, we do not have a state, we are all free peasants”, by S. Gutiérrez Lloret.

Juan Antonio Quirós C astillo

Equal and Unequal Societies in Early Medieval Europe An Introduction

Introduction The 2018 Oxfam inequality report launched for the World Economic Forum in Davos (Switzerland) highlighted the sharp increase in inequality on a global scale over the last few years. According to this report, eighty-two per cent of the wealth generated in 2017 went to the richest one per cent of the global population, while the 3.7 billion people who make up the poorest half of the world saw no increase in their wealth.1 In the same way, authors like Branko Milanovic have pointed out that we live in the most unequal era of history, and that this is a highly globalised and interconnected phenomenon which cannot be dealt with at the national level.2 The effect of the world wars and the crisis of the 1920s made it possible to reduce inequality and build a social model which, at least in the developed world, allowed for the emergence of a middle class and the so-called welfare state. This model was deeply shaken first by the conservative revolution of the 1980s and, more recently, by the global crisis of 2007–2008. All of this has in turn led to the emergence of national populism, as well as an alarming increase not only in economic and geographical inequalities, but in intergenerational ones as well. For the first time in decades, today young people in Western countries will have a lower standard of living than their parents had.3 Growing social inequality has not only become one of the main concerns of progressive politicians, but also of the Social Sciences and the Humanities. In recent years, economists, sociologists, anthropologists and other experts have made important contributions to the analysis of social inequality in current societies from a dual perspective.4 On the one hand, the causes of the increase or decrease 1 https://www.oxfam.org/en/research/reward-work-not-wealth. 2 B. Milanovic, Global Inequality. A New Approach for the Age of Globalization, Harvard, 2016. 3 J. Brusuelas, The End of the Middle Class: What Went Wrong and What We Can Do about It, New York, 2014. 4 E. Margolis and M. Romero eds, The Blackwell Companion to Social Inequalities, New York, 2005; B. Nolan, W. Salverda and T. Smeeding eds, The Oxford Handbook of Economic Inequality, Oxford, 2011. Juan Antonio Quirós Castillo  •  University of the Basque Country Social Inequality in Early Medieval Europe: Local Societies and Beyond, ed. by Juan Antonio Quirós Castillo, Turnhout : Brepols, 2020 (HAMA 39), pp. 11–29 © FHG 10.1484/M.HAMA-EB.5.118443

12

j ua n a n to n i o q u i ró s c as t i l lo

in economic inequality throughout the twentieth century have been described and analysed, showing that there is no direct relationship between economic growth and the evolution of inequalities, as the Nobel Prize laureate Simon Kuznets already foresaw decades ago.5 On the other hand, scholars have also explored how it might be possible to overcome this situation via compensatory measures such as the creation of basic income, new forms of taxation, technological development, equal access to education, promotion of economic growth, etc.6 History and Archaeology have by no means stood on the sidelines of these debates. In fact, in recent years, work analysing the unequal distribution of resources in the pre-industrial period has multiplied from a plurality of perspectives, of which we can outline at least four main approaches. On the one hand, some scholars have explored the historical dimension and background of current economic inequalities, hoping in this way to identify patterns on which to formulate more solid theories and causal explanations. A long-term perspective would allow us to consider the importance of wars, pandemics and other natural disasters in reducing economic inequalities throughout history,7 as well as to explore the complex relationship between inequality and economic development.8 A second line of work, on the other hand, has analysed the emergence and consolidation of different forms of inequality in past societies from an evolutionary perspective, in terms of complexity and scale changes in socio-systems. Anthropological archaeology has taken the fore in this type of study, and has managed to build highly refined models.9 A third line of research aims to make long-term cross-cultural comparative analyses of wealth inequality by measuring the degrees of inequality in the past using instruments such as the Gini coefficient10 or Lorenz curves.11 The use of these theoretical and methodological approaches has revealed, for example, the existence of significant differences in terms of inequality between New and Old World societies, questioning some of the theories that until

5 T. Piketty, Le Capital au xxie siècle, Paris 2013. 6 See Piketty, Le Capital; Milanovic, Global Inequality; A. Atkinson, Inequality: What Can Be Done?, Harvard, 2015. 7 W. Scheidel, The Great Leveler: Violence and the History of Inequality from the Stone Age to the TwentyFirst Century, Princeton, 2017. 8 B. Milanovic, P. H. Lindert and J. G. Williamson, ‘Pre-industrial inequality’, The Economic Journal 121 (March 2010), pp. 255–72. 9 T. D. Price and G. M. Feinman ed., Foundations of social inequality, New York, 1995; T. D. Price and G. M. Feinman, 2010, Pathways to power: new perspectives on the emergence of social inequality, New York, 2010; S. Bowles, E. A. Smith and M. Borgerhoff Mulder, ‘The emergence and persistence of inequality in premodern societies: introduction to the special section’, Current Anthropology 51.1 (2010), pp. 7–17; A. W. Johnson and T. Earle, The evolution of human societies: From foraging group to agrarian state, Stanford, 2000; K. Flannery and J. Marcus, The Creation of Inequality: How our Prehistoric Ancestors set the Stage for Monarchy, Slavery, and Empire, Cambridge and London, 2012. 10 T. A. Kohler and M. E. Smith eds, Ten Thousand Years of Inequality: The Archaeology of Wealth Differences, Tucson, 2018. 11 K. I. Wright, ‘Domestication and inequality? Households, corporate groups and food processing tools at Neolithic Çatalhöyük’, Journal or Anthropological Archaeology 33 (2014), pp. 1–33; J. Bermejo Tirado, ‘Household Archaeology y el análisis de las sociedades antiguas en la peninsula Ibérica: definiciones, aplicaciones y posibilidades’, Materialidades. Perspectivas en cultura material 2 (2014), pp. 48–92.

eq ual an d u n eq ual s o c i e t i e s i n e arly  me d i e val e u ro pe

now had explained the formation of state structures, the increase and persistence of social inequality, the factors affecting the degree of inequality, and the variation in inequality across time.12 Another group of studies, with which the present volume is aligned, has focused on the analysis of inequalities not only in economic terms, but also in social, political, cultural or ideological terms as well. For example, a recent study devoted to defining the grand challenges facing archaeology in the next decade and beyond has identified among other priorities “why and how social inequalities emerge, grow, persist, and diminish, and with what consequences”.13 Again, the thematic and theoretical approaches employed have been very heterogeneous. In some cases social inequality has been analysed in specific territories, establishing connections between the processes of social hierarchy and the main historical processes.14 Inequality and social identities in the past have also been discussed through the analysis of specific groups, such as the seminal project “Les élites dans le haut Moyen Âge” (2003–2009), which has resulted in the publication of a large collection of monographs.15 The studies of gender or sexual inequality, race or ethnicity are also some of the topics around which there has been an intense scholarly production dedicated to the analysis of social inequality.16 The increasing interest in these topics can also be seen in the number of conferences organised in recent years, such as the meetings “Ruralia IX: Hierarchies in Rural Settlements”,17 “Social Inequality as a Topic in Archaeology”18 or “Measuring and Explaining Household Inequalities in Antiquity: Inequality from the Bottom Up”,19 among 12 W. Scheidel, ‘Stratification, deprivation and quality of life’, in M. Atkins and R. Osborne eds, Poverty in the Roman World, Cambridge, 2006, pp. 40–59; M. E. Smith, T. A. Kohler and G. M. Feinman, ‘Studying Inequality’s Deep Past’, in T. A. Kohler and M. E. Smith eds, Ten Thousand years of inequality. pp. 3–38; T. A. Kohler, M. E. Smith, A. Bogaard, C. E. Peterson, A. Betzenhauser, G. M. Feinman, R. C. Oka, M. Pailes, A. M. Prentiss, E. C. Stone, T. J. Denneby and L. J. Ellysoin, ‘Deep Inequality. Summary and Conclusions’, in T. A. Kohler and M. E. Smith eds, Ten Thousand years of inequality, pp. 289–317; M. E. Smith, ‘Quality of Life and Prosperity in Ancient Households and Communities’, in C. Isendahl and D. Stump, The Oxford Handbook of Historical Ecology and Applied Archaeology, Oxford, forthcoming. 13 K. W. Kintigh, J. H. Altschul, M. C. Beaudry, R. D. Drennan, A. P. Kinzig, T. A. Kohler, W. F. Limp, H. D. G. Maschner, W. K. Michener, T. R. Pauketat, P. Peregrine, J. A. Sabloff, T. J. Wilkinson, H. T. Wright and M. A. Zeder, ‘Grand Challenges for Archaeology’, American Antiquity 79.1 (2014), pp. 5–24. 14 See the recent book by E. Van Onacker, Village Elites and Social Structures in the Late Medieval Campine Region, Turhnout, 2017. 15 F. Bougard, G. Büher-Thierry and R. Le Jan, ‘Elites in Early Middle Ages. Identities, Strategies, Mobility’, Annales HSS 68.4 (October-December 2013), pp. 735–68. 16 For example, R. McGuire and R. Paynter ed., The Archaeology of Inequality, Oxford and Cambridge, 1991. 17 J. Klápštē ed., Hierarchies in rural settlements / Hierarchien in ländlichen Siedlungen / Des hierarchies dans l’habitat rural, Ruralia IX (26 September–2 October 2011, Götzis, Austria), Turhnout, 2013. 18 S. Hansen and J. Müller eds, Rebellion and inequality in archaeology: proceedings of the Kiel workshops ‘Archaeology of rebellion’ (2014) and ‘Social inequality as a topic in archaeology’ (2015), Habelt, Bonn 2017. 19 T. A. Kohler and M. E. Smith eds, Ten Thousand Years of Inequality.

13

14

j ua n a n to n i o q u i ró s c as t i l lo

others. Along these lines, several collective volumes have been published,20 and a number of research projects and networks have been created, as in the case of the “Inequality and History Network”.21 In short, social inequality has become a fruitful area for reflection and for the construction of new narratives in numerous disciplines, and remains one of the main vectors around which history and social archaeology are articulated from a multiplicity of perspectives.22 In this book dedicated to Chris Wickham we intend to explore this subject in the context of early medieval societies. The topic of inequality and social complexity explicitly runs through all of Chris Wickham’s work, regardless of whether the focus is settlement patterns, political dynamics, territorial or sub-regional analysis, cultural history, economic exchanges, or socioeconomic studies. In addition, the Early Middle Ages is a particularly interesting laboratory for the study of social inequality. Although each historical stage is interesting by itself, the half-millennium that runs between the end of the Roman world in the West and the year one thousand is a highly fluid and experimental period in political and social terms throughout Western Europe.23 The construction, fission, coalescence and collapse of political structures, the mobility of populations or the profound economic and cultural transformations that took place at the time created the conditions not only for the development of elaborate forms of social inequality, but also for the constant redefinition of social structures at different scales. The objective of this collective volume is therefore to explore, through several case studies, the phenomenon of social inequality in Western Europe in the Early Middle Ages based on both material and textual evidence. This brief introduction is divided into three main parts. Firstly, the structure and contents of the volume will be explained. Next, we will survey the main topics addressed by the different authors and, lastly, we will suggest new lines of work for the future.

20 For example, Timothy A. Kohler, Michael E. Smith, Amy Bogaard, Gary M. Feinman, Christian E. Peterson, Alleen Betzenhauser, Matthew Pailes, Elizabeth C. Stone, Anna Marie Prentiss, Timothy J. Dennehy, Laura J. Ellyson, Linda M. Nicholas, Ronald K. Faulseit, Amy Styring, Jade Whitlam, Mattia Fochesato, Thomas A. Foor and Samuel Bowles. ‘Greater post-Neolithic wealth disparities in Eurasia than in North America and Mesoamerica’, Nature 551 (2017), pp. 619–22; J. A. Quirós Castillo ed., Social complexity in Early Medieval Rural Communities. The north-western Iberia archaeological record, Archaeopress Archaeology, Oxford, 2016. 21 http://inequalityandhistory.blogspot.com/. 22 See for example the recent work by C. A. Hastorf, The Social Archaeology of Food: Thinking about Eating from Prehistory to the Present, Cambridge, 2017. 23 In the last few years, several interpretative essays have been published about the period, offering particularly interesting panoramas from very different perspectives. Among others, we might mention C. J. Wickham, Framing the Early Middle Ages: Europe and the Mediterranean, 400–800, Oxford, 2005; J. M. H. Smith, Europe after Roman: A New Cultural History, 500–1000, Oxford, 2005; B. Ward-Perkins, The Fall of Rome and the End of Civilization, Oxford, 2006; G. Halsall, Barbarian Migrations and the Roman West, Cambridge, 2007; E. James, Europe’s Barbarians, ad 200–600, Season Longman, London, 2009; M. Costambeys, M. Innes and S. MacLean, The Carolingian World, Cambridge, 2011.

eq ual an d u n eq ual s o c i e t i e s i n e arly  me d i e val e u ro pe

Contents and Outline of the Book The thirteen papers included in this volume are by scholars from seven different countries, and they represent a broad range of topics and approaches to the analysis of social inequality in the early medieval period from historical and archaeological perspectives. The reader will notice that most papers are centred on the study of Southern Europe (fig. 1), even though for comparative purposes other examples have been included from Carolingian areas, as well as one study on North America. As Chris Wickham has pointed out on multiple occasions, national frameworks have heavily conditioned the study of medieval societies in Europe, limiting our understanding of regional and transnational processes, falling into metanarratives of national identities or using teleological approaches.24 According to Wickham, societal comparison using powerful variables is the methodological and conceptual key to overcoming these risks.25 Although in this collection only some papers establish systematic comparisons, the use of strong theoretical frameworks and the exploration of numerous links with other territories, studies and discussions situate these single case studies within a larger context. Moreover, most of the papers address methodological issues, often combining different kinds of evidence. Given the nature and the diversity of the papers, they are not easy to group together into closed sections. Still, we can say that they are concerned with two main scales of analysis (states and elites, and local societies or “small worlds”26) and three major topics (socio-political complexity, economic specialisation, and social inequality in local societies). However, some papers necessarily span multiple categories. State Formation and Socio-Political Complexity

State formation has been one of the key topics in the study of social inequality and social complexity in Social Anthropology and other social sciences in the last decades.27 During this time, the traditional unilinear evolutionist approach to state emergence has been challenged by new multilinear and multiscale perspectives, giving rise to new lines of enquiry.28 Indeed, most scholars now acknowledge that state formation processes have laid the foundation for the stabilisation and reproduction of social

24 C. Wickham, Framing the Early Middle Ages. Europe and the Mediterranean, 400–800, Oxford, pp. 2–3, pp. 825–27. 25 C. Wickham, ‘Problems in doing comparative History’, in P. Skinner ed., Challenging the Boundaries of Medieval History. The legacy of Timothy Reuter, Turnhout, 2009, p. 5-28. 26 W. Davies, Small Worlds: The Village Community in Early Medieval Brittany, Oxford, 1988. 27 Indeed, this is one of the major challenges for Archaeology in the near future, according to K. W. Kintigh, J. H. Altschul, M. C. Beaudry, R. D. Drennan, A. P. Kinzing, T. A. Kohler, W. F. Limp, H. D. G. Maschner, W. K. Michener, T. R. Pauketat, P. Peregrine, J. A. Sabloff, T. J. Wilkinson, H. T. Wright and M. A Zeder, Grand Challenges for Archaeology, American Antiquity 79.1 (2014), pp. 5–24. 28 T. D. Price and G. M. Feinman, eds, Foundations of Social Complexity, New York and London, 1995; A. W. Johnson and T. Earle, The Evolution of Human Societies. From Foraging group to agrarian state, Sandford, 2000; K. Flannery and M. Marcus, The creation of inequality. How our prehistoric ancestors set the stage from monarchy, slavery and empire, Cambridge, 2012; T. D. Price and G. M. Feinman, eds, Pathways to Power. New perspectives on the Emergence of Social Inequality, New York, 2012.

15

Fig. 1. Map with the main European sites cited in the book.

16 j ua n a n to n i o q u i ró s c as t i l lo

eq ual an d u n eq ual s o c i e t i e s i n e arly  me d i e val e u ro pe

inequality following the notion of durable inequality proposed by Charles Tilly.29 Early medieval Europe is a very interesting laboratory for the study of this topic given the diversity and complexity of state formation and collapse processes attested after the fall of the Roman Empire, the existence of failed states, warlords and non-state societies, as well as the coexistence of different kinds of states (tax-based, land-based, non-states).30 In this book, there are two articles that address this topic. In his paper, Julio Escalona provides an up-to-date and wide-ranging perspective on the topic, exploring the different traditions existing in different areas of the continent, and underlining the characteristics of secondary state building in early medieval Europe. This is an important point in Southern Europe, where states and political formations emerged and collapsed in an environment dominated by the legacy of Roman political culture. He then focuses his attention on northern Iberia, where different states emerged and collapsed during the Early Middle Ages, showing how these processes provided abundant opportunities for social promotion, mobility and the establishment of durable inequality. In particular, he explores the convergence of different developments (direct inheritance of state structure and culture; reintroduction of statehood elements; “satellitisation” in the periphery of strong states, and incorporation of non-state societies) and their archaeological visibility through monumentality and landscapes of defence.31 The volatility of early medieval states in north-western Iberia paradoxically determined the great stability of elite domination and two-way agency in the process of building the Astur-Leonese kingdom. However, it is in North America where state formation theory has seen the biggest developments in recent years. In his paper, Robin Beck has analysed the formation of durable inequality using the Mississippian culture as a case study. The participation of an extra-European scholar in a conference and a book devoted to early medieval Europe is due to the will to introduce theoretically transversal perspectives into the discussion of social inequality, as well as to propose new avenues for the study of the topic from different backgrounds. Robin Beck’s study of Cahokia, one of the largest and most remarkable urban settlements of the Mississippian culture, provides a good example for the analysis of complex chiefdoms or small states.32 His theoretical starting point is the contraposition between status and class as an explanation for

29 C. Tilly, Durable Inequality, Berkeley, Los Angeles, London, 1998. In this book Julio Escalona and Robin Beck have explicitly used his framework. 30 C. Wickham, Framing the Early Middle Ages, pp. 56–150. 31 In the Vitoria-Gasteiz conference another two papers focused on state formation. S. Gutiérrez has explored state building in Al-Andalus from the perspective of Islamization processes (see S. Gutiérrez Lloret, ‘Early Al-Andalus: An Archaeological Approach to the Process of Islamization in the Iberian Peninsula (7th to 10th centuries)’, in S. Gelichi and R. Hodges eds, New Directions in Early Medieval European Archaeology: Spain and Italy compared. Essays for Riccardo Francovich, Collection Haut Moyen Âge 24, Turhnout, pp. 43–86). A. Reynolds has studied the formation of Anglo-Saxon states studying landscapes of governance and administrative structures (see A. Reynolds, ‘Judicial culture and social complexity: a general model from Anglo-Saxon England’, World Archaeology 45.5 (2013), pp. 699–713). Videos of their talks can be seen at https://goo.gl/CHfo4g (S. Gutiérrez Lloret) and https://goo.gl/yqVPii (A. Reynolds). 32 T. R. Pauketat, Chiefdoms and other Archaeological Delusions, Lanham, 2007.

17

18

j ua n a n to n i o q u i ró s c as t i l lo

the nature of the social change that occurred between Early Mississippian Cahokia and a later phase (1200 ce), likewise appealing to political economy frameworks and the economy bottleneck model proposed by T. Earle and M. Spriggs.33 Landscape transformation in the wake of environmental crisis created the opportunity for the emergence of central power and social hierarchy, when land control rather than labour created the framework for durable inequality. Economic Specialisation, Elite Demand and Social Inequality

A second group of papers is focused on economic and political perspectives, exploring, via archaeology, the connection between elite demand, economic specialisation and social inequality. This is another important issue in the European and American anthropological and archaeological agenda, since authors have established a direct connection between specialisation, the emergence of elites and the development of social inequality.34 However, some authors have pointed out that increasing socio-political complexity may be more closely linked to a growing variety of simultaneously existing forms of craft organisation, rather than the emergence of specialisations.35 The simplification of material culture and the socio-political characterisation of economic systems have played a central role in the recent interpretation of the European Early Middle Ages. Although projects such as “The Transformation of the Roman World”, financed by the European Science Foundation,36 have reviewed the most alarmist and pessimistic comparative approaches to this historical period, early medieval political economy continues to be the subject of debate among scholars.37 The study of pottery in terms of exchange systems and the transformation of production modes has been one of the main axes around which the processes of economic and political regionalisation, as well as growing social simplification, have been studied.38 The availability of a growing number of local studies on contexts of 33 T. Early and M. Spriggs, ‘Political Economy in Prehistory. A Marxist Approach to Pacific Sequences’, Current Anthropology 56.4 (2015), pp. 515–40. 34 E. M. Brumfiel and T. K. Earle ed., Specialization, Exchange and Complex Societies, Cambridge, 1987; C. L. Costin, ‘Craft Production Systems’, in G. M. Feinman and T. D. Price eds, Archaeology of the Millennium, New York, 2001, pp. 273–327; K. Flannery and J. Marcus, The Creation of Inequality. How our Prehistoric Ancestors set the Stage for Monarchy, Slavery and Empire, Harvard, 2014. 35 C. L. Costin, ‘Craft Economies of the Ancient Andean States’, in G. Feinman, G. M., Nicholas, L. M., and Skibo, J. M., eds, Archaeological Perspectives on Political Economies, Salt Lake City, 2004, pp. 189–223. (p. 218). 36 The outcomes of the project have been published in more than a dozen of seminal volumes: https:// brill.com/view/serial/TRW. 37 B. Ward-Perkins, The fall of Rome and the end of civilization, Oxford, 2005; R. Hodges, Dark Age Economics. A new audit, London, 2012; J. Moreland, ‘Concepts of the Early Medieval Economy’, in J. Moreland, Archaeology. Theory and the Middle Ages, London, pp. 75–101. 38 While there is a vast literature on the subject, the following list includes some of the most noteworthy: C. Panella, ‘Merci e scambi nel Mediterraneo in età tardoantica’, in A. Carandini, L. Cracco Ruggini and A. Giardina eds, Storia di Roma III.2. I luoghi e le culture, Turin, 1993, pp. 613– 97; M. McCormick, Origins of the European Economy. Communications and Commerce ad 300–900, Cambridge, 2001, pp. 53–60; P. Arthur, ‘Form, function and technology in pottery production

eq ual an d u n eq ual s o c i e t i e s i n e arly  me d i e val e u ro pe

ceramics consumption in Europe and the Mediterranean has allowed us to observe the processes of progressive fragmentation and depletion of the exchange system based on mechanisms of fiscal-based distribution from the Roman period, as well as to analyse the emergence of new systems that were more local and much less sophisticated. Experts have therefore investigated the extent to which consumption patterns, and the mechanisms of production and distribution of such modest products as pottery, are capable of reflecting historical background trends, in economic, cultural, social and political terms. The answers have been varied. Some authors have explicitly questioned the usefulness of pottery and objects of everyday use for analysing cultural and economic change during the Early Middle Ages.39 Other researchers are much more optimistic (or possibilistic) and consider that the study of pottery and other objects of everyday use constitutes a good guide for the definition of social categories through the identification of lifestyle signatures.40 However, it is Chris Wickham who has most explicitly established a connection between elite demand and economies of scale, relating the complexity of exchanges at the local and regional scales to elite wealth.41 What is more, according to Wickham, the complexity of pottery exchanges and the articulation of settlement hierarchies are two key guides to the analysis of social complexity in early medieval societies, and aid in the identification of aristocracies.42 Although some recent work emphasises the role of peasant demand in generating economies of scale,43 pottery remains a central theme in the study of economics and social inequality in early medieval Europe. Two papers in this book deal with this topic. Sauro Gelichi specifically asks to what degree pottery reflects social diversity in the Early Middle Ages in light of the urban and rural evidence in northern Italy from before the seventh century, when large-scale Mediterranean exchange ceased. In order to tackle the topic, he discusses, in the first place, the utility and the limits of the Peacock mode of productions approach, extensively used in the analysis of early medieval European wares.44 The main conclusion is that these models should from Late Antiquity to the Early Middle Ages’ in L. Lavan, E. Zanini and ([A-Z].)( )([A-Z][a-z] {1;})A. Sarantis eds, Technology in Transition A.D. 300–650, Leiden and Boston 2007, pp. 159–86; P. Reynolds, Hispania and the Roman Mediterranean, ad 100–700. Ceramics and Trade, London, 2010. 39 For example, C. Hills, ‘History and archaeology: the state of play in early medieval Europe’, Antiquity 81 (2007), pp. 191–200; J. Blair, Building Anglo-Saxon England, Princeton, 2018, p. 67. 40 C. Loveluck, Northwest Europe in the Early Middle Ages, c. ad 600–1150. A comparative archaeology, Cambridge, 2013, pp. 98–99. 41 C. Wickham, ‘Rethinking the Structure of the Early Medieval Economy’, in J. R. Davis and M. McCormick, The long morning of Medieval Europe: New Directions in Early Medieval Studies, New York and London, pp. 19–31. 42 C. Wickahm, Framing the Early Middle Ages, p. 493. 43 F. Theuws, ‘Early medieval transformations: aristocrats and dwellers in the pagus Texandria. A publication programme’, Medieval and Modern Matters 1 (2010), pp. 37–72 (p. 58); F. Theuws and M. van Haperen, The Merovingian cemetery of Bergeijk-Fazantlaan, Bonn, 2012, pp. 180–87. 44 The heuristic problems of these ethnographic-based typological approaches have been discussed by different authors, such as C. L. Costin, ‘Craft Production’, in M. Maschner and C. Chippindale eds, Handbook of Methods in Archaeology, Lanham, 2005, pp. 1032–1105 (p. 1037), and more recently by K. Duistermaat, ‘The Organization of Pottery Production: Toward a Relational Approach’, in A. Hunt

19

20

j ua n a n to n i o q u i ró s c as t i l lo

be reformulated, taking into consideration different variables when assessing the complexity of the production systems. Then, he compares pottery evidence from ancient emporia and cities in order to understand the connections between urban elite demand and social consumption patterns. The degree of local production standardisation, the importation of pottery, and the quality of the associated buildings show that the meaning of similar consumption patterns varies in different contexts. In emporia, some kinds of pottery could be identified as elements of distinction, but not necessarily of inequality. Conversely, in ancient cities the same evidence could indeed be related to social inequalities. Consequently, the signature value of pottery should be considered unstable. Similar conclusions have been reached by another scholars studying signatures such as faunal remains.45 If archaeological markers do not have a universal and objectively social meaning, we should consider regional or local frameworks in order to decodify contextual implications and practices. This is the main focus of Francesca Grassi’s paper, centred on the study of pottery production in Álava (Basque Country) during the sixth and the tenth centuries. As this was a peripheral region in relation to the early medieval Iberian states, local leadership possessed significant political and economic clout. However, the scale of the sociopolitical systems conditioned the nature of elite demand and its archaeological visibility. As F. Grassi points out, although a simple analysis of pottery production might lead us to conclude that there were no elites in some periods of early medieval Álava, other archaeological evidence indicates there were indeed different forms of social display. However, the interpretative key for this territory is the social, rather than the economic, interpretation of craft activities and everyday consumption patterns, from a substantivist perspective.46 Imported red painted pottery should be interpreted in terms of social value in relationship with the emergence of the Castilian political elites, rather than in commercial terms. In recent years, the immense development of rescue archaeology and more structured territorial projects has allowed us to discuss specialisation and social inequality while considering other archaeological markers, including, for the first time, bioarchaeological records.47 The set of papers most explicitly connecting specialist production and elite demand includes two studies devoted to estate economy in the Carolingian core area and in Portugal. The study of the Sermersheim site presented by Edith Peytremann is significant because it introduces the notion of “specific activity areas” in order to identify distinctive archaeological sites characterised by specialisation in craft and agrarian activities and the lack of domestic structures. Metallurgical activity, specialised agrarian activity, and long-term textile production have been attested at the site. The analysis of the documentary evidence has made it possible to identify the site

ed., The Oxford Handbook of Archaeological Ceramic Analysis, Oxford, 2016, pp. 115–47 (116–18). 45 N. Sykes, Beastly Questions: Animal Answers to Archaeological Issues, London, 2014, pp. 165–68. 46 J. Moreland, Concepts of the Early Medieval Economy. 47 See, among others, Loveluck, Northwest Europe; M. McKerracher, Farming Transformed in AngloSaxon England. Agriculture in the Long Eighth Century, Oxford, 2018.

eq ual an d u n eq ual s o c i e t i e s i n e arly  me d i e val e u ro pe

with an estate centre related to Ebersmunster Abbey, even though other monasteries are attested in the area as well. The archaeology of estate centres is a major topic of early medieval rural archaeology in Carolingian Europe because it has shed light upon elite agency, social domination practices and the ways in which agrarian and non-agrarian activities were conducted.48 Even though scholars are divided about the role of monastic production in early medieval economies,49 the definition, in political and economic terms, of the great number of rural sites discovered all around Europe in the last few decades through rescue archaeology is bound to transform our perception of rural societies over the next few years. Indeed, the increasing quality of the available archaeological data is making it possible to define patterns of social inequality, even in contexts characterised by homogeneous and uniform domestic structures. The systematic study of faunal and botanical remains and craft activities has provided a new avenue for the interpretation of the early medieval mountain societies of central-northern Portugal analysed by Catarina Tente. The excavation of four short-term rural sites located in the vicinity of Viseu has revealed complementary economies based on specialisation patterns. Even though there is almost no evidence of social status signatures, the nature of the attested agrarian and craft activities denotes single-purpose sites, and is interpreted as being integrated within an elite-driven system. In addition, the high degree of specialisation explains the fragility of the sites and why they were abandoned in the tenth century. However, the relationship between social inequality and economics is complex, and cannot be simply reduced to confrontation between local elites and communities. In the great narratives about the European Early Middle Ages, this confrontation has been defined within the framework of phases of growth and economic development. Such phases would have made it possible to overcome the contraction that took place after the fall of the Roman Empire, while at the same time laying the groundwork for the emergence of the lordly powers. This interpretation is challenged by Richard Hodges in his paper devoted to the analysis of the peasantry in early medieval Italy. In this brief paper, Hodges challenges the optimistic interpretations of the material culture and lifestyle of subaltern groups promoted by authors supporting the idea of the early medieval period as a peasant “golden age”. His diagnosis is radical: before the ninth century, peasant culture in Italy was miserable, following the influential ideas of scholars such as B. Ward-Perkins.50 However, the most interesting point of the paper is connected to the explanation of Carolingian economic growth, which changed the standard of living not only of the elites, but of the peasantry as

48 On this topic see R. Francovich and R. Hodges, Villa to Village. The transformation of the Roman Countryside in Italy, c. 400–1000, London, 2003; Loveluck, Northwest Europe, pp. 57–75. 49 Compare J. Henning, ‘Early Medieval towns. The development of the economy of the Frankism realm between dynamism and deceleration ad 500–1100’, in J. Henning ed., Post-Roman Towns, Trade and Settlement in Europe and Byzantium, vol. 1. The Heirs of the Roman West, Berlin-New York, pp. 3–31 (17–21) with P. Toubert, Dalla terra ai castelli. Paesaggio, agricoltura e poteri nell’Italia medievale, Turin, 1995, pp. 115–250. 50 B. Ward-Perkins, The fall of Rome and the end of civilization.

21

22

j ua n a n to n i o q u i ró s c as t i l lo

well. Using a Weberian perspective, Hodges proposes a mutualistic approach, overcoming the contraposition between elite demand and bottom-up spontaneous agency. Following Frans Theuws’s accounts, he proposes that the commoditisation attested in this period is the result of a convergence of interests and the creation of new opportunities for all social groups. This “mutualistic approach” can be found in other papers in the collection, and is proving increasing important in the study of structural changes in medieval archaeology. “Small Worlds” and Social Inequalities

Over the last few decades both history and social archaeology have been characterised by a tension between two sorts of approaches, one emphasising structures and the other individual or group agency.51 This has resulted in a difficult dialogue between “model” and empirical evidence, between generalisations and singularities. The search for regularities to which processualist archaeology aspires can at times end up conditioning how evidence is interpreted, in the same way that the generalisations drawn from the particular analyses promoted by post-processualism can in some cases grow inappropriate. In general, humanities disciplines seek to generalise from particular case studies illuminated by the available sources of evidence, and so determining methodologies and scales of analysis is a key factor to avoid these risks. One of the most interesting recent developments in early medieval social history and archaeology is the increasing attention being paid to the analysis of local societies. Little interest has generally been paid to very local scales of analysis, due to the limits of the evidence available and to an emphasis on the big questions and processes. However, some important pioneering studies have demonstrated the power of studying general processes from the local scale.52 The viability of this type of studies depends on the existence of dense textual records (which in the Early Middle Ages are almost exclusively found in monastic domains), or on long-term and high-intensity archaeological projects based on multiple deposits located in small areas, in order to analyse diversity and variability at the local level. In other words, such studies can only be carried out in specific spaces, despite the fact that the main social practices in the Early Middle Ages themselves took place at this scale. The seven papers included in this section constitute the most homogenous set of chapters in the book, despite the diversity of the evidence considered, the territories analysed and the approaches used.53 These studies share three methodological and

51 J. Bintliff, ‘Time, Structure, and Agency: The Annales, Emergent Complexity, and Archaeology’, in J. Bintliff ed., A Companion to Archaeology, Oxford, pp. 174–94. 52 W. Davies, Small Worlds, The Village Community in Early Medieval Brittany, London, 1988; C. Wickham, The mountains and the city: The Tuscan Apennines in the Early Middle Ages, Oxford; M. Innes, State and society in the Early Middle Ages, Cambridge, 2000. Eric Hobsbawm has called this scale of analysis ‘little worlds’ in E. J. Hobsbawm, ‘Peasant and Politics’, Journal of Peasant Studies 1 (1973), pp. 3–22. 53 In the Vitoria-Gasteiz conference another three papers focused on ‘small worlds’. A. Vigil-Escalera analysed forms of social exclusion and the construction of collective identities using funerary evidence (see video https://goo.gl/ZZJPYB and A. Vigil-Escalera Guirado, ‘Invisible social

eq ual an d u n eq ual s o c i e t i e s i n e arly  me d i e val e u ro pe

thematic characteristics. Firstly, unlike the “history from below” of the 60s and 70s that focused primarily on the study of subaltern groups,54 the studies on local societies included in this volume are more closely related to the Italian micro-history of the 70s and 80s.55 The local scale is considered as an observatory through which to analyse, from a bottom-up perspective, historical dynamics affecting the entire social body.56 And although the ambitions and the scale of analysis differ from chapter to chapter, the local scale is employed throughout as a strategy to approach the ever-elusive early medieval societies with a certain level of detail. Secondly, from all this it follows that another of the axes that characterise these works is the analysis of the relationships between local communities and the supra-local structures or encompassing societies in which they are inserted or participate. From this perspective, the notion of local societies is considered more like an arena of dense interaction between different agents than an object of analysis in itself.57 And although documentary filters significantly determine the ways in which peasantry, intermediate groups, and elites are conceptualised and analysed, the diversity of local realities allows us to spotlight the mosaic of socio-political situations that characterised early medieval Europe. The exercise of power in the Early Middle Ages took place at the local scale, and for this reason it is essential to penetrate through the dearth of documentation in order to decode the forms of cooperation, solidarity, competitiveness and conflict that took place at this scale. The third thematic axis that appears in all these works is the heterogeneity of local communities. Challenging traditional assumptions, it is argued that local communities in general, and peasant societies in particular, are not the static, conservative or homogeneous realities they have been portrayed as in many syntheses about this period. Despite the difficulties offered by both textual and material records when approaching the local scale, the collection of articles included in this volume illuminates forms of social mobility (which do not always end up stabilising in intergenerational terms), the existence of peasant elites, forms of intermediation with encompassing societies before the implantation of lordly domains, forms of social grouping and communitarian action, as well as the existence of true political agendas developed

54 55 56 57

inequalities in early medieval communities: the bare bones of household slavery’, in J. A. Quirós Castillo eds, Social complexity in early medieval rural communities: the north-western Iberia archaeological record, Oxford, 2016, pp. 113–23), and D. Tys studied local societies in the salt marshes of Flanders in the context of the emergence of warlords (see https://goo.gl/qYLwzy, and D. Tys, ‘Medieval moated sites in coastal Flanders: the impact of social groups on the formation of the landscape in relation to the early estates of the Count of Flanders’, in K. De Groot, D. Tys and M. Pieters eds, Exchanging Medieval Material Culture. Studies on archaeology and history presented to Frans Verhaeghe, Brussels, pp. 289–301). A third talk, by Anne Nissen Jaubert, was unfortunately not recorded. D. M. MacRaild and A. Taylor, Social Theory and Social History, New York, 2014, pp. 125–28. C. Ginzburg, ‘Microhistory: Two or Three Things that I Know about it’, Critical Inquiry 20.1 (1993), pp. 10–35; G. Levi, ‘On microhistory’, in P. Burke ed., New Perspectives on Historical Writing, Cambridge, 1991, pp. 93–113. This is, in fact, the perspective used by the previous works on ‘small worlds’ in early medieval societies mentioned in note 52. About the notion of encompassing societies see H. Mendras, Les sociétés paysannes, Paris, 1995.

23

24

j ua n a n to n i o q u i ró s c as t i l lo

in the local sphere.58 This perspective opens up the possibility of exploring topics such as the relationships between the action of the state or aristocratic power, supra-local structures and the local sphere. Another important topic is the study of the heterotemporality or multitemporality of historical processes that took place at different political scales.59 In geographical and political terms, the chapters devoted to local societies included in this volume encompass two broad areas. On the one hand, we have local Carolingian societies connoted by the predominance of public goods and strongly dominated, at least since the ninth century, by an aristocracy and monasteries. On the other hand, we have territories in north-western Iberia where the presence of such powers is less important at the local level at least until the tenth century. The work by J. P. Devroey and N. Schroeder is centred on the village of Villance, in Ardennes, one of the Carolingian core areas, and the location of an important estate of the nearby monastery of Prüm. The documentation of the monastery has enabled a critical analysis of the village’s social structure, revealing not only the complexity of the local community, but also the existence of local elites linked to the management of the monastery’s goods. Likewise, the study has exposed the difficulties of interpretation posed by the use of mansi as an instrument for the analysis of local society, and the weight of documentary filters when scanning microhistories. However, the texts make it possible to analyse economic practices based on mixed farming, integrating craft activities with infield-outfield agrarian systems and convertible husbandry practices. Consequently, oxen and lands are recognised as crucial indicators of wealth and distinction in the Villance society. The tax assets at the heart the powerful Abbey of Nonantola’s operations are the example analysed by Igor Santos. Throughout the ninth century the influence of the monastery was spreading, coming into conflict with local communities, whose social and political contours came into view precisely through these frictions. The use of judicial documentation makes it possible to observe the emergence of local elites linked to the monastery, at the same time that a dense network of patronage takes shape. This network involved agents from very heterogeneous social backgrounds, a fact exploited by Nonantola within local societies. In other words, through the analysis of individual case studies (territory of Salata, the Homines Flexicani, Lizzano and Gabba) this study illuminates the complex mechanisms through which monastic domination took root, but also the consequences that the imposition of this domain had on socio-political articulation in different local contexts. It is interesting to contrast this work with the chapter by Fabio Saggioro, also focusing on the study of local societies in northern Italy. From the archaeological 58 These topics have been analysed in other territories and from other perspectives in F. Menant and J.-P. Jessenne, Les élites rurales dans l’Europe médiévale et moderne, Actes des XXVIIe Journées Internationales d’Histoire de l’Abbaye de Flaran, Toulouse, 2007; S. Carocci a cura di, La mobilità sociale nel Medioevo, Rome, 2010; C. Loveluck, Northwest Europe in the Early Middle Ages, p. 274–301. 59 C. Witmore, ‘Vision, media, noise and the percolation of time. Symmetrical approaches to the mediation of the material world’, Journal of Material Culture, 11.3 (2006), pp. 267–92.

eq ual an d u n eq ual s o c i e t i e s i n e arly  me d i e val e u ro pe

studies carried out in the village of Nogara, the monastery of San Benedetto di Leno, the settlement of Bovolone and Canolo di Mezzo or in several sectors of the Po Valley, the chapter analyses the importance of local communities when modifying and building rural landscapes, in both economic and social terms. The concise survey presented here summarises the results of intensive archaeological projects in different sectors of the Po Valley. It is particularly suggestive in that it shows the difficulties in detecting the material signs of the local elites present in the documentation of Nonantola and other ninth-century monasteries. And although there is no territorial overlap between Igor Santos and Fabio Saggioro’s studies, it is interesting to note the diversity of information provided by the analysis of textual and material documentation at the local level.60 However, the most significant group of studies devoted to local societies is the one on north-western Iberia. Unlike the previous examples, we are faced with small worlds by and large made up of smallholders. They lack formalized political structures and are in constant negotiation with ambitious agents acting at the supra-local scale. These agents, in turn, are key in the construction of the emerging early medieval political formations.61 As noted in the preface, the long tradition of history and social archaeology in this territory has made peasant communities and local societies a subject of study since at least the 1960s.62 However, this line of enquiry has undergone an important renovation in recent years, as modelling on major issues, such as the implementation of lordships, has been replaced by research agendas attentive to the processes of social change at the local scale.63 Despite their diversity, the four papers in this block illustrate this trend very well. Two of them examine written evidence, while the other two make use of new archaeological records that have been developed in recent decades. Of all areas of northwest Iberia, León produced the most sizeable corpus of early medieval documentation, so it should come as no surprise that the two text-based papers focus on this region. In the first one, Iñaki Martín Viso presents a very articulate

60 About this contrast see C. Wickham, ‘Fonti Archeologiche e Fonti Storiche: Un dialogo complesso’, in S. Carocci ed., Storia d’Europa e del Mediterraneo. IV. Il Medioevo (secoli V–XV), Rome, 2007, pp. 15–49. 61 Noteworthy recent studies include J. Escalona, ‘In the name of a distant King: representing royal authority in the county of Castile, c. 900–1038’, Early Medieval Europe 24.1 (2016), pp. 74–102; A. Carvajal Castro, Bajo la máscara del Regnum. La monarquía Asturleonesa en León (854–1036), Madrid, 2018. 62 Although we lack up-to-date historiographic studies about local medieval communities, some general reflections can be read in C. Estepa, ‘Comunidades de aldea y formación del feudalismo. Revisión, estado de la cuestión y perspectivas’, en M. J. Hidalgo, D. Pérez and M. J. R. Gervás ed., ‘Romanización’ y ‘Reconquista’ en la Península Ibérica: Nuevas Perspectivas, Universidad de Salamanca, Salamanca, pp. 271–82: J. Escalona, ‘De señores y campesinos a poderes feudales y comunidades. Elementos para definir la articulación entre territorio y clases sociales en la Alta Edad Media castellana’, in I. Álvarez Borge ed., Comunidades locales y poderes feudales en la Edad Media, Logroño, pp. 115–57. 63 See the monograph issue published in the journal Studia Historica. Historia Medieval 31 (2013) entitled ‘Los ‘pequeños mundos’: Sociedades locales en la Alta Edad Media’.

25

26

j ua n a n to n i o q u i ró s c as t i l lo

picture of local societies in León in the tenth century. What we find are societies dominated by smallholders structured into communities which, despite having a defined identity and appearing as owners of local churches, do not necessarily limit individual agency. The peasantry is not an even block, as the existence of a high class of peasants is detected, and there are indications of other forms of social hierarchy stemming from relationships with monasteries and other external agents. Similarly, the documentation shows there were diverse processes of mobility and social promotion, from the increase of individual wealth (in terms of land ownership), to the establishment of relations with external agents, to the performance of certain mediation functions with external powers. However, within this general picture, micro-level differences are significant, so that the analysis of the examples of Marialba de la Ribera, Villa de Bera or Cañones sheds light on the diversity of active agents at the local level. The main aim of this piece is to show that the existence of deep social inequalities, which find their counterpart in the economic and political spheres, and are inherent to the very existence of smallholder communities sparingly mediated by the aristocracy or the king. Álvaro Carvajal Castro’s work is situated along similar intellectual coordinates, although he deals with the analysis of forms of collective action from the dual perspective of conflict analysis and peasant struggles that have left a documentary imprint, as well as from the viewpoint of the existence of different forms of social grouping. Based on the documentary testimonies in the León area, Carvajal questions the static, monolithic representation of the peasant community as traditionally represented by medievalists, showing how boundaries within a community could be fluid. Moreover, the communities involved in the various documented conflicts do not necessarily represent the voices of a homogeneous and cohesive body. Rather, within these communities, there were different — and at times contradictory — individual or group interests. As such, it is necessary to revise the very notion of peasant community as understood until now. And while his work mainly explores the political dimension of communities, there is no doubt that these are collectives characterized by the existence of important internal social hierarchies, and the consolidated presence of local leaders who play a leading role in the conflicts with regional monasteries. Social grouping is also the subject of analysis in the paper presented by J. A. Quirós. Village formation has been one of the main topics addressed by Medieval Archaeology in much of Western Europe in recent decades. Considered by many as a critical stage in the configuration of European landscapes and rural societies, generations of archaeologists have sought to define the moment, the processes and the reasons behind the nucleation of previously dispersed settlements into compact rural communities endowed with solid territorial identities. The new archaeological records available for the north-western area of Iberia allow us to uncover how the forms of social aggregation in the Early Middle Ages were much more fluid and complex than previously thought. Comparing the archaeological records of Madrid and Álava in terms of settlement patterns, household duration and funerary evidence, it is argued that different village formation processes took place during the Early Middle Ages in a dynamic context of creation of local identities, involving different kinds of grouping and communities. The two case studies considered show that

eq ual an d u n eq ual s o c i e t i e s i n e arly  me d i e val e u ro pe

village and farm formation processes are a meeting point, but also a source of friction, among a diversity of social subjects. In Madrid, the first documented post-Roman communities date as far back as the fifth century, and built their collective identities mainly through funerary rituals. By contrast, in Álava, village formation took much longer and followed more complex modalities. The oldest collective cemeteries were the privileged scenarios for the deployment of social display practices, but also for the construction of social memories. When the process of village formation dissociated from the first cemeteries began, starting in the seventh century, the practices of legitimation of social memory moved away from burial and into the domestic spaces. Throughout this whole process, local elites played a critical role in a context marked, once again, by the predominance of smallholders. Finally, Carlos Tejerizo’s paper discusses the notion of the “peasant mode of production”, one of Chris Wickham’s main theoretical contributions for conceptualizing preindustrial peasant societies and generating frameworks for interterritorial analysis and comparison.64 Using as a starting point a set of archaeological sites in the Duero Valley that have been investigated over the last two decades through rescue archaeological projects, it is convincingly argued that the social and economic logic of these local societies can be explained using the concept of the peasant mode of production. This identifies certain “archaeological markers” — organization of settlements, the centrality of households, architectural characteristics, and in particular storage mechanisms — which characterise these peasant societies. In some of these villages it has been possible to observe the existence of the profound social transformations that Tejerizo conceptualizes, again following Wickham’s proposal, in terms of the implantation of a feudal mode of production. To this end, he suggests the existence of two possible scenarios: the first was marked by the emergence of local elites who would end up consolidating their status, ultimately establishing class-based intergenerational forms of domain. The second path is characterised by the intervention of external powers strong enough to dominate the community. In short, this chapter shows how useful these analytical categories are in assessing early medieval archaeological records.

Some General Trends The studies collected in this volume provide different intellectual and contextual paths for the study of social asymmetries in the Early Middle Ages. Before we conclude, I would like to briefly point out what I see as the main contributions to be drawn from these works. Studying inequality in early medieval societies is no easy task. By the same token, textual records are not always explicit enough for a proper analysis of social dynamics, or numerous enough for detailed comparative study. On the other hand, although more and more early medieval archaeological records are becoming available in 64 Wickham, Framing the Early Middle Ages, pp. 535–50.

27

28

j ua n a n to n i o q u i ró s c as t i l lo

Western Europe, more than a few authors in this volume have brought to light the hostility of these records when visualizing hierarchies and their opacity on social practices. Thus, the study of this subject requires solid theoretical anchors and very refined methodologies, particularly when privileging analysis on the local scale. For decades both archaeology and history of social inequalities have aimed at defining markers or signatures of social inequality in order to detect asymmetries in normative terms. And although most of the papers in this volume still seek to explain some of the useful mechanisms for analysing social complexity in archaeological records (monumentality, defence systems, storage systems, long-term maintenance of domestic structures, imported pottery, specialised rural production, bioarchaeological evidence) or textual records (local-scale actions of the powerful, construction of local churches, forms of local-scale cooperation and collaboration, possession of resources such as oxen, lands, etc.), a holistic and critical approach to the aseptic identification of “markers” prevails. Indeed, several papers have underlined how the meaning of inequality practices and signatures is tremendously changeable and situational. Therefore, only a contextual approach can allow for the proper interpretation of the different “markers”. In addition, the papers gathered here show that the different forms of social inequality are, in the Early Middle Ages, in constant negotiation and transformation in multiscale terms, both vertically and horizontally. In local societies there is a tension between the tendencies to accentuate the forms of social differentiation, and other centripetal forces that tend, instead, to limit local distances within peasant communities. On the scale of encompassing societies, the fluid and changing nature of political structures generates opportunities for promotion and social dominance, as seen in Iberia, while constituting a destabilising factor for ambitions aspiring to intergenerational transmission. Consequently, the history and archaeology of social asymmetries cannot ultimately have as their main aim the definition of markers of inequality. Rather, they must aspire to dynamically analyse the formation, failure and contraction of these asymmetries in their contexts. In terms of theory, not all the papers in this volume make explicit reference to their theoretical underpinnings. However, perhaps the most important conclusion that the reader will come away with is the high degree of intellectual crossbreeding that currently characterises the practice of history and social archaeology. While some pieces are explicitly or implicitly aligned with historical materialism, several authors have followed theoretical approaches that replace class conflict with forms of mutualism, where the simple opposition between “lords” and “peasants” gives way to narratives that are much more nuanced and complex in social terms. Other works have resorted to the use of analytical resources such as bottlenecks or the study of contextual meanings of material culture. Despite the relative weight given to the study of local societies, no exclusive bottom-up perspectives are adopted; on the contrary, some pieces point to the need to integrate them with top-down analyses. This very emphasis on local societies is another aspect worth underlining. While Mesoamerican Social Anthropology has underlined the need to study social inequalities through macro-level approaches based on cross-cultural comparison, this volume shows the growing importance of micro-level studies that privilege the contextual analysis of material culture, and processes documented through material sources. This

eq ual an d u n eq ual s o c i e t i e s i n e arly  me d i e val e u ro pe

tendency towards the most localised analyses should be read not only in terms of the quality of the records and the detail of the analysed dynamics, but also from a purely theoretical perspective. And while a priori this particularised approach can hinder comparative analysis, Chris Wickham has shown that it is not an impossible task.65

Final Remarks In short, this volume aims to make an important contribution both to the study of the complexity of early medieval societies, through the analysis of issues of social inequality, and to the analysis of the forms of inequality in the past from the perspective of material and narrative sources in different areas of Western Europe. As opposed to teleological approaches that have considered the Early Middle Ages in terms of the simplification of ancient societies following the collapse of the Roman Empire, or homogeneity against the subsequent development of lordships and feudal monarchies, these papers bring to the fore the specificities of this historical period. If there is one common thread that runs through nearly all of the texts in this volume, it is the express will to analyse, in explanatory terms, the emergence and functioning of social inequalities, questioning the “natural” or given character of the asymmetries that characterise complex societies on a wide variety of interconnected scales. And despite its apparent banality, this is perhaps one of the main contributions of this book. A volume of this nature always leaves in its wake numerous open topics, uncovered territories and unfinished tasks for the future. However, only one of these merits mentioning in these concluding remarks. So far experts have focused their attention on the analysis of inequalities and their forms of long-term institutionalisation in the Early Middle Ages in economic and social terms. And although, as we have seen, it is not easy to identify and conceptualise inequality in the period under analysis, there is no doubt that the main challenge is to address equality and the mechanisms of compensation for social imbalances used in this period. While social differentiation precedes the institutionalisation of inequality, hierarchies existed and played a key role even in egalitarian contexts.66 Based on the idea that there are no egalitarian societies in the Middle Ages, but egalitarian scenes or situations,67 the analysis of moral economy practices, the forms of acquiring prestige without accumulating power, the practices of concealment of social differences in material culture, or the use of other mechanisms of social stress compensation,68 constitute some of the thematic axes that can guide an archaeology of equality in unequal societies such as those of the Middle Ages. But this is a subject for another volume altogether. 65 C. Wickham, Framing the Early Middle Ages. 66 R. Paynter, ‘The Archaeology of Equality and Inequality’, Annual Review of Anthropology 18 (1989), pp. 369–99. 67 J. G: Flanagan, ‘Hierarchy in simple “egalitarian” societies’, Annual Review of Anthropology 18 (1989), pp. 245–66. 68 Including social conflicts, J. Müller, ‘Rebellion and inequality: Hierarchy and Balance’, in S. Hansen and J. Müller eds, Rebellion and inequality, pp. 23–32.

29

1

State Formation and Socio-Political Complexity

Julio Escalona*

Towards an Archaeology of State Formation in North-Western Iberia

To discuss state formation in the context of a more general reflection on social inequality inevitably feels a bit like ‘building the house from the roof down’, as we say in Spain. In many ways the state can be seen as a point of arrival in social stratification; a point at which social inequality becomes embedded in a formal organisation that outlives individuals and facilitates its transgenerational reproduction. How such systems of ‘durable inequality’ — to borrow Charles Tilly’s term — came to exist in early medieval Europe has long been recognised as a crucial topic for both historians and archaeologists.1 There are nevertheless important divergences in the approaches to this subject in both camps, to the point of complicating any effort to combine the two of them into a single theorisation. Archaeological approaches mainly derive from the social anthropological theory of the mid-twentieth century on, which was dominated by linear evolutionary models that seek to define a sequence of stages of social structuring eventually leading to statehood.2 Linear evolutionary models have been, and in many respects still are,



* This paper was produced with support from the project HAR2013–47889-C3–2-P, funded by the Spanish Programa Estatal de Investigación Científica y Técnica de Excelencia and under the framework of the ‘Grupo de Estudios del Mundo Rural Medieval’ (Unidad Asociada Universidad del País Vasco — CSIC). The theoretical strand it follows is heavily dependent on the author’s collaborative work over the past several years with Orri Vésteinsson and Stuart Brookes. A preliminary version was presented in 2016 at the Vitoria-Gasteiz conference, and the author is most grateful for the feedback received, especially from Juanjo Larrea and Robin Beck. Huge thanks are also due to the editor, Juan Antonio Quirós, for his patience and encouragement. While all of them contributed a great deal to improving the final text, none of its possible shortcomings should be attributed to anyone but the author. 1 C. Tilly, Durable Inequality, Berkeley, 1998; see also C. Tilly, Identities, boundaries and social ties, Boulder, USA, 2005, pp. 71–90. On the development and impact of Tilly’s work see K. Voss, ‘Enduring Legacy? Charles Tilly and Durable Inequality’, The American Sociologist, 41–4, 11/06 2010, pp. 368–74. See also Robin Beck’s contribution to this volume. 2 See especially M. H. Fried, The Evolution of Political Society: An Essay in Political Anthropology, New York, 1967; E. Service, Primitive Social Organization: An Evolutionary Perspective, New York, 1971. Julio Escalona  •  Instituto de Historia — CSIC Social Inequality in Early Medieval Europe: Local Societies and Beyond, ed. by Juan Antonio Quirós Castillo, Turnhout : Brepols, 2020 (HAMA 39), pp. 33–53 © FHG 10.1484/M.HAMA-EB.5.118444

34

J u l i o E s c a lo n a

hugely influential; they have a lot to say, and they are important to take into account.3 However, applying evolutionary theory to early medieval state formation often feels like running in circles, even though elements of evolutionary thinking come up time and again in early medieval studies. Anthropological and archaeological evolutionary models were originally developed with a focus on the early states of the Near East and Mesoamerica, which emerged under environmental and economic conditions that were very different from those of early medieval Europe, as was their long-term evolution into large-scale units that from a medieval perspective would be called empires.4 Furthermore, and perhaps more importantly, linear evolutionism naturally privileged the study of ‘primary states’, that is, stateless societies that turned into states through different mechanisms and due to different causes, but always representing the first cases of statehood in areas where states did not previously exist. This is, of course, at odds with the fact that the predecessor of many early medieval polities was the Roman state, meaning they do not qualify as ‘primary states’. Indeed, primary states represent an absolute minority among the processes of state formation throughout history, while the majority of cases are difficult to assign to a single evolutionary track from pre-statehood to statehood.5 Accordingly, scholars gradually shifted the focus towards the differential developments that, in areas like the Near East, led to the creation of a world-system comprising a ‘spatial hierarchy of dependent social formations’ organised in centres and peripheries that ‘advanced the frontier of civilisation, interrupted by cycles of collapse and resistance’.6 In the late twentieth century, world systems theory and centre-periphery approaches highlighted the importance of short- and long-distance interactions between non-state and state societies in the study of social and political change.7 As the focus shifted from linear to non-linear approaches to state formation, a richer anthropological and archaeological agenda emerged, which accommodated for stagnation, disruptions, interruptions, forks, and regressions on the road to statehood and beyond. Among them, the issue of ‘collapse’ became a favourite theoretical motif in two steps. In the context of the end of the Cold War and the fall of the Soviet Union, the late twentieth-century literature was most interested in collapse derived from





3 K. V. Flannery and J. Marcus, The Creation of Inequality: How Our Prehistoric Ancestors Set the Stage for Monarchy, Slavery, and Empire, Cambridge, MS, 2012. 4 Hence maybe the marginality of the Middle Ages in global approaches to empires; see S. E. Alcock, T. N. D. D’Altroy, K. D. Morrison and C. M. Sinopoli (eds), Empires. Perspectives from archaeology and history, Cambridge, 2001. 5 For an early attempt towards theorisation, see B. Price, ‘Secondary State Formation: An Exploratory Model’, R. Cohen and E. R. Service (eds), Origins of the State: The Anthropology of Political Evolution, Institute for the Study of Human Issues, Philadelphia, 1978, pp. 161–224. 6 The quotations are from K. Kristiansen, ‘Chiefdoms, states and systems of social evolution’, T. Earle (ed.), Chiefdoms: Power, Economy and Ideology, 2, Cambridge, 1991, pp. 16–43, at p. 41. See also the critical remarks in S. J. Shennan, ‘The history of social hierarchies’, European Journal of Archaeology, 6, 2003, pp. 91–97. 7 For a paramount example, P. S. Wells, Farms, villages and cities. Commerce and urban origins in late prehistoric Europe, Cornell, 1984.

towards an archaeology of state formation in north-western iberia

structural economic, military and organisational mismatches.8 Then, around the turn of the millennium, rising concerns about climate change and global sustainability triggered an eco-determinist view with apocalyptic overtones that gained remarkable popularity, despite the theoretical and empirical weakness of its application to the demise of past polities.9 A natural development of scholarly reflections on collapse was the theorisation of the aftermath of collapse, through concepts such as ‘resilience’ and ‘regeneration’, as the realisation set in that collapse rarely leads to annihilation, but rather to different versions of recovery, and eventually to the reconstruction of complexity under alternative terms and conditions.10 These recent developments are better suited to the early Middle Ages. Of particular importance is the progressive replacement of unilinear evolutionism in favour of more complex views of interconnected socio-political change — including collapse and regeneration — within networks of polities having different degrees of complexity. The new understanding that most cases of state formation are in fact instances of ‘secondary state formation’ is also of obvious interest to medievalists. However, the impact of modern anthropological and archaeological theories on the study of early medieval state formation is limited and uneven. They are most often applied to European regions without a Roman past, such as Central Europe, Ireland, and especially Scandinavia, where archaeology takes the lead and the direct transition from Late Prehistory into the Middle Ages welcomes evolutionary and post-evolutionary views. However, such theories are similarly detectable in British archaeology as well. In trying to make sense of how the northern stateless societies made their way to statehood, notions of bottom-up agency play a major role, especially through the operation of local or micro-regional assemblies that are seen to constitute the first steps in the construction of a political body.11 The existence of small-scale political

8 J. A. Tainter, The collapse of complex societies, Cambridge, 1988 and N. Yoffee and G. L. Cowgill (eds), The collapse of ancient states and civilizations, Tucson - London, 1995 were probably the most influential among archaeologists, while P. Kennedy, The Rise and Fall of the Great Powers: Economic Change and Military Conflict From 1500 to 2000, New York, 1987 had considerable impact on the general public. See also the Special Issue on ‘The Rise and Fall of States and Empires’ in Journal of World-Systems Research, 10–13, 2004. 9 The flagship study was J. Diamond, Collapse: How societies choose to fail or succeed, New York, 2005. See the criticism in P. A. McAnany and N. Yoffee (eds), Questioning Collapse. Human Resilience, Ecological Vulnerability, and the Aftermath of Empire, Cambridge, 2009, esp. pp. 1–17. 10 K. D. Morrison, ‘All is not lost when the center does not hold’, Science and Archaeology, 315, 2007, pp. 42–43; G. M. Schwartz and J. J. Nichols (eds), After collapse. The regeneration of complex societies, Tucson, 2006. 11 See A. Pantos and S. Semple (eds), Assembly places and practices in Medieval Europe, Dublin, 2004; F. Iversen, ‘Concilium and Pagus: Revisiting the Early Germanic Thing System of Northern Europe’, Journal of the North Atlantic (Special Volume), 5, 2013, pp. 5–17; F. Iversen, ‘Community and Society. The Thing at the Edge of Europe’, Journal of the North Atlantic, 8, 2015, pp. 5–21. A wider European discussion is to be found in C. Wickham, ‘Consensus and Assemblies in the Romano-Germanic Kingdoms: A Comparative Approach’, V. Epp and C. H. F. Meyer (eds), Recht und Konsens im Frühen Mittelalter, Ostfildern, 2017, pp. 389–426. On the role of political assemblies in early medieval polities see T. Reuter, ‘Assembly politics in Western Europe from the Eighth Century to the Twelfth’,

35

36

J u l i o E s c a lo n a

communities with a notion of representative rather than coercive rulership is seen as a crucial factor for the transition toward increasingly complex, larger-scale political systems.12 Scale growth seems to have played a key part in the evolution from representational politics into statehood. Within larger territorial aggregates, it is possible for central powers to break free from too much dependence on local communities and override local constraints, gradually limiting the role of the lesser assemblies — and communities — to strictly local affairs. The best illustration of this dialogue between small basic political units and larger-scale constructs may be the ‘classical’ three-tiered view of the early Irish polities.13 The fundamental unit, the túath, was a small community based on an ‘organic’ relationship of proximity between the society and a petty king who in anthropological terms was essentially a ‘chieftain’.14 Above the túath level, the political articulation consisted of overkings — ideally in two levels — who established their power over the petty túath kings through links of kinship and clientele. In this scheme the túath was both the most stable level, and also the one in which the king’s domination was least coercive, as it depended closely upon the community he embodied. By contrast, at the higher levels, power relations depended upon the pyramidal networks that higher kings were able to develop and maintain with their sub-kings and their retinues of military aristocrats. Such networks were larger and more powerful, in military terms, than any of the small tuatha, but also much more volatile, because — in the absence of institutionalisation at the top level — they depended on the contingencies of warfare, kinship, marriage and political developments of all sorts, and this kind of relationships can easily fall apart within a person’s lifetime.15 In a model like this, in which complexity grows from the bottom up through the aggregation of smaller blocks, the institutionalisation of the upper levels entails a separation between the higher political spheres and the local communities, as in the case of Norway, where the local things tended to become gradually disconnected from

12 13

14 15

P. Linehan and J. L. Nelson (eds), The Medieval World, London, 2001, pp. 432–50 and P. S. Barnwell and M. Mostert (eds), Political assemblies in the earlier Middle Ages, Studies in the early Middle Ages, 7, Turnhout, 2003. The similarities with the anthropological concepts of ‘chieftain’ or ‘big man’ are obvious. See D. Quigley, ‘Introduction: The Character of Kingship’, D. Quigley (ed.), The Character of Kingship, Oxford - New York, 2005, pp. 1–23. The highly idealistic traditional model of early Irish kingship is largely derived from the discourse of legal texts. The key references are D. Binchy, ‘Celtic and Anglo-Saxon Kingship’, Oxford, 1970 and F. J. Byrne, Irish Kings and High Kings, London, 1973. Subsequent research has highlighted the differences between that notional model and the actual workings of power relationships. See, among others, D. Ó Corráin, ‘Kingship and the Royal Dynasties (Prehistoric and Early Christian Ireland)’, R. F. Foster (ed.), The Oxford Illustrated History of Ireland, Oxford 1989, pp. 25–31; D. B. Gibson, From chiefdom to state in early Ireland, Cambridge, 2012 and C. Doherty, ‘The evolution of kingship and government in early medieval Ireland’, H. Oudart, J.-M. Picard and J. Quaghebeur (eds), Le prince, son peuple et le bien commun: De l’antiquite tardive a la fin du Moyen Âge, Rennes, 2013, pp. 397–410. For more details, see T. M. Charles-Edwards, ‘A contract between king and people in early medieval Ireland: Críth Gablach on Kingship’, Peritia, 8, 1994, pp. 107–19. See D. Ó Corráin, ‘Prehistoric’; N. B. Aitchison, ‘Kingship, Society, and Sacrality: Rank, Power, and Ideology in Early Medieval Ireland’, Traditio, 49, 2016, pp. 45–75.

towards an archaeology of state formation in north-western iberia

upper-tier politics and limited to local affairs.16 In Anglo-Saxon England, the growth of the great kingships of the eighth and ninth centuries, like Mercia or Wessex, was brought about by erasing many smaller pre-existing polities and replacing them with a different, top-down driven organisation: the hundredal network.17 By contrast to Northern Europe, in the areas that were once part of the Roman Empire, archaeological theory has made much less of an impact. This is not just because of the different historical trajectories of the two areas. Crucially, by the time archaeology entered the scene of medieval studies, the debate on post-Roman state formation had largely run its course among historians. Their views were therefore mostly based on written sources and determined by legal conceptions about social structuring and the state that were established independently from any archaeological evidence. The twentieth-century interwar period may perhaps be taken as the crux of the debate for most of the following decades. During those years, and largely as a reaction against the highly idealistic conceptions favoured by the German ‘Verfassungsgeschichte’, early and high medieval political structures were reformulated by emphasising the role of ties of personal dependence: royal power relied upon the king’s personal estates and the networks of fidelity he could control.18 Personal ties between kings and aristocrats became the main factor to shape the stateless polities that succeeded the Roman Empire — the so-called ‘Germanic’ or ‘Barbarian’ kingdoms. This became a widespread paradigm, although often hybridised with many elements of old constitutional history, which was never completely abandoned.19 Seen from an archaeological/anthropological perspective, a notion of political power based upon person-to-person patronage links would seem easy to reconcile with the chiefdom phase of evolutionary anthropology mentioned above, and could be handily applied, for example, to the upper layers of Irish political construction. However, the primary target of such theories was instead the large-scale, far more complex Merovingian and Carolingian kingdoms, and especially the latter. To what extent Carolingian Francia was a successor to Rome, what its real character was, or how it ended and what kind of society replaced it, became crucial questions that could easily be exported to other medieval contexts — most notably Anglo-Saxon England — but were never given definitive answers. As Susan Reynolds put it, ‘Historians of medieval Europe may in

16 F. Iversen, ‘Community and Society’; F. Iversen, ‘Concilium and Pagus’. 17 The literature is vast. See for reference S. R. Bassett (ed.), The Origins of Anglo-Saxon Kingdoms, Leicester, 1989; B. Yorke, Kings and Kingdoms of Early Anglo-Saxon England, London, 1990; M. P. Brown and C. A. Farr (eds), Mercia: An Anglo-Saxon Kingdom in Europe, Continuum, London, 2001. On the development of the hundredal system, compare the classical works by H. R. Loyn, ‘The hundred in England in the Tenth and Eleventh Centuries’, and H. Hearder and H. R. Loyn (eds), British Government and Administration Studies presented to S. B. Chrimes, Cardiff 1974, pp. 1–15, with recent developments in J. Baker, S. Brookes and A. Reynolds (eds), Landscapes of Governance: legal geographies and political order in Anglo-Saxon England, ad 400–1066, forthcoming. 18 S. Reynolds, ‘The historiography of the medieval state’, M. Bentley (ed.), Companion to Historiography, London, 1997, pp. 109–38. 19 See the sharp remarks in M. Innes, State and Society in the Early Middle Ages. The Middle Rhine Valley, 400–1000, Cambridge, 2000, pp. 4–12.

37

38

J u l i o E s c a lo n a

some ways have found their view of the state impeded by all the work that has been done’, and the race seems to go on and on.20 Clearly, a model focused on networks of personal dependence cannot, on its own, account for the political articulation of the Frankish kingdoms. They simply contained too many elements of statehood, even though personal links did play a significant role (as was also the case in Rome). This may explain how in the mid-twentieth century, during the heyday of French influence over medieval studies and the pushback against the ‘Germanic kingships’ models, an alternative narrative was developed in Francophone scholarship arguing that Late Antique statehood never fully faded out in the Frankish realm. The fundamental components of state governance survived — reinterpreted within a Christian framework — until the year 1000, at which point the so-called ‘feudal revolution’ or ‘feudal mutation’ swiftly led into the feudal dismantling of state structures. This model flourished and became dominant in the post-World War II period thanks to historians such as Duby, Bonnassie, Poly or Bournazel. One particular strand — the so-called ‘fiscalist’ school — took on aspects of Ganshof ’s neo-constitutionalism to produce a vision of the post-Roman Frankish state as essentially identical to Rome.21 The notion of a continuation of ‘public power’ after Rome and until the crisis of the Carolingian state, followed by a ‘feudal revolution’, was a hugely influential model that still colours the way we think about most regions inside or on the periphery of the Carolingian core, and tended to set the standards by which all other neighbouring statehoods were gauged. Alternative visions emerged as the end of the century approached, giving way to a well-known international debate on the topic.22 Scholars are now, in general, more open to recognising early medieval forms of statehood in Carolingia and elsewhere, although the focus seems to be firmly set upon the higher political elites and the legal and religious foundations of power and authority. Bottom-up approaches are preferentially explored in the turn-of-century historiography through ideological notions such as ‘ethnicity’ and ‘identity’, to the detriment of more structural and

20 S. Reynolds, ‘The historiography of the medieval state’, p. 109. Compare R. Davies, ‘The medieval state: the tyranny of a concept’, Journal of Historical Sociology, 16, 2003, pp. 280–300 and S. Reynolds, ‘There were states in medieval Europe: A response to Rees Davies’, ibid., pp. 550–58. 21 F.-L. Ganshof, Qu’est-ce que la féodalité, Office de publicité, Bruxelles, 1944. The model is pervasive, even among authors with very different historiographical orientations. Compare G. Duby, Guerriers et paysans, vii-xiie siècle; premier essor de l’économie européenne, Paris, 1973; P. Bonnassie, La Catalogne du milieu du xe à la fin du xie sècle: croissance et mutations d’une société, Toulouse, 1976 or G. Bois, La mutation de l’an mil. Lournand, village mâconnais de l’Antiquité au féodalisme, Paris, 1989. From that trunk further debates branched out, such as the one on the continuation of Roman taxation advocated by J. Durliat, Les finances publiques. De Diocletien aux Carolingiens (284–888), Thorbecke, 1990 (on which see C. Wickham, ‘La chute de Rome n’aura pas lieu’, Le Moyen Age, 99, 1, 1993, pp. 107–26), the debate on ‘incastellamento’ initiated by P. Toubert, Les structures du Latium médiéval. Le Latium méridional et la Sabine du ixe siècle à la fin du xiie siècle, Rome, 1973, or the ‘feudal revolution’ or ‘feudal mutation’, a favourite of the turn of the twenty-first century that prompted much literature from opposing historiographical fields (see C. West, Reframing the Feudal Revolution: Political and Social Transformation Between Marne and Moselle, c. 800-c. 1100, Cambridge, 2013). 22 See especially S. Reynolds, Kingdoms and Communities in Western Europe, 900–1300, Oxford, 1984.

towards an archaeology of state formation in north-western iberia

social analyses.23 In 2005 Chris Wickham called on scholars to break free from the continuity/discontinuity straightjacket by focusing upon the enormous regional differences within the Roman Empire and the many possible ways of ‘ceasing to be Roman’.24 Regarding states, he distinguished three main variants in the fragmentation of the Roman ‘strong state’: a) the still ‘strong’ Byzantine state, where taxation fuelled complex bureaucratic and military machineries; b) the ‘weak’ post-Roman Barbarian kingdoms, which preserved formal governance and some bureaucracy, but where taxation gradually faded out as the funding for the army shifted from state payments to land assignments; and c) the non-state polities that emerged especially in the non-Roman parts of Europe, or wherever Roman structures were fundamentally destroyed, as in Britain25. Wickham’s proposal focuses upon the resources that states were able to control and the capabilities they developed with these resources. In doing so, he departs significantly from mainstream historical arguments on medieval state formation, which tend to revolve around the emergence of political identities26 and the institutionalisation of governance, instead broadening his scope to include crucial functions such as warfare, justice and lawmaking, or coinage and taxation, or even Weber-inspired sociological approaches based on state control of violence/force.27 23 See J. L. Nelson, ‘Kingship and Royal Government’, R. McKitterick (ed.), The New Cambridge Medieval History, 2 (c. 700-c. 900), Cambridge, 1994, pp. 343–430; J. Nelson, ‘Rulers and Government’, T. Reuter (ed.), The New Cambridge Medieval History III c. 900–c. 1024, Cambridge, 2000, pp. 95–129; P. Fouracre, ‘Conflict, Power and Legitimation in Francia in the Late Seventh and Eighth Centuries’, I. Alfonso Anton, H. Kennedy and J. Escalona (eds), Building Legitimacy. Political Discourses and Forms of Legitimation in Medieval Societies, Leiden/Boston, 2004, pp. 3–26. On ecclesiastical involvement in the workings of the state, see especially M. De Jong, ‘The state of the church: ecclesia and early medieval state formation’, W. Pohl and V. Wieser (eds), Der Frühmittelalterliche Staat: Europäische Perspektiven, Wien, Institut für Mittelalterforschung der Österreichischen Akademie der Wissenschaften, Wien, 2009, pp. 241–54. See more generally the contributions to S. Airlie, W. Pohl and H. Reimitz (eds), Staat im Frühen Mittelalter, Forschungen Zur Geschichte Des Mittelalters, 11, Wien, 2006; W. Pohl and V. Wieser (eds), Der frühmittelalterliche Staat - Europäische Perspektiven, Forschungen zur Geschichte des Mittelalters, 17, Wien 2009. 24 C. Wickham, Framing the early middle ages. Europe and the Mediterranean 400–800, Oxford, 2005, pp. 3–14. 25 Ibid., p. 56. 26 The study of collective identities linked to and legitimating the emergent early medieval polities (whether state or non-state) grew immensely in the late twentieth century, especially thanks to Walter Pohl’s notion of ‘ethnogenesis’ and Patrick Geary’s work on the roots of European nations, with a host of further studies. See, among many other titles, W. Pohl (ed.), Kingdoms of the Empire: The Integration of Barbarians in Late Antiquity, The Transformation of the Roman World 1, Leiden-New York-Köln, 1997; W. Pohl and H. Reimitz (eds), Strategies of Distinction. The Construction of Ethnic Communities, 300–800, The Transformation of the Roman World, 2, Leiden-New York-Köln, 1998; S. Forde, L. Johnson and A. V. Murray (eds), Concepts of national identity in the Middle Ages, Leeds, 1995; P. Geary, The Myth of Nations: The Medieval Origins of Europe, Princeton, NJ, 2003; I. H. Garipzanov, P. Geary and P. Urbanczyk (eds), Franks, Northmen and Slavs. Identities and state formation in early medieval Europe, Cursor Mundi, 5, Turnhout, 2008; W. Pohl, C. Gantner and R. Payne (eds), Visions of Community in the Post-Roman World: The West, Byzantium and the Islamic World, 300–1100, Farnham, 2012. 27 S. Airlie, W. Pohl and H. Reimitz (eds), Staat im Frühen Mittelalter; W. Pohl and V. Wieser (eds), Der frühmittelalterliche Staat. On state control of force see quintessentially S. Reynolds, ‘The historiography of the medieval state’.

39

40

J u l i o E s c a lo n a

Wickham’s rejection of the continuity/discontinuity dilemma as a false problem could be easily extended to the theoretical approaches to state formation that I have briefly discussed above. In a wider context of secondary state formation and close interconnections between areas, none of these models can fully account for the political developments in any region of early medieval Europe. Rather, certain elements of each of them can be detected in all situations, whether due to internal factors or external influence. Interestingly, the role of inequality and domination in state building tends to be neglected from most points of view.28

State Formation in North-West Iberia The second part of this paper is an exploratory approach to early medieval state formation in the north-western quadrant of the Iberian Peninsula (hereinafter NW Iberia, for the sake of brevity). In NW Iberia there was no primary state formation. All the polities involved represent secondary developments through one or more of four basic vectors:29 a. ‘Direct inheritance’, that is, states that developed from an earlier form of state, as is the case of the post-Roman Barbarian kingdoms. b. ‘Reintroduction’ of elements of statehood from outside into areas where state structures had collapsed. This is an important factor in the Christianisation of the north, where the spread of the Church brought with it crucial notions of statehood.30 c. ‘Satellitisation’ of territories on the periphery of states that do not operate directly upon them, but influence their internal political dynamics. d. ‘Incorporation’ of non-state societies into states. a. Direct Inheritance

Early medieval state formation in NW Iberia must be considered within the framework of the dismantling first of the Roman state in the fifth century, and then of its Visigothic successor in the eighth. The region had always been peripheral and subjected to ‘colonial’ dynamics in the Roman period. Its elites never rose to the 28 Though hardly the only one, R. Faith is a notable exception: R. Faith, The English Peasantry and the Growth of Lordship, Leicester, 1997; and recently R. Faith, ‘Forms of Dominance and the Early Medieval Landscape’, Medieval Settlement Research, 23, 2008, pp. 9–13. See also M. Innes, ‘Property, politics and the problem of the Carolingian state’, W. Pohl and V. Wieser (eds), Der Frühmittelalterliche Staat: Europäische Perspektiven, Wien, 2009, pp. 299–313. 29 These categories are presented and thoroughly discussed in a collaborative work with Orri Vésteinsson and Stuart Brookes for a forthcoming book: J. Escalona, O. Vésteinsson and S. Brookes, ‘Polities, Neighbourhoods and things in between’, J. Escalona, O. Vésteinsson and S. Brookes (eds), Polity and Neighbourhood in Early Medieval Europe, The Medieval Countryside, 21, Turnhout, forthcoming. They are as much my contribution as that of my two co-editors. 30 S. Bagge, ‘Christianization and State Formation in Early Medieval Norway’, Scandinavian Journal of History, 30–32, 2005, pp. 107–34.

towards an archaeology of state formation in north-western iberia

highest senatorial echelons of imperial society, as those from Baetica sometimes did. The richest layer of Late Roman villae in the Duero Plateau points instead to external magnates who owned luxury estates there. The urban network was highly dependent upon state agency and the town-countryside relationship was vulnerable.31 When in the early fifth century the westernmost part of the Iberian north fell into the hands of the Sueves (fig. 1) the region’s large-scale economic networks and territorial control were severely disrupted.32 In the fifth and sixth centuries a long-distance trade route survived, still linking the Mediterranean with high-status sites in south-west Britain and Ireland, arguably with Bordeaux as its main middle point. Recent archaeological research has unveiled traces of this maritime circuit at certain sites in Galicia like Vigo,33 and similar cases could be made for Portugal and the Cantabrian coast as well, although the evidence for these latter areas is so far less compelling.34 Trade along this route was active throughout the Suevic period. Nevertheless, just how far inland its effects permeated, or whether it was a strictly coastal phenomenon, are questions that remain unanswered. Meanwhile, in the Duero Plateau the high-status Roman villae vanished at a quick pace35 as the entire settlement pattern underwent crucial changes, with hilltop sites emerging on the peripheries of waning over-sized municipal territories, a process that seems to mark the rise of a different kind of regional rural elites, as suggested by Iñaki Martín Viso.36 Other medium- and low-status rural settlements suffered a similar

31 See a more developed argument in J. Escalona, ‘Patrones de fragmentación territorial: el fin del mundo romano en la Meseta del Duero’, U. Espinosa and S. Castellanos García (eds), Comunidades locales y dinámicas de poder en el norte de la Península Ibérica durante la Antigüedad Tardía, Logroño, 2006, pp. 165–200. 32 See P. C. Díaz Martínez, El reino suevo (411–585), Madrid, 2011, esp. pp. 55–68. 33 A. Fernández, El comercio tardoantiguo (ss. IV–VII) en el Noroeste peninsular a través del registro cerámico de la ría de Vigo, Oxford, 2014, (Roman and Late Antique Mediterranean Pottery, 5). 34 J. M. Iglesias, ‘Establecimientos portuarios en el sector central del mar Cantábrico’, M. Urteaga and M. J. Noain (eds), Mar exterior: el Occidente atlántico en época romana, Roma - Irún - San Sebastián, 2006, pp. 109–22; J. A. Quirós Castillo, A. Azkarate Garai-Olaun, R. Bohigas Roldán, I. García Camino, Á. L. Palomino Lázaro and J. M. Tejado Sebastián, ‘Arqueología de la Alta Edad Media en el Cantábrico Oriental’, A. Llanos (ed.), Medio siglo de arqueología en el Cantábrico Oriental y su entorno, Vitoria-Gasteiz, 2009, pp. 449–500. 35 A. Chavarría Arnau, ‘Interpreting the Transformation of Late Roman Villas: The case of Hispania’, N. Christie (ed.), Landscapes of Change. Rural evolutions in Late Antiquity and the Early Middle Ages, Aldershot, 2004, pp. 67–102. See also T. Lewit, ‘‘Vanishing villas’: what happened to élite rural habitation in the West in the 5th-6th c?’, Journal of Roman Archaeology, 16, 2003, pp. 260–74. 36 I. Martín Viso, ‘Central places and the territorial organization of communities: the occupation of hilltop sites in early medieval northern Castile’, W. Davies, G. Halsall and A. Reynolds (eds), People and Space in the Middle Ages (300–1300), Turnhout, 2006, pp. 187–208; I. Martín Viso, ‘La sociedad rural en el suroeste de la meseta del Duero (siglos VI–VII)’, G. Del Ser Quijano (ed.), Espacios de poder y formas sociales en la edad Media: estudios dedicados a Ángel Barrios, Salamanca, 2007, pp. 171–88; I. Martín Viso, ‘Tremisses y potentes en el nordeste de Lusitania (siglos VI–VII)’, Mélanges de la Casa de Velázquez. Nouvelle série, 38–31, 2008, pp. 175–200; A. Vigil-Escalera Guirado and C. Tejerizo García, ‘Asentamientos fortificados altomedievales en la meseta. Algunas distorsiones historiográficas’, R. Catalán Ramos, P. Fuentes and J. C. Sastre (eds), Las fortificaciones en la tardoantigüedad: Élites y articulación del territorio (siglos V–VIII d. C.), Madrid, 2014, pp. 229–46;

41

42

J u l i o E s c a lo n a

contraction that ultimately led to the first phase in the creation of the village-type peasant settlements that by 900AD were dominant in the regional landscape. The recognition of these rural sites represents perhaps the most important advance in Iberian medieval archaeology over the last two decades.37 Between the fifth-century Roman collapse and the reconstruction of an Iberia-wide state by the Visigothic monarchy in the early sixth century, the peninsula underwent decades of instability. Some regions, like Baetica, had well-developed cities and a strong town-countryside articulation, which allowed them to remain firmly in the hands of the regional aristocracies and navigate the period so as to maintain the basic elements of statehood.38 This was not the case of NW Iberia, where the small Suevic kingdom stayed independent in the far west until the 580s, while the lands east of it became a sort of marginal cul-de-sac which, until well into the sixth century, knew no effective state agency other than sporadic military campaigns launched from the Roman-controlled Tarraconensis or, later, from the Visigothic kingdom of Toulouse. This can only have accentuated the decline of the towns, bringing about the severe disruption of the territorial patterns in the region. The radical changes that show up in the episcopal network and the ecclesiastical divisions until the mid-sixth century are an additional expression of the region’s overall disarticulation.39 This process of fragmentation placed serious limitations on the reintroduction of state power in the sixth century, including the annexation of the Suevic territory in the

C. Tejerizo García and J. Canosa-Betés, ‘Power, control and social agency in post-Roman northern Iberia: An archaeological analysis of hillfort occupations’, Journal of Medieval Iberian Studies, 10–13, 2018, pp. 295–323. 37 J. A. Quirós Castillo and A. Vigil-Escalera Guirado, ‘Networks of peasant villages between Toledo and Velegia Alabense, northwestern Spain (V-Xth centuries)’, Archeologia Medievale, 33, 2006, pp. 79–128; A. Vigil-Escalera Guirado and J. A. Quirós Castillo, ‘Early medieval rural societies in North-Western Spain: Archaeological reflections of fragmentation and convergence’, J. Escalona and A. Reynolds (eds), Scale and Scale Change in the Early Middle Ages: exploring landscape, local society and the world beyond, Turnhout, 2011, pp. 33–60; A. Vigil-Escalera Guirado, Los primeros paisajes altomedievales en el interior de Hispania. Registros campesinos del siglo quinto d.C., Bilbao, 2015; C. Tejerizo García, Arqueología de las sociedades campesinas en la cuenca del Duero durante la Primera Alta Edad Media, Bilbao, 2017. 38 Recent archaeological research in Córdoba has revealed the creation of a dense, dynamic network of high-status villas in the Emirate and Caliphate periods that was strongly dependent upon an existing network of peri-urban Roman villae and their associated irrigation infrastructures. More than straightforward continuity in settlement patterns, this seems to indicate a relatively healthy town-to-countryside relationship in the post-Roman period, which may be a sounder mark of operating regional elites and governance. See J. F. Murillo, A. León Muñoz, E. Castro, M. T. Casal, R. Ortiz and J. González, ‘La transición de la civitas clásica cristianizada a la madina islámica a través de las transformaciones operadas en las áreas suburbiales’, D. Vaquerizo Gil and J. F. Murillo (eds), El Anfitetatro Romano de Córdoba y su Entorno Urbano. Análisis Arqueológico (ss. I–XIII d. C.), Córdoba, 2010, pp. 530–47; J. F. Murillo Redondo, ‘Grandes Residencias Suburbanas en la Córdoba Omeya. Estado de la Cuestión’, Al-Mulk, 12, 2014, pp. 85–108. 39 A. Barbero de Aguilera, ‘Las divisiones eclesiásticas y las relaciones entre la Iglesia y el Estado en la España de los siglos VI y VII’, M. J. Hidalgo de la Vega (ed.), Homenaje a Marcelo Vigil Pascual, Salamanca, 1989, pp. 169–90; C. Martin, ‘Las cartas de Montano y la autonomía episcopal de la Hispania septentrional en el siglo VI’, Hispania Antiqua, 22, 1998, pp. 403–26.

towards an archaeology of state formation in north-western iberia

580s. The regional elites that the Visigothic state had to deal with were very different from the contemporary aristocracies of south and east Iberia or the Narbonensis, and they seem to have maintained an unstable relationship with central power that became increasingly rarefied as state taxation retreated.40 The Visigothic state — a case of Wickham’s post-Roman ‘weak states’ — could not reverse the region’s situation. If anything, it arguably slowed down a process that resumed at high speed in the eighth century, after the Arab conquerors brought most of the Visigothic kingdom under the control of the Umayyad Caliphate of Damascus, but failed to establish similarly permanent bases in the north-west, a fact that became clear from the mid-eighth century onwards.41 The situation of NW Iberia between the mid-eighth and the mid-ninth century is reminiscent of the fifth-century statelessness discussed above. The region became a disarticulated periphery on the borders of the large-scale Emirate of Al-Ᾱndalus. From the 720s on, the emergence of a small polity in the central Asturian mountains and its later expansion is an essential component of the standard narratives of Spanish history centred upon the notion of Reconquista that were dominant in the mid-twentieth century and, in many respects, still enjoy much currency today.42 The Asturian expansion proceeded southwards in successive bands until the late eleventh century, when it came to dominate the entire Duero Plateau (fig. 2). However, the earlier phases of the Asturian kingdom are ill-understood, among other reasons because of the wholesale revision of the Asturian past that was effected during the reign of Alfonso III (866–910). This period marks the real take-off of the kingdom, and it finds its best expression in the group of texts commonly known as the Asturian chronicles.43 Precisely because it is only at this relatively late date that the first articulate narratives were produced, and because the number of preserved charters from before the 860s is almost negligible, historians have tended to overstress the social disruption across NW Iberia after 711. Starting in the 1920s, Sánchez-Albornoz advocated the theory that the Duero Plateau had essentially been depopulated in the eight century, and was not repopulated until efforts by the Asturian kings starting with Ordoño I (850–866).44 It was not until the late 1970s that this theory was strongly refuted by Barbero and Vigil. However, the alternative they offered was an overly primitivist model by which

40 See I. Martín Viso, ‘Tributación y escenarios locales en el centro de la Península Ibérica: algunas hipótesis a partir del análisis de las pizarras ‘visigodas’’, Antiquité Tardive, 14, 2006, pp. 263–90 and I. Martín Viso, ‘Tremisses y potentes’. On the kingdom’s higher aristocratic levels and their role in the construction of the Visigothic monarchy, see S. Castellanos, Los godos y la cruz. Recaredo y la unidad de Spania, Madrid, 2007. 41 E. Manzano Moreno, Conquistadores, Emires y Califas: Los Omeyas y la formación de Al-Andalus, Barcelona, 2006, esp. pp. 34–120; A. García Sanjuán, La conquista islámica de la Península Ibérica y la tergiversación del pasado, Madrid, 2013, pp. 384–439. 42 See the criticism in J. Escalona and I. Martín Viso, ‘The Life and Death of an Historiographical Folly: The Early Medieval Depopulation and Repopulation of the Duero Basin’, S. Barton and R. Portass (eds), Beyond the Reconquest, forthcoming. 43 J. Gil Fernández, J. L. Moralejo and J. I. Ruiz de la Peña (eds), Crónicas Asturianas: crónica de Alfonso III (Rotense y ‘A Sebastián’), Crónica Albeldense (y ‘Profética’), Oviedo, 1985. 44 C. Sánchez-Albornoz, Despoblación y Repoblación en el Valle del Duero, Buenos Aires, 1966; on the success and decline of this theory see J. Escalona and I. Martín Viso, ‘The Life and Death’.

43

44

J u l i o E s c a lo n a

all of NW Iberia would have been dominated by tribal social structures of pre-Roman origin that would have survived under the Roman and Visigothic dominations and, in the absence of any superior state, would have become widespread in the region until the emergence of feudalism.45 Neither of these views is accepted by most scholars today. Perhaps as a reaction, in the last two decades there has been a trend to consider that the Cantabrian north was as ‘Roman’ as any other part of Roman Hispania or, later, as ‘Visigothic’ as the rest of the kingdom. However, disregarding the extreme regional diversity within both states is unfortunately not very helpful. We can gain a more nuanced insight by considering that the strong Roman state could very well have had enough capillarity to permeate local settings and, thus, some aspects of governance and social structuring, which at first sight seem to depend on continuing state enforcement, could in fact have become embedded in the local societies and outlived the Roman or Visigothic states themselves. In recent years Wendy Davies has shown that the charters from NW Iberia contain traces of a legal culture of Late Antique pedigree that lived on locally but was gradually superseded in the tenth century by more recent uses fuelled by the Leonese kingship and the great monasteries. This included not only notions of ‘public’ justice in dispute settlement, but also significant elements of a documentary culture and charter-making practices that were strongly rooted in local secular contexts.46 Graham Barrett’s recent study on literacy points in a similar direction regarding the continuation of written culture.47 All this might indicate ‘residual statehood’, directly inherited from Late Antiquity, and expressed in a range of state-like ways of doing things — dispute settlement, the recording of legal affairs, and notions of law, rightfulness and authority — that went on without an operating state, simply as ‘the way things should be done’. Notions of public authority and the common good could very well form part of this ‘residual statehood’, and indeed could have favoured the reconstruction of state governance under the Leonese kings starting in the mid-ninth century, although they may also have run up against other elements, like the spread of aristocratic and ecclesiastical lordly powers.48 This can be seen as an example of the resilience of disrupted complex societies.49

45 A. Barbero de Aguilera and M. Vigil Pascual, La formación del feudalismo en la Península Ibérica, Barcelona, 1978. 46 W. Davies, Acts of giving. Individual, community and Church in tenth-century Christian Spain, Oxford, 2007; W. Davies, Windows on Justice in Northern Iberia, 800–1000, Abingdon, 2016; W. Davies, ‘Local Priests and the Writing of Charters in Northern Iberia in the Tenth Century’, J. Escalona and H. Sirantoine (eds), Chartes et cartulaires comme instruments de pouvoir. Espagne et Occident chrétien (viiie-xiie siècles), Toulouse, 2013, pp. 29–43; W. Davies, ‘Summary Justice and Seigneurial Justice in Northern Iberia on the Eve of the Millennium’, The Haskins Society Journal, 22, 2010, pp. 43–58. 47 G. Barrett, The Written and the World in Early Medieval Iberia, unpublished doctoral thesis defended at the University of Oxford, 2015, forthcoming soon in Oxford Historical Monographs (Oxford). I am most grateful to Dr Barrett for letting me use a copy of his unpublished work. 48 Á. Carvajal Castro, Bajo la máscara del Regnum: la monarquía asturleonesa en León (854–1037), Madrid, 2017, pp. 165–206. 49 See K. D. Morrison, ‘All is not lost when the center does not hold’ and, specifically for NW Iberia, I. Martín Viso, ‘Colapso político y sociedades locales: el Noroeste de la península ibérica (siglos VIII–IX)’, Reti Medievali Rivista, 17–12, 2016, pp. 335–69.

towards an archaeology of state formation in north-western iberia

b. Reintroduction

In parallel to the partial survival of Late Antique statehood, we may interpret as ‘re-introduction from outside’ the much undervalued role played by the Christian clerics from Al-Ᾱndalus who in the late ninth century relocated in the north and became the political aides of Asturian king Alfonso III. The newly arrived ecclesiastical elites brought with them a powerful political discourse that dramatically modified and enhanced the self-representation of the monarchy as the natural successor of the Visigothic state. This meant the deployment of a new ideological and legal weaponry built upon material from the Visigothic kingdom (the laws, a complex ecclesiastical organisation, notions of ideology and royal authority), which combined with the remnants of statehood that were already present in the north to create a powerful political programme aimed at buttressing Alfonso III’s position in the Iberian north.50 c. Satellitisation

The notion of ‘satellitisation’ can also be applied to NW Iberia, in at least two different directions. On the one hand, being the neighbour of the Emirate entailed steady contact with a powerful state that had a complex economy and an intellectual and artistic culture which necessarily had an impact beyond its borders. One aspect of this influence might be the adoption of an accounting system for valuations and transactions moulded upon Andalusi coinage. This seems to have expanded first in the plateau and more slowly in the further reaches like Galicia and Liébana, where the Visigothic system based on the gold solidus remained dominant for accounting until the early tenth century.51 Archaeologists are also increasingly aware of the fact that the material culture of NW Iberia, especially in the Duero Basin, includes a range of objects — mainly pottery — that can be interpreted as either imports from Al-Ᾱndalus or local production following southern models.52 The widespread occurrence of Arab personal names (often extending beyond given names to encompass the full Arabic onomastic system) within an utterly Christian context also suggests that, at least in the centre and south of the plateau, cultural contacts had resulted in some degree of

50 J. Escalona, ‘Family Memories. Inventing Alfonso I of Asturias’, I. Alfonso, H. Kennedy and J. Escalona (eds), Building legitimacy. Political discourses and forms of legitimation in medieval societies, Leiden / Boston, 2004, pp. 223–62. 51 E. Manzano and A. Canto, ‘The value of wealth: coins and coinage in Iberian early medieval documents (9th-11th centuries)’, R. Portass and S. Barton (eds), Beyond Reconquista, Leiden, 2019 (forthcoming). 52 A. Gutiérrez González and F. Miguel Hernández, ‘La cerámica altomedieval en León: producciones locales y andalusíes de Puerta Obispo’, J. Zozaya, M. Retuerce Velasco, M. Á. Hervás and A. De Juan (eds), Actas del VIII Congreso Internacional de Cerámica Medieval, 1, Ciudad Real-Almagro, 2009 pp. 443–62; A. Gutiérrez González and F. Miguel Hernández, ‘Cerámica altomedieval en León y su entorno’, A. Vigil-Escalera Guirado and J. A. Quirós Castillo (eds), La cerámica de la Alta Edad Media en el cuadrante noroeste de la Península Ibérica (siglos V–X): sistemas de producción, mecanismos de distribución y patrones de consumo, Bilbao, 2016, pp. 139–72.

45

46

J u l i o E s c a lo n a

bilingualism.53 On the other hand, the Asturian kingship seems to have undergone a crucial phase of growth in the late eighth and early ninth centuries, when establishing links with the Carolingian rulers became an important factor in gaining legitimacy and strengthening royal power, especially under Alfonso II (783 and 791–842) and up to the end of the reign of Louis the Pious (814–840).54 d. Incorporation

In the Iberian NW, the reconstruction of statehood was largely a consequence of its ‘incorporation’ into the Asturian-Leonese kingdom. However, this was a much smaller and weaker form of state and, crucially, in its earlier phases, amounted to hardly more than a state in the making. The Asturian expansion happened in several stages and differed greatly from one area to another, not only because of regional diversity but also because the Asturian kingship itself changed radically throughout the process. By the time this expansion gained momentum with Alfonso III, it had completely reinvented itself with the indispensable help of clerical aides from Al-Ᾱndalus, who brought with them the idea of a ‘Visigothic revival’.55 Alfonso III was presented as the legitimate heir of the Visigothic monarchs, and the champion of all Iberian Christians. Such a political programme has traditionally been considered the point of departure of the so-called ‘Reconquista’, but at the time it could hardly have been considered anything more than wishful thinking. Alfonso III could not foster much hope of actually ridding Spain of the Muslims, but claiming to be the only legitimate ruler aspiring to do so worked well for strengthening the roots of royal power within his realm and in the eyes of other Christian rulers. A crucial aspect of incorporation — in the Asturian case as in so many others — is two-way agency.56 Recent work by Álvaro Carvajal has shown that beneath the apparently uniform blanket of Asturian-Leonese royal power, the areas where kings really held direct control were rather limited. The county of Castile on the eastern borders is an extreme case of this situation.57 Most of the kingdom’s territory was connected to central power by means of two different kinds of interposed actors: either the high aristocracy, which was firmly linked to the monarchy, or lesser, less defined groups that held local control of specific areas. Carvajal’s work is particularly important in that he identifies some of the latter as the native elites of the Duero

53 See the pioneering work by V. Aguilar Sebastián and F. Rodríguez Mediano, ‘Antroponimia de origen árabe en la documentación leonesa (siglos VIII–XIII)’, El Reino de León en la Alta Edad Media, 6, 1994, pp. 497–633. 54 J. Escalona, ‘Family Memories’, esp. at pp. 226–32. 55 For a detailed analysis of the neo-Visigothic ideology, see T. Deswarte, De la destruction à la restauration: l’idéologie du royaume d’Oviedo-León (viiie-xie siècles), Turnhout, 2003. 56 J. Escalona, ‘The early Middle Ages: a scale-based approach’, J. Escalona and A. Reynolds (eds), Scale and Scale Change in the Early Middle Ages: Exploring landscape, local society and the world beyond, Turnhout, 2011, pp. 9–30, pp. 17–20. 57 J. Escalona, ‘In the name of a distant king: Representing royal authority in the county of Castile, c. 900–1038’, Early Medieval Europe, 24–21, 2016, pp. 74–102.

towards an archaeology of state formation in north-western iberia

Plateau, who had to negotiate their incorporation into the emergent polity.58 The Asturian-Leonese kingdom is thus reframed as a patchwork of different patterns of authority that became increasingly interwoven thanks not only to the development of royal governance, but also to the spread of great lordly, and especially monastic, domains. Given the paucity of the resources that the ninth- and tenth-century kings could deploy, it is unlikely that they could have subjected the vast regions south of their original Asturian core without local involvement. The local and micro-regional elites of the Duero Basin might have played an important role in the incorporation into the Asturian realm, a process that has traditionally been seen only as a top-down imposition, and in which the importance of bottom-up agency is only beginning to emerge.59 The notions of ‘satellitisation’ and ‘incorporation’ can also help theorise the situation of specific territories within the Asturian-Leonese kingdom, like Castile. Rather than just an administrative district, Castile was more of a satellite on the kingdom’s eastern fringes. The fact that since the 880s Castile underwent its own territorial expansion, as another independent polity, is a crucial, often overlooked indicator. Given the limited direct agency of the Asturian-Leonese kings inside Castile, and the great autonomy of its counts, I have suggested elsewhere that the Castilian elites conceivably found in their connection to the Asturian kingdom the fuel to join forces and expand their territory up to a scale at which, despite their limited seigneurial resources, they could consolidate their social dominance.60 Nevertheless, ‘elite’ is an eminently relational notion. Any elite must be defined in the terms of the society over which it enjoys superiority, and there is much ground for variation. In tenth-century NW Iberia there were local elites in the rural communities, and also in central places (whether ‘urban’, ‘proto-urban’ or otherwise),61 and there were different scales of supralocal elites up to the aristocracy in the royal entourage. The development of statehood might have been in the interest of many or all of those elites, albeit in different ways. The local elites of the Duero Basin that were incorporated — willingly or not — into the Asturian kingdom found themselves in a wholly new social landscape, with a huge social distance between the political core and the localities. There was much more room for social differentiation and personal promotion. Even if the local communities might have tended to resist the 58 Á. Carvajal Castro, Bajo la máscara del Regnum, pp. 128–30. 59 J. Escalona and F. Reyes Téllez, ‘Scale change on the border: The county of Castile in the tenth century’, J. Escalona and A. Reynolds (eds), Scale and Scale Change in the Early Middle Ages: exploring landscape, local society and the world beyond, Turnhout, 2011, pp. 153–83; J. Escalona, ‘Military stress, central power and local response in the county of Castile in the tenth century’, J. Baker, S. Brookes and A. Reynolds (eds), Landscapes of Defence in Early Medieval Europe, Turnhout, 2013, pp. 341–67. 60 J. Escalona, ‘In the name of a distant king’. 61 I think the towns of this period are better understood not in terms of economic specialisation or market functions, but as communities controlling a hinterland with subordinated rural settlements and some sort of surplus flow towards the centre. See J. Escalona, Sociedad y territorio en la Alta Edad Media castellana: la formación del alfoz de Lara, Oxford, 2002, pp. 223–31. See also J. Escalona, ‘SupraLocal Territorial Units: on the origins of Behetría lordship’, C. Jular and C. Estepa (eds), Land, Power and Society in medieval Castile, Turnhout, 2009, pp. 29–58.

47

48

J u l i o E s c a lo n a

newly imposed power, their elites had a choice of strategies at their disposal. They could perhaps reinforce their distinction from ordinary community members (for example by emulating the aristocratic way of life: holding weaponry, having a horse, perhaps building a chapel or a tower); they could try to escape the community and join the lower ranks of the aristocracy; or, failing that, they could use their local leadership to become seigneurial agents. All of these avenues are documented in the tenth-century charters from NW Iberia62.

The Archaeological Visibility of State Formation in NW Iberia To detect the archaeological traces of the state has long been a major challenge, and it remains so in many respects. Archaeologists naturally tend to focus upon the kinds of material culture that can be more readily linked to state organisation, like large-scale irrigation networks, complex systems of military defence, large-scale monumentality (secular or sacred) or coinage.63 All these, however, are easier to relate to fully operating states than to state formation processes. The aforementioned caveat about the resilience of state-related cultural traits after collapse applies here too. In NW Iberia the situation is no different. In the absence of coinage (the Asturian-Leonese monarchy did not strike any), maybe the most obvious — and most commented upon — archaeological indicator is the investment in monumentality, namely churches. Building and patronising churches is a prominent attribute of the Asturian kings beginning in the early or mid-ninth century (depending on whether the royal-founded church of Santullano (Oviedo) is considered as dating from the reign of Alfonso II (783 and 791–842) or Alfonso III (866–910) as recently proposed by M. A. Utrero).64 During the reign of Ramiro I (842–850) there seems to have emerged a building programme fostered by the monarchy in the so-called ‘Asturian style’ that irradiated from the royal see in Oviedo to the surrounding region, but — significantly — not to the whole kingdom. Starting one half century later, and extending well into the tenth century, the growth of the Asturian-Leonese kingship under Alfonso III produced a new constructive programme, this time heavily influenced by the sophisticated architecture of the Caliphate of Córdoba, a style that is commonly — if inaccurately — termed ‘Mozarabic’. This programme had its centre in the new royal see of León, from which it mainly irradiated to the surrounding areas in the central Duero Plateau. The result is an assemblage of top-quality buildings that, albeit quite different from one another structurally, keep to a common aesthetic language. Over the tenth century, however, the same style was adopted by a wider range of buildings. Therefore, not every church in the 62 See the chapters by Iñaki Martín Viso and Álvaro Carvajal in this volume. 63 See the classical discussion in B. Trigger, ‘The archaeology of government’, World Archaeology, 6, 1, 1974, pp. 95–106. 64 M. Á. Utrero Agudo, ‘La Basílica de San Julián de los Prados, Alfonso III, Selgas y Schlunk’, M. Á. Utrero Agudo (ed.), Iglesias altomedievales en Asturias: arqueología y arquitectura, Madrid, 2016, pp. 11–38.

towards an archaeology of state formation in north-western iberia

‘Mozarabic’ style must necessarily be seen as the product of royal initiative. It is safer to simply consider that they all speak a language that was ultimately promoted from the political centre. Taken together, the two Asturian styles seem to mark two crucial steps in the development of political centralisation. Not every church is an expression of statehood, though. Most of the hundreds of churches that are recorded in the ninth- and tenth-century charters from these regions should be ascribed to the initiative of local communities, local elites, or aristocrats, with no primary state involvement. Moreover, building churches and monasteries seems to have been a strategy for middle-ranking elites seeking to buttress or enhance their position and make it durable.65 Defence arrangements are much more difficult to gauge. A large number of defensive sites and fortifications have been detected that are coetaneous with the rise of the Asturian monarchy, but most of them are impossible to connect to any centre-driven large-scale programme of defence.66 The written sources sometimes trace royal associations for specific sites, but only one such case has been thoroughly investigated. This is the site of Gauzón (Asturias), which recent excavations define as a coastal stronghold controlling access to the Avilés estuary.67 There is no indication, however, that Gauzón formed part of a defensive network with other fortifications. On the other hand, linear earthworks similar to, but shorter than Wansdyke or Offa’s Dyke in England, have recently been detected on the summits of the Asturian mountains, controlling the old Roman road that connected the plateau to the port of Gijón on the Asturian coast68. Fortifications of this kind are very difficult to date with any precision. Unsurprisingly, given the bearing that inherited historical commonplaces

65 J. J. Larrea Conde, ‘Construir iglesias, construir territorio: Las dos fases altomedievales de San Román de Tobillas (Álava)’, J. López Quiroga, A. M. Martínez Tejera and J. Morín de Pablos (eds), Monasteria et Territoria. Élites, edilicia y territorio en el Mediterráneo medieval (siglos V–XI), S1720, Oxford, 2007, pp. 321–36; J. A. Quirós Castillo and I. Santos Salazar, ‘Founding and Owning Churches in Early Medieval Álava (North Spain): The Creation, Transmission and Monumentalization of Memory’, J. C. Sánchez-Pardo and M. Shapland (eds), Churches and Social Power in Early Medieval Europe. Integrating Archaeological and Historical Processes, Turnhout, 2015, pp. 35–68. On the archaeological recognition of early medieval fabrics, see L. Sánchez Zufiaurre, Técnicas constructivas medievales. Nuevos documentos arqueológicos para el estudio de la Alta Edad Media en Álava, VitoriaGasteiz, 2007. 66 For a recent review of fortifications, see J. A. Quirós Castillo, ‘Defensive Sites of the Early Middle Ages in North-West Spain’, J. Baker, S. Brookes and A. Reynolds (eds), Landscapes of Defence in Early Medieval Europe, Turnhout, Brepols, 2013, pp. 301–39. See also J. A. Quirós Castillo and J. M. Tejado Sebastián (eds), Los castillos altomedievales en el noroeste de la Península Ibérica, Documentos de Arqueología Medieval, 4, Bilbao, 2012. 67 I. Muñiz López and A. Álvarez Busto, ‘El Castillo de Gauzón (Asturias. España) y el proceso europeo de feudalización entre la Antigüedad tardía y la Edad Media a través de las fortificaciones’, Munibe, 61, 2010, pp. 289–328. 68 J. Camino Mayor, R. Estrada García and Y. Viniegra Pacheco, ‘A propósito de las fortificaciones lineales ástures de El Homón de Faro (La Carisa) y El Muro (La Mesa)’, Territorio, sociedad y poder, 2, 2007, pp. 53–64; A. Gutiérrez González, ‘Fortificaciones visigodas y conquista islámica del norte hispano (c. 711)’, Zona Arqueológica (Special Issue 711. Arqueología e historia entre dos mundos, I), 15, 2011, pp. 335–52.

49

50

J u l i o E s c a lo n a

have on the interpretation of any sort of Asturian archaeological evidence, these linear works have been presented as a response to the Arab invaders of 711AD by the Asturian rebels that first challenged them in the probably mythical battle of Covadonga, and then proceeded to found the Asturian kingship under their first king Pelayo. In theory, to build, man and maintain such a wall would require some kind of continuous centralised organisation, which seems more likely in a situation of repeated, systematic attacks, as occurred in NW Iberia from the very late eighth century onwards. However, in the current state of research it seems safer not to put too much interpretative weight upon the linear earthworks. The best evidence so far for large-scale military arrangements comes from the Asturian expansion on the Duero River in the first years of the tenth century. On the western side, the textual evidence from the chronicles finds a very good match in Iñaki Martín Viso and Álvaro Carvajal’s analysis of the strongholds south of the Duero River that were ephemerally seized and fortified by the Asturian monarchy, but were soon lost to the newly-founded Caliphate when it sought to reinforce its positions on the north-western frontier.69 The creation of a line of fortresses was accompanied by an effort to incorporate the area’s elites that left an imprint via the apparition of a distinct kind of neatly-arranged cemetery associated with the strongholds, which stand in contrast to the existing pattern of smaller, less organised burial sites.70 On the eastern side of the Duero Valley, similar processes formally took place in the name of the kings, but were actually carried out under the aegis of a group of Castilian aristocrats.71 The textual evidence shows how the defence system on that stretch of the frontier was organised on the basis of fortified strongholds evenly spaced along the natural border formed by the river, together with a network of lookouts where garrisons provided by the local elites conducted surveillance service. The latter, however, have only occasionally been investigated archaeologically.72

69 I. Martín Viso and Á. Carvajal Castro, ‘Historias regionales de la repoblación: Los reyes asturleoneses y las ‘políticas de la tierra’ en el oeste de la meseta del Duero’, P. C. Díaz Martínez, F. Luis Corral and I. Martín Viso (eds), El Historiador y la Sociedad. Homenaje al Profesor José Mª. Mínguez, Salamanca, 2013, pp. 39–52. 70 I. Martín Viso, ‘Ancestors and Landscape: Early Medieval Burial Sites in the Central-Western Regions of the Iberian Peninsula’, J. Escalona, O. Vésteinsson and S. Brookes (eds), Polity and Neighbourhood in Early Medieval Europe, Turnhout, forthcoming. 71 C. Estepa Díez, ‘La Castilla primitiva (750–931): condes, territorios y villas’, F. J. Fernández Conde and C. García de Castro Valdés (eds), Poder y Simbología en Europa, siglos VIII–X (Actas del Symposium Internacional, Oviedo, 22–27 de septiembre del 2008), Oviedo, 2009, pp. 261–78; I. Santos Salazar, ‘Competition in the frontiers of the Asturian kingdom: The comites of Castile, Lantarón and Álava (860–940)’, R. Le Jan, G. Bührer-Thierry and S. Gasparri (eds), Coopétition. Rivaliser, coopérer dans les sociétés du Haut Moyen Âge (500–1100), Turnhout, 2018, pp. 231–51. 72 J. Escalona, ‘Comunidades, territorios y poder condal en la Castilla del Duero en el siglo X’, Studia Historica. Historia Medieval, 18–19, 2000–2001, pp. 85–120; J. Escalona, ‘Military stress’. An early — but now rather outdated — archaeological exploration of fortifications and watchtowers along the Duero frontier is F. Reyes Téllez and M. L. Menendez Robles, ‘Sistemas defensivos altomedievales en las comarcas del Duratón- Riaza (siglos VIII–X)’, II Congreso de Arqueología Medieval Española, 3, Madrid, 1987, pp. 631–39.

towards an archaeology of state formation in north-western iberia

Perhaps the most obvious conclusion to be drawn from this rather meager survey is that the quest for state formation cannot limit itself to prominent top-down expressions of centralisation like monuments and defences. In fact, no single site or group of sites holds the key to understanding state formation. It is necessary to adopt a more systemic approach focusing upon pivotal changes in social complexity. As noted at the outset, state formation is a way of stabilising and reproducing social inequality. The traces of this sort of evolution can be expected to appear, on the one hand, at elite sites that, even if not royal, can help define the class interests behind state formation. So far NW Iberia has not revealed any secular aristocratic residence comparable to the luxurious late-seventh or early-eighth century villa of Pla de Nadal (Valencia) or the lavish Islamic-style villas (almunias) surrounding tenth-century Córdoba.73 However, from the tenth century there is significant written evidence for the existence of estate centres (palatia) that served to organise the largely dispersed ensembles of property that constituted most aristocratic estates and where rent in kind and livestock was stored.74 It is only recently that pioneering work by J. A. Quirós and his group has begun to unearth the first material evidence for this kind of sites, but much more work remains to be done in order to characterise the different types of palatia, from royal ones to those of the regional elites.75 In detecting and investigating such sites, it will be crucial to bear in mind that although estate centres were locally situated, they nevertheless operated well beyond their surrounding territory, through territorially discontinuous long-distance connections.76 In many senses, building churches and estate centres can be seen as an expression of the growth of the aristocracies, the preferential beneficiaries of state formation. The process, however, is wider and more complex. It is connected to social change (an increase in social difference and the creation of a bounded peasantry),77 and

73 E. Juan Navarro and J. V. Lerma Alegría, ‘La villa aúlica del “Pla de Nadal” (Riba-Roja de Túria)’, A. Ribera i Lacomba (ed.), Los orígenes del cristianismo en Valencia y su entorno, Valencia, 2000, pp. 135–42. For Córdoba, see above, n. 39. 74 J. Escalona and I. Martín Viso, ‘Los palatia, puntos de centralización de rentas en la Meseta del Duero (siglos IX–XI)’, A. Vigil-Escalera, G. Bianchi and J. A. Quirós (eds), Horrea, barns and silos. Storage and Incomes in Early Medieval Europe, Bilbao, 2013, pp. 103–26. J. Á. García de Cortázar y Ruiz de Aguirre and M. E. Peña Bocos, ‘El palatium, símbolo y centro de poder en los reinos de Navarra y Castilla en los siglos XI a XII’, Mayurqa, 22 (Homenatge al Prof. Alvaro Santamaría), 1989, pp. 281–96. 75 I. Santos Salazar, ‘Words, things and social inequality: The village of Torrentejo’, J. A. Quirós Castillo (ed.), Social Complexity in Early Medieval Rural Communities. The north-western Iberia archaeological record, Oxford, 2016, pp. 125–33; see a summary of the site’s archaeological sequence in J. A. Quirós Castillo and C. Tejerizo García, ‘Despoblado de Torrentejo o Santa Lucía’, Arkeoikuska, 16, 2016, pp. 81–85. 76 See the enlightening discussion of dispersed Carolingian domains in J.-P. Devroey, ‘Gérer et exploiter la distance. Pratiques de gestion et perception du monde dans les livres fonciers carolingiens’, P. Depreux, F. Bougard and R. Le Jan (eds), Les élites et leurs espaces: mobilité, rayonnement, domination (du vie au xie siècle), Turnhout, 2007, pp. 49–65. For Castile, see the discussion and cartography of the vast domains of the ‘comital’ monasteries of Covarrubias and Oña in J. Á. García de Cortázar y Ruiz de Aguirre and E. Peña Bocos, ‘Poder condal ¿y mutación feudal? en la Castilla del año mil’, M. I. Loring Garcia (ed.), Historia social, pensamiento historiográfico y Edad Media. Homenaje al Prof. Abilio Barbero de Aguilera, Madrid, 1997, pp. 273–98. 77 See the always illuminating work of R. Faith, The English Peasantry.

51

52

J u l i o E s c a lo n a

this in turn depends on overall economic growth. The connection between early medieval economic growth and increasing social and political complexity has been formulated, among other areas, for Scotland and Ireland, and can be pursued for other European regions as well.78 Increasing division of labour, with specialised personnel (craftsmen, bureaucrats, warriors, clerics) working full time for the elite is one of its visible traces, as is the spread of technological advances such as the water mill, extensively evidenced in the surviving charters from NW Iberia, but as of yet largely under-investigated archaeologically. The archaeology of agricultural fields and irrigation could similarly benefit from fresh approaches.79 Specialised production resulting in supra-local economic dependencies, and centre-to-hinterland articulations is another promising vector. The charters from the Duero Plateau contain indicators that seem to point to an early phase of urban development around the 950s and 960s which is not yet archaeologically defined but could be in the near future.80 A more traditional field, such as the study of pottery, could contribute a great deal by helping to differentiate local and micro-regional productions from long-distance imports and from the first wave of native large-scale ceramic styles, as seems to be the case of the red painted pottery in the county of Castile.81 In recent years, the development of archaeobiology is providing new avenues for the study of inequality at the village level.82 This crucial but difficult issue can be studied through the anthropological evidence of cemeteries or plant and animal remains, but also through the spread of churches and the creation of a parish system. However, a view to the broader picture is indispensable. While wealth or diet differences in a local cemetery may reveal inequality within the community, socially ‘flat’ cemeteries may actually indicate the consolidation of stronger social inequality beyond the village level.

Conclusions By way of conclusion, it is first of all worth admitting that, as of today, the archaeology of state formation in NW Iberia looks more like a wish list than anything else.

78 See S. Driscoll, W. S. Hanson and E. A. Slater, ‘The archaeology of State formation in Scotland’, 1991, pp. 81–111 and W. Davies, ‘Economic change in Early Medieval Ireland: the case for growth’, Settimane di studio (Centro Italiano di Studi sull’Alto Medioevo), Spoleto, 2010, pp. 111–33. 79 Ongoing work by Wendy Davies on gardens and by Margarita Fernández Mier and Juan Antonio Quirós on common land and common rights may provide interesting developments on this issue in the near future. 80 J. Escalona, ‘Aspectos del Pastoreo en la Península Ibérica’, Historia Agraria, 11, 1996, pp. 181–86. 81 F. Grassi, ‘Production, consumption and political complexity: early medieval pottery in Castile and Southern Tuscany (7th-10th centuries)’, J. A. Quirós Castillo (ed.), Social complexity in early medieval rural communities. The north-western Iberia archaeological record, Oxford, 2016, pp. 91–112; see also her contribution to this volume. 82 See the contributions to J. A. Quirós Castillo (ed.), Demografía, paleopatologías y desigualdad social en el noroeste peninsular en época medieval, Documentos de Arqueología Medieval, 10, Bilbao, 2016 and to J. A. Quirós Castillo (ed.), Social Complexity in Early Medieval Rural Communities. The north-western Iberia archaeological record, Oxford, 2016.

towards an archaeology of state formation in north-western iberia

There are, however, reasons for anticipating major advances on this issue. The most important is perhaps that the decades-long gap between historians and archaeologists is shrinking, not just in methodological terms (that is, the joint use of texts and the archaeological record) but, above all, theoretically. Younger scholars are willing to engage critically on the two strands of thought about state formation that I discussed in the first part of this paper. Those diverging paths can and should converge, if only because it is now clear that the topic cannot be dealt with satisfactorily from either of them separately. As usual with early medieval NW Iberian archaeology, the main problem is one of critical mass. The corpus of excavated sites is still too small and fragmentary to enable sound generalisations, let alone inter-regional comparisons. Unless a new building and development frenzy takes hold in the country for a second time this century (something that no one in their right mind would hope for), the body of empirical data can only be expected to increase gradually by means of painstaking, characteristically under-funded, research projects. This state of affairs does, however, have a silver lining. Slower progress may pay off if it allows the fieldwork to be steered towards more consistent research-driven strategies.83 Regarding state formation, it is nevertheless essential to bear in mind the signs that we will recognise as indicators of early statehood. I hope that my discussion has made clear that no single criterion — whether textual or archaeological — will suffice to spot our nascent early medieval states. A systemic, holistic approach is needed, and its major target must be the detection of durable, institutionalised inequality. There is no magic wand for this task. In general, there is much work to be done, but also a very promising future for thinking about state formation in NW Iberia. Every aspect of archaeological research ultimately ties in to this all-encompassing topic. The study of state formation calls for the consideration of changes in agrarian technology and in landscape change in general, as well as monument building or defence as indicators of elite growth, the structuring of inequality, and the implementation of institutionalised governance. Imaginative approaches are needed in order to build a case for the growth of statehood, and they should bear as a major component the four variants of secondary state formation outlined above (as well as other possible variants). Distinguishing emerging centre-driven from locally resilient statehood will be a major challenge. Ultimately, early medieval states did not have any anniversary to celebrate. State formation is a gradual, non-linear process. Future models will have to allow not only for the slow development of state procedures, but also, crucially, for very uneven implementation in local and regional contexts. The extremely variable patterns of seigneurial, ecclesiastical and royal authority that have been so masterfully described by the late Carlos Estepa for medieval central Castile are obviously rooted in the early medieval period.84

83 J. Escalona, ‘L’archéologie médiévale chrétienne en Espagne: entre recherche et gestion du patrimoine’, E. Magnani (ed.), Le Moyen Âge vu d’ailleurs: voix croisées d’Amérique latine et d’Europe, Dijon, 2009, pp. 253–74. 84 C. Estepa Díez, Las behetrías castellanas, Valladolid, 2003.

53

Robin A. Beck

Maize, Mounds, and Cosmos Durable Inequality in the Mississippian World (ad 1000–1250)

The concept of inequality has long been central to archaeological research on the Mississippian phenomenon, which spread throughout much of what is now the eastern United States from ad 1000 to 1600. Social inequality seems to be among those factors that make Mississippian Mississippian — that distinguish the social organizations of Mississippian peoples from those of others — their antecedents, contemporaries, and descendants — who structured their social relationships according to other principles. It is just as critical, at the same time, for us to appreciate the extraordinary range of diversity among the varied peoples, societies, or cultures that we are inclined to bundle together under this Mississippian label. Given that such problems are relevant to recent studies of Early Medieval Europe, there is much to recommend considering the Mississippian emergence as part of a broad, comparative perspective on the Early Medieval. Both Mississippian and the Early Medieval witnessed a widespread intensification of agricultural economies across their respective places — along with the related developments of social inequality and political hierarchy — and both phenomena unfolded over continental or semi-continental scales. In this chapter, I offer a perspective on the emergence of social inequality in eastern North America through the lens of the greatest Mississippian site, Cahokia, located along the Mississippi River near modern-day St Louis. I suspect that many of the same processes that unfolded at Cahokia were at play in Europe during the Early Medieval. Yet before exploring, if only briefly, the archaeology of Cahokia, I would like to discuss what we mean by inequality: what it is (and maybe what it is not), how it happens, what it means, and how we recognize it. To do so, I draw both on anthropological perspectives — particularly those of Gerald Berreman and Eric Wolf — and the work of sociologist Charles Tilly. Drawing these different approaches together, I argue that we should distinguish status and class as distinct pathways to inequality, the former structuring inequality between individuals, the latter between groups. I then turn to Mississippian Cahokia to show how structural bottlenecks provide opportunities for groups to create what Tilly refers to as durable inequalities, those that persist across generations and thus constitute the basis for

Robin A. Beck  •  University of Michigan Social Inequality in Early Medieval Europe: Local Societies and Beyond, ed. by Juan Antonio Quirós Castillo, Turnhout : Brepols, 2020 (HAMA 39), pp. 55–72 © FHG10.1484/M.HAMA-EB.5.118445

56

ro bin  a . bec k

class relations. I suggest that any such perspectives need to be grounded in historical relations rather than the general, evolutionary categories and stages.

Perspectives on Inequality Over the past few years, I have grown dissatisfied with the foundations of our current approach to inequality. And by ‘our’ I mean many of us working from the perspective of anthropological archaeology in the United States. It is an approach that I believe traces its foundations back to Ralph Linton, whose highly influential Study of Man, published in 1936, still drives much of our current thinking about inequality, whether we always recognize it or not.1 It was here that he defined the two terms most American archaeologists use almost automatically when writing about inequality: achieved status and ascribed status. In Linton’s original description, ascribed statuses are those ‘assigned to individuals without reference to their innate differences or abilities. They can be predicted and trained for from the moment of birth.’2 Achieved statuses, on the other hand, are ‘at a minimum, those requiring special qualities…They are not assigned to individuals from birth but are left open to be filled through competition and individual effort.’3 Significantly, he continues that ‘The majority of the statuses in all social systems are of the ascribed type, and those which take care of the ordinary day-to-day business of living are practically always of this type.’4 This is not, to my reading, how we usually think about achieved and ascribed status today. Linton, for example, defined status as simply ‘a position in a particular pattern…a collection of rights and duties.’5 For Linton, then, status differences — whether of the achieved or ascribed type — need not be associated with vertical ranking. What has happened in the intervening years, it seems, at least in archaeological use, is that Linton’s treatment of status has come to be replaced by one that equates status with rank. Achievement (rank by personal accomplishment) and ascription (rank by birthright) are thus the two pathways for reckoning privilege and together constitute the foundational principles of social inequality. This shift from status as position to status as rank coincided with the search for social types during the 1960s, and Linton’s distinction between achieved and ascribed statuses was drawn inexorably into a growing neoevolutionary paradigm. In neoevolutionary logic, ascribed status — now explicitly associated with social inheritance (as opposed to biological inheritance: characteristics such as sex, eye color, or skin pigmentation) — evolved from achieved status as a defining feature of complex societies. Band-level societies were described as egalitarian and lacking



1 R. Linton, The Study of Man: An Introduction, New York, 1963. 2 R. Linton, op. cit. not. 1, p. 115. 3 R. Linton, op. cit. not. 1, p. 115. 4 R. Linton, op. cit. not. 1, p. 115. 5 R. Linton, op. cit. not. 1, p. 113.

m ai ze , mo u nd s, and co smo s

rank altogether; achieved status came to be associated with rank in tribal societies; and ascribed status, believed to have first emerged in chiefdoms, was fully developed in the context of state-level institutions. As competing models and social typologies proliferated, we saw (in addition to band, tribe, chiefdom, and state) the identification of rank societies, class societies, stratified societies, transegalitarian societies, complex hunter-gatherers, middle range societies, and others still. Any attempt to put order to these different terminologies, most all of which are products of neoevolutionary logic, seems impossible today, if not counterproductive. Why counterproductive? Unilineal evolution may be in ruins, but many of its foundational logics remain embedded in our discourse, in the way we think about the problems of human social organizations and the ways that these organizations have changed through time. My aim is not to flog a dead horse, nor to suggest that every vestige of neoevolutionary thinking needs to be purged from our theoretical toolboxes. But I do think that using achieved and ascribed statuses as proxies for different social types has for too long impeded a better understanding of inequality itself. Ultimately, treating such statuses as markers for social types in an evolutionary scale has kept us from the business of historicizing inequality, of treating it as a contingent phenomenon rather than as a universally inevitable condition of social life. What follows is a work-in-progress. What I hope it offers is a less trodden way of looking at inequality from a general, comparative perspective, yet one that takes history and context at least as seriously as it takes social types. Status and Class

Let me begin by observing that all inequalities are fundamentally rooted in what I refer to as qualities of difference; we can think of these as the building blocks of inequality. One of the most pervasive problems with the distinction between achieved status and ascribed status is that it focuses almost exclusively on the relative positions of individuals in social hierarchies and the means of securing such individual positions. I suggest instead that qualities of difference are relevant both at the scale of the individual and at the scale of the group (see fig. 1). We can refer to the practice of attaching social significance to such qualities at the scale of the individual as differentiation. Attaching significance at the scale of the group — that is, constructing culturally meaningful categories of people — is categorization. Yet neither differentiation nor categorization constitute inequality in themselves. We can recognize differences between ourselves and others, as well as form and recognize different kinds of social groups, without ranking these differences or assigning them relative social worth. Here it is useful to consider Gerald Berreman’s ideas on inequality, which he developed from years of ethnographic fieldwork among the social castes of India.6 His general perspective links three terms that merit discussion: inequality, dominance,



6 G. Berreman, ‘Social inequality: A cross-cultural analysis’, in G. Berreman and K. M. Zaretsky eds, Social inequality: Comparative and developmental approaches, New York, 1981, pp. 3–40.

57

58

ro bin  a . bec k

and social inequality. Berreman defined inequality as ‘the social [moral] evaluation of whatever differences are regarded as relevant in a given society or situation;’ dominance is ‘the behavioral expression of these differences;’ social inequality, then, is ‘the combination of these two processes, inequality and dominance’.7 I would like to briefly consider these three concepts and their usefulness. First, for Berreman, it is the moral evaluation of difference that distinguishes inequality from cultural sorting (that is, from the recognition of difference). This seems reasonable. Yet Berreman’s framework does not recognize differentiation and categorization as distinct processes; moreover, the terminological redundancy in his use of ‘inequality’ and ‘social inequality’ is an unnecessary source of potential confusion. Instead, I suggest that a moral evaluation of difference at the scale of the individual results in discrimination. At the scale of the group, it results in segregation (fig. 1). Each of these terms denotes a different route to what Berreman glosses as inequality. He goes on to argue, though, that this condition of inequality only becomes ‘social inequality’ when joined with dominance. As Kenneth Ames points out, Berreman’s use of dominance differs from its more conventional use, and we might think here of power instead of domination.8 But in the context of establishing social inequality, the relevant mode of power is less power over — the power to impose one’s will through social interaction — and more power to, a capacity to wield what Eric Wolf refers to as structural power, the power to set the agenda: By this I mean the power manifest in relationships that not only operates within settings and domains but also organizes and orchestrates the settings themselves, and that specifies the direction and distribution of energy flows. In Marxian terms, this refers to the power to deploy and allocate social labor.9 Pulling these conceptual strands together, and dropping the ‘social’ from Berreman’s ‘social inequality,’ I suggest that the application of structural power to the moral evaluation of difference creates inequality at the scale of the individual and at the scale of the group. When we apply structural power to discrimination (me vs. you at the scale of the individual), we create status. When we apply structural power to segregation (us vs. them at the scale of the group), we create class (fig. 1). In either case, inequality is only established, to return to Wolf ’s point, when the moral evaluation of difference meets the allocation of labor. Earlier I observed that qualities of difference are the building blocks that make inequality through status and class relations possible. Those qualities amenable for the reckoning of status are usually more apparent than those amenable for the reckoning of class. This is unsurprising, since discrimination is based on specific qualities. Segregation, however, is based on general and — more often than not — arbitrary qualities Let me give some examples.

7 G. Berreman, op. cit. not. 1, p. 4. 8 K. M. Ames, ‘The archaeology of rank’, in R. A. Bentley, H. D. G. Maschner and C. Chippendale eds, Handbook of archaeological theories, Lanham, 2007, pp. 487–513. 9 E. Wolf, Envisioning power: Ideologies of dominance and crisis, Berkeley, 1999, p. 4.

m ai ze , mo u nd s, and co smo s

Fig. 1. Modes of inequality: status and class.

The qualities of difference that inhere in individuals are specific and can often be evaluated as continua. We sort one another by age; by color of hair, eyes, and skin; by height and weight; by speed, strength, or testing ability. This is not simply a means for re-labelling achieved status. Indeed, we inherit many of these qualities as part of our individual genetic packages, and the moral values we attach to those we inherit, just as to those we achieve, are potential sources of discrimination that can be combined with structural power to establish status inequality. It is ironic, given the degree to which we have used the genetically ascribed trait of light skin for such purposes, that an increased pigmentation achieved through tanning can now be seen to benefit one’s social standing. Many individual qualities that directly affect status fall somewhere in between strict inheritance and strict achievement. Adult height, for example, is determined both by genetic inheritance and childhood nutrition, itself the product of economic opportunity (Wolf ’s allocation of social labor). Thus, the lines between ascribed and achieved status are hopelessly tangled at the scale of individual qualities of difference.

59

60

ro bin  a . bec k

In contrast to the specific qualities that inhere in individuals, those qualities of difference that promote and justify segregation among groups, thus creating the context for class, are general and arbitrary. They are typically understood in terms of binary oppositions, including clean and unclean, feminine and masculine, upper and lower, superior and inferior, elite and commoner, believer and infidel, friend and foe. Binary oppositions are so prevalent in contexts of class and segregation because they ultimately all derive from the same, primordial binary of kin vs. non-kin. Different groups may claim the same morally superior qualities for themselves while imposing the same morally inferior qualities on others. It should be clear, I hope, that such qualities are in no inherently obvious way manifested either at the scale of the group itself or at the scale of those individuals that comprise it at any given point in time. The qualities of difference that promote segregation and class at the scale of the group are not inherent to the group; rather, they are claimed by the group. As such, they must be manifested in ways that make them seem inherent and that are difficult for outsiders — those not recognized as members of the class — to emulate. These include: ornamentation and dress, particularly incorporating rare or exotic materials; ritual or religious prerogatives; specific foodways, tastes, and table manners; tattooing and cranial modification; and spatial segregation in life and in death. Durable Inequality

Status inequalities are less durable than those that derive from class because they are inherent to specific individuals with necessarily limited life spans. This brings me to the work of American sociologist Charles Tilly and his efforts to account for how social categories often result in systematic, long-term inequalities. Tilly’s most extensive treatment of the issue was his widely acclaimed book Durable Inequality, published in 1998 but referenced only infrequently by archaeologists. As its title suggests, Tilly focuses on durable inequalities, ‘those that last from one social interaction to the next, with special attention to those that persist over whole careers, lifetimes, and organizational histories’.10 Accordingly, he turns attention from those continua that distinguish status relations — (tall/short), (light/dark), (young/old) — to those categorical inequalities that pertain to ‘distinctly bounded pairs such as female/male, aristocrat/plebian, citizen/foreigner [recall the binary oppositions that I discussed earlier] and more complex classifications based on religious affiliation, ethnic origin, or race.’11 Categorical inequality, in sum, constitutes class. Tilly’s extended argument is multi-faceted, and I cannot properly do it justice here. What I do want to focus on are four fundamental processes that Tilly argues create and sustain durable inequality. The first two processes, which he refers to as causal mechanisms, are exploitation and opportunity hoarding; the second two, which he calls supporting mechanisms, are emulation and adaptation. Exploitation ‘operates when powerful, connected people command resources from which they

10 C. Tilly, Durable Inequality, Berkeley, 1998, p. 6. 11 C. Tilley, op. cit. not. 10, p. 4.

m ai ze , mo u nd s, and co smo s

draw significantly increased returns by coordinating the effort of outsiders whom they exclude from the full value added by that effort.’12 Let me emphasize that Tilly’s distinction between powerful, connected people whom we might think of as insiders and the outsiders excluded from the full value of their effort (or labor) combines what I refer to as segregation and what Eric Wolf refers to as structural power, ‘the power to deploy and allocate social labor.’ I suggest, in fact, that exploitation is the specific manifestation of structural power that finances class, which in turn is legitimized through segregation. Where class pertains, structural power also unfolds as opportunity hoarding — Tilly’s second causal mechanism — which he defines as operating ‘when members of a categorically bounded network acquire access to a resource that is valuable, renewable, subject to monopoly, supportive of network activities, and enhanced by the network’s modus operandi.’13 Opportunity hoarding differs from exploitation in that it is usually practiced by non-elites who lack the ability to appropriate either labor or surplus from others, but who seek rather to secure the benefits of a particular kind of work to the members of their own group. Rather than using the labor of outsiders to perform this work, as in exploitation, they exclude outsiders on the basis of categorical distinctions, creating beliefs and practices that sustain their control over the resource and its relevant practices. Unsurprisingly, perhaps, Tilly observes that exploitation and opportunity hoarding are often complementary, if not mutually reinforcing. Tilly’s supporting mechanisms, emulation and adaptation, are relatively straightforward. Emulation is ‘the copying of established organizational models and/or the transplanting of existing social relations from one setting to another,’ while adaptation is ‘the elaboration of daily routines such as mutual aid, political influence, courtship, and information gathering on the basis of categorically unequal structures.’14 That is, adaptation reproduces and even reinforces a system of exploitation and class inequality.15 Sociologist Michael Mann published a salient critique of Durable Inequality shortly after its publication, noting that Tilly’s causal mechanisms are each more mechanism than cause.16 While Tilly combines these concepts to give us a far-reaching theory of how durable inequality works, he does little to explain how such inequality arises. Indeed, Tilly himself, in considering exploitation, begs off the problem of causation: ‘Let me vault over the crevasse of a fascinating, important question: how by force, ruse, purchase, inheritance, or legal device groups of actors acquire control over valuable resources in the first place.’17 Of course, this is a question that has captivated archaeologists for decades now — what are the origins of durable inequality? If Tilly’s theory helps to explain how it 12 13 14 15 16

C. Tilley, op. cit. not. 10, p. 10. C. Tilley, op. cit. not. 10, p. 10. C. Tilley, op. cit. not. 10, p. 10. C. Tilley, op. cit. not. 10, p. 98. M. Mann, ‘The history of all previous society is the history of durable dichotomies’, in Contemporary Sociology, 28(1), 1999, pp. 29–30. 17 C. Tilley, op. cit. not. 10, p. 87.

61

62

ro bin  a . bec k

operates, then how do we explain why it occurs at all? Douglas Price and Gary Feinman pose the problem this way: ‘The important question for archaeologists considering the emergence of inequality becomes not when and how status differentiation emerged, but rather when and how it became formalized and institutionalized in society. By institutionalized, we mean that these differences become inherited and socially reproduced.’18 As I noted earlier, I am unconvinced that the issue is a shift from achieved to inherited differentiation; nor do I think — as Price and Feinman do — that ‘institutionalized’ and ‘socially reproduced’ are synonymous with ‘inherited.’ So instead, I pose the question this way: What historical conditions enable a categorically defined group of actors, through an exercise of structural power, to transform segregation into class? Just recently, Tim Earle and Matthew Spriggs have offered an approach that I believe may address this question.19 But before introducing it, let me break from the abstract and theoretical side of things and turn to my case study, Mississippian Cahokia. After briefly discussing this greatest of all North American archaeological sites, I return to the problem of those historical conditions that enable the creation of class.

Mississippian Class Origins at Ancient Cahokia Mississippian was the last great moundbuilding tradition of the Eastern Woodlands. The term itself — whether in its current form or in its earlier Middle Mississippi iteration — has a long history in American archaeology, beginning with W. H. Holmes’ monumental analyses of native ceramics during the last two decades of the nineteenth century.20 It has persisted through a series of shifts in archaeological method and theory, from the Midwestern Taxonomic System of the 1920s and 30s21 through the culture-historical perspectives of the Post-WWII era22 and the ecological-adaptionist scenarios of the 1970s23 (Smith 1978). Today, few North American archaeologists would propose a single way of defining Mississippian, nor rely on a set list of material

18 D. Price and G. Feinman, ‘Foundations of prehistoric social inequality’, in T. D. Price and G. M. Feinman eds, Foundations of social inequality, New York, 1995, p. 2. 19 T. K. Earle and M. Spriggs, ‘Political economy in prehistory: A Marxist approach to Pacific Sequences’, in Current Anthropology, 56(4), 2015, pp. 515–44. 20 W. H. Holmes, ‘Ancient Pottery of the Mississippi Valley’, in C. Thomas ed., 4th Annual Report of the Bureau of Ethnology, Washington, D. C., 1886, pp. 361–436. W. H. Holmes, ‘Aboriginal Pottery of the Eastern United States’, in C. Thomas ed., 20th Annual Report of the Bureau of Ethnology Washington, D. C., 1903, pp. 1–201. 21 T. Deuel, ‘Appendix I: The Application of a Classificatory Method to Mississippi Valley Archaeology’, in F. C. Cole and T. Deuel, Rediscovering Illinois, Chicago, 1937, pp. 207–19. 22 J. B. Griffin, ‘Culture periods in Eastern United States archaeology’, in J. B. Griffin ed., Archaeology of Eastern United States, University of Chicago Press, Chicago, 1952, pp. 352–64. G. R. Willey, An introduction to American archaeology, Vol. 1: North and Middle America, Englewood, 1966. G. Willey and P. Phillips, Method and theory in American archaeology, Chicago, 1958. 23 B. Smith, ‘Variation in Mississippian settlement patterns’, in B. Smith ed., Mississippian settlement patterns, New York, 1978, pp. 479–503.

m ai ze , mo u nd s, and co smo s

traits (Thorne Deuel once proposed no fewer than 16 such distinguishing characteristics). Most of us, rather, recognize that the concept subsumes an enormous range of cultural, linguistic, geographic, and political variability. Yet the term must surely refer to something more than just a block of time; it is hardly sufficient for each of us to know Mississippian when we see it. Here, I return to the idea that Mississippian was the last great moundbuilding tradition in the Eastern Woodlands, a tradition that stretches back more than 5000 years. What makes Mississippian different from earlier earth-moving traditions is that its political economy was based on an agricultural lifeway introduced from Mesoamerica by way of the American Southwest. And maize — the product of that lifeway — carried with it the potential for radical and durable transformations in social and political structures. Mississippian, then, lies at the intersection of mounds and maize. The piling of earth into mounds has a deep history in eastern North America. The earliest known mounds are those at Watson Brake in Louisiana, begun approximately 3500 bc. Watson Brake, indeed, is the oldest dated architectural complex of its kind in the Americas, and its eleven mounds surrounding an oval plaza would prefigure the mound and plaza arrangement of Mississippian mound groups more than 4000 years later.24 The Lower Mississippi Valley and adjacent areas remained the center of moundbuilding practices through the end of the Archaic Period, culminating with the construction of Poverty Point (1750–1350 bc), located east of Watson Brake in north-eastern Louisiana. After a lull of several centuries, moundbuilding flourished again during the Early and Middle Woodland periods, though its center of gravity had shifted north to the Ohio Valley, where Hopewell (200 bc–ad 400) peoples created some of the most extensive geometric earthworks in the world; the largest of these, at Newark, once covered more than 1200 ha. Although Midwestern moundbuilding declined with Hopewell after ad 400, the Late Woodland Troyville (ad 400–700) and Coles Creek (ad 700–1200) cultures, located in the Lower Mississippi Valley not far from both Watson Brake and Poverty Point, would continue the moundbuilding tradition. Significantly, the Coles Creek peoples would introduce a new kind of mound, a truncated pyramid that served as a foundation for wood buildings, which they arranged around open plazas. This platform mound and plaza complex later became the prototype of Mississippian ceremonial spaces.25 The development of Mississippian society is inextricably tied to the growing significance of maize agriculture throughout most of the Eastern Woodlands during the period from ad 500–1000, which archaeologists identify as the Late Woodland. Maize (Zea mays) was not native to this region, having first been domesticated in the highlands of southern Mexico by c. 6000 bc. Small amounts of maize have been recovered from Ohio Hopewell sites that date to the first two centuries ad, indicating 24 J. W. Saunders, R. D. Mandel, C. Garth Sampson, C. M. Allen, E. T. Allen, D. A. Bush, J. K. Feathers, K. J. Gremillion, C. T. Hallmark, H. E. Jackson, J. K. Johnson, R. Jones, R. T. Saucier, G. L. Stringer, and M. F. Vidrine, ‘Watson Brake, a Middle Archaic mound complex in Northeast Louisiana’, in American Antiquity 70(4), 2005, pp. 631–68. 25 V. Steponaitis, M. C. Kassabaum, and J. W. O’Hear, ‘Coles Creek Antecedents’, in T. Pauketat and S. Alt eds, The Medieval Mississippians, Santa Fe, 2015, pp. 13–19.

63

64

ro bin  a . bec k

that use of this plant has a long history in eastern North America (Riley et al. 1994).26 During Hopewell times, however, its consumption was likely restricted to special or ritual events, such that it never made an important contribution to Hopewell diet. Rather, Middle Woodland peoples such as the Hopewell and their contemporaries joined a hunting and gathering subsistence with the cultivation of certain seed crops that were native to the Eastern Woodlands, particularly maygrass, sunflower, and sumpweed. As native farmers continued to experiment with maize during the Late Woodland period, they recognized that it had clear benefits over those traditional seed crops. First, maize provided higher-yield harvests than any native cultigens and had a much higher caloric value. This was no small matter, for the Late Woodland was a period of population growth in much of eastern North America, and as the size of local populations grew, that of their hunting and gathering territories shrank. By focusing more attention on maize, Late Woodland groups were able to intensify food production in smaller catchment zones, even though the clearing and maintenance of maize fields required a higher investment of labour. Just as importantly, it is far easier to harvest and store maize than any of the native seed crops so vital to previous generations. Nimble fingers were needed to harvest the latter, but younger children and the elderly could help to pick maize given the relative ease of removing cobs from their stalks. This shift to maize agriculture thus offered substantial benefits, up to a point, for having access to larger groups and their potential labour. Finally, since maize is storable for long periods of time, people had greater incentives to produce surpluses. As George Milner observes, however, we need not presume that there was significant top-down management of farming regimes; rather, ‘the impetus to grow maize bubbled up from below; that is, from decisions made by individual households’.27 This link between labour and surplus was a key part of the cultural context in which larger polities began to develop across the Midwest and South after ad 1000. As competition for the best agricultural lands intensified, so too did competition for the labour needed to clear, plant, and harvest such lands and defend them from others. The leaders of those families and communities with claim to particularly desirable locations actively sought to draw new people to their groups, as more people produced larger surpluses and offered better defence. Also, drawing new people could create debt relationships, as families new to a particular location found that the better soils for farming had already been claimed by others. Many newcomers must have found themselves periodically dependent on their more affluent neighbours in times of want. They could then repay any debts by working in their benefactors’ fields or by giving up a portion of their own surplus in times of plenty. These relations reached beyond the community to the wider region, as those communities located in rich locations could attract more people than those in less desirable places. Larger communities

26 T. J. Riley, G. R. Walz, C. J. Bareis, A. C. Fortier, and K. E. Parker, ‘Accelerator Mass Spectrometry (AMS) dates confirm early Zea mays in the Mississippi River Valley’, in American Antiquity 59(3), 2004, pp. 490–98. 27 G. Milner, The moundbuilders: Ancient peoples of eastern North America, London, 2004, p. 123.

m ai ze , mo u nd s, and co smo s

enjoyed an obvious military advantage over smaller neighbours and also had more labour and surplus to apply toward political activities such as making mounds and hosting feasts. Maize may not have caused Mississippian, but maize agriculture did the create conditions that made Mississippian possible.

Early Mississippian at Cahokia Unprecedented opportunities for polity building began in the American Bottom region of the American Midwest once such maize-related changes to the Late Woodland political economy were underway by ad 900. Here, as would unfold in other parts of the Eastern Woodlands over the centuries to come, well-positioned communities began to grow rapidly at the expense of their neighbours. The particular socio-religious form that this process took in the Midcontinent, for which we now use the term Mississippian, was thus in a temporal position to influence other portions of the Eastern Woodlands.28 In the American Bottom, Kelly has suggested that an initial Emergent Mississippian pattern of small communities accelerated to one of the nucleated towns some decades prior to ad 1000.29 Cahokia itself may have grown to house more than 2000 people,30 while farther south in the Bottom, small, dispersed communities aggregated into larger towns such as the Pulcher and Washausen sites.31 Platform mound construction also likely began at Cahokia, Lohmann, Pulcher, East St Louis, and Washausen. There is now an archaeological consensus that the time right around ad 1050 is when a diverse set of Southeastern and Midwestern practices and ideas were brought together at Cahokia to create a way of life we refer to as Early Mississippian (fig. 2). In less than fifty years, the mound and plaza arrangement, the use of shell-tempered pottery, and the shift to wall trench architecture would become recognizable as a cultural entity — Mississippian — across the American Midcontinent and Southeast. While Cahokia is the foremost of these early sites in the American Bottom region, two others — East St Louis and St Louis — became giants themselves: East St Louis and St Louis are, respectively, the second and fourth largest of all Mississippian sites.

28 D. Anderson, ‘The role of Cahokia in the evolution of southeastern Mississippian society’, in T. R. Pauketat and T. E. Emerson eds, Cahokia: Domination and ideology in the Mississippian world, Lincoln, 1997, pp. 248–68. V. J. Knight, ‘Some developmental parallels between Cahokia and Moundville’, in T. R. Pauketat and T. E. Emerson eds, Cahokia: Domination and ideology in the Mississippian world, Lincoln, 1997, pp. 229–47. T. R. Pauketat and T. E. Emerson, ‘Introduction: Domination and ideology in the Mississippian world’, in T. R. Pauketat and T. E. Emerson eds, Cahokia: Domination and ideology in the Mississippian world, Lincoln, 1997, pp. 1–29. 29 J. E. Kelly, ‘The Pulcher tradition and the ritualization of Cahokia: A perspective from Cahokia’s southern neighbor’, in Southeastern Archaeology 21(2), 2002, pp. 136–48. 30 T. R. Pauketat and N. H. Lopinot, ‘Cahokian population dynamics’, in T. R. Pauketat and T. E. Emerson eds, Cahokia: Domination and ideology in the Mississippian world, Lincoln, 1997, p. 115. 31 C. R. Barrier and T. J. Horsely, ‘Shifting communities: Demographic profiles of early village population growth and decline in the central American Bottomm’, in American Antiquity, (79), 2014, pp. 295–313. J. E. Kelly, op. cit. not. 29.

65

66

ro bin  a . bec k

Fig. 2. Cahokia’s location in the Eastern Woodlands.

Cahokia transformed from a large community of perhaps 2000 people to a massive center of as many as 10,000.32 It became an attractor, drawing people from a diverse range of places across the Midcontinent.33 Much of this attraction seems to

32 T. R. Pauketat and N. H. Lopinot, op. cit. not. 30. 33 G. R. Milner, The Cahokia chiefdom: The archaeology of a Mississippian society, Washington, D. C., 1998. T. R. Pauketat, The ascent of chiefs, Tuscaloosa, 2004. T. R. Pauketat, An archaeology of the cosmos: Rethinking agency and religion in ancient America, Routledge, New York, 2013. P. A. Slater, K. M. Hedman, and T. E. Emerson, ‘mmigrants at the Mississippian polity of Cahokia: Strontium isotope evidence for population movement’ in Journal of Archaeological Science 44, 2014, pp. 117–27.

m ai ze , mo u nd s, and co smo s

Fig. 3. Plan view of Cahokia, ad 1000–1250.

have been the center’s emphasis on public works, including Monks Mound and the Grand Plaza (fig. 3).34 In the Mississippian world, the activities of mound and plaza construction were tied to the forging of communal or group identities. These were inclusive practices aimed at creating a sense of community and common purpose among the diverse peoples rapidly converging on this place. Elsewhere, I have referred to these as group-building activities.35 What is so interesting about Early Mississippian organization at Cahokia is the lack of evidence for group-distancing or exclusive activities and practices. That is, across a range of analytical measures (including the distribution of shell beads and other exotic materials, house size, and feasting debris), there is remarkably little evidence of durable inequality or class as

34 R. A. Dalan, ‘The construction of Mississippian Cahokia’, in T. R. Pauketat and T. E. Emerson eds, Cahokia: Domination and ideology in the Mississippian world, Lincoln, 1997, pp. 89–102. R. A. Dalan, G. R. Holley, W. I. Woods, H. W. Watters, Jr. and J. A. Koepke, Envisioning Cahokia: A landscape perspective, DeKalb, 2003. T. Schilling, ‘The chronology of Monks Mound’, in Southeastern Archaeology 32(1), 2013, pp. 14–28. 35 R. A. Beck, ‘Consolidation and hierarchy: Chiefdom variability in the Mississippian southeast’, in American Antiquity 68(4), 2003, pp. 641–61. R. A. Beck, ‘Persuasive politics and domination at Cahokia and Moundville’, in B. M. Butler and P. D. Welch eds, Leadership and polity in Mississippian society, Carbondale, 2006, pp. 19–42.

67

68

ro bin  a . bec k

I define it at the time of Cahokia’s demographic and architectural explosion, from ad 1050 to ad 1200. Cahokia was a religious center, and the mounds that dominated its landscape were a key part of the Early Mississippian religious experience. Vernon Knight and other scholars have long argued that mounds were earth symbols, and that the act of mound construction reenacted the creation of the earth; moreover, mound surfaces were continually decommissioned or ritually killed with a new surface built atop the old, a combination of acts linked to death and rebirth, fertility and regeneration.36 Throughout the Early Mississippian, Cahokia’s religious practice also centered on two especially prominent culture heroes, Earth or Corn Mother and Morning Star, sometimes referred to as Birdman. Corn Mother is usually associated with small redstone statuary. Most have been recovered from so-called rural temples on the outskirts of the main centers, and most were smashed.37 The best-known example is the Birger Figurine. Guy Prentice has argued that it and all others of the same class are an expression of the Earth-Mother concept — the goddess of life and death, creator of people and plants. With her help the ‘Earth-Serpent,’ symbol of death and the underworld, provides the agricultural crops that humans need. On her back the goddess carries the pack or sacred bundle that is the symbol of fertility goddesses. The pack may be symbolic of rain, fertility, the source of knowledge, and the power of deities. It may contain human bones or the souls of men and women in the process of being returned to the land of the living. It may also contain the first seeds of the plants given by the fertility goddess to humankind.38 Morning Star, on the other hand, was linked to what James Brown refers to as the male progenitive principal, as well as with warfare and the conquest of life over death.39 He was closely associated with the game known during historical times as chunkey, when it was a popular sport of high stakes gambling among native peoples throughout the Eastern Woodlands. During the Mississippian period, though, the stakes may have been even higher, as depictions of chunkey in Mississippian iconography identify it with a narrative of sacrifice, life, and death. The hero of the narrative ventures into the underworld where he challenges its lords to a game of chunkey. These are the same lords who had defeated and beheaded his ancestors at some earlier point in time. The hero defeats his underworld enemies, takes their heads in victory, and returns to our middle world, where he restores his ancestors’ heads to their proper place of veneration. In doing so, he is apotheosized as Morning Star and takes his place in the heavens of the upper world. What I want to emphasize is that — throughout the 36 V. J. Knight, ‘The institutional organization of Mississippian religion’, in American Antiquity 51, 1986, pp. 675–87. 37 T. E. Emerson, Cahokia and the archaeology of power, Tuscaloosa, 1997. 38 G. Prentice, ‘An analysis of the symbolism expressed by the Birger Figurine’, in American Antiquity 51(2), 1986, p. 262. 39 J. A. Brown, ‘The Cahokia Mound 72-Sub1 burials as collective representation’, in The Wisconsin Archaeologist 84 (1&2), 2003, pp. 83–99.

m ai ze , mo u nd s, and co smo s

Early Mississippian — it seems as though the veneration of both culture heroes was a communal affair and part of the religious synthesis that brought so many people to Cahokia in the first place.

Transformation at Cahokia Then, about ad 1200, Cahokia witnessed a significant decrease in its resident population; in mound and plaza construction; and in the importation of exotic materials. As people moved away from Cahokia and as the rate of new mound construction decreased, we see the first evidence of a massive palisade within the ceremonial precinct. The Central Palisade enclosed Monks Mound and the Grand Plaza and was rebuilt three times over the next century. Milner argues that these palisades demanded more labor organization than all of the earth-moving activities at Cahokia, including Monks Mound, combined.40 While this wall could certainly have satisfied defensive needs, it also acted as a barrier to what we may presume were among the town’s most sacred zones — zones that were, until this act, communal and open to public view. It is significant that people never aggregated behind these walls, but rather moved away from Cahokia in large numbers. Cahokia’s palisade did not serve to draw people inside, but to create a cosmological distance between those within and those without. At about the same time that people at Cahokia built their first palisade, they also built a new mound and plaza group just to the east of the Grand Plaza. This new complex, the Ramey Plaza, seems to have been restricted to elite households and craft production areas. Mary Beth Trubitt’s study of household data suggests that ‘differentiation of household units into smaller/poorer and larger/wealthier is more pronounced in the later phases of the Mississippian period’.41 Comparing the density of non-local items at Cahokia with the density of such materials at outlying sites, she found that there was ‘more status or wealth differentiation between households and between sites after ad 1200’.42 What is more, as larger households began to withhold a significantly greater share of exotic materials, they also began to centralize the production and exchange of craft objects. The decrease in shell bead production after ad 1200 coincided with a new focus on one-of-a-kind objects like engraved shell cups and gorgets and repoussé copper plates. Large households at Cahokia controlled their production, and most were likely produced in the Ramey Plaza.43 Both the production and distribution of these symbolically-loaded objects was quite different from that of shell beads during Early Mississippian times. While the latter were mass produced and meant 40 G. R. Milner, op. cit. not. 33. 41 M. B. D. Trubitt, ‘Mound building and prestige goods exchange: Changing strategies in the Cahokia chiefdom’, in American Antiquity 65(4), 2000, p. 673. 42 M. B. D. Trubitt, op. cit. not. 41, p. 674. 43 J. E. Kelly, J. A. Brown, and L. S. Kelly, ‘The context of religion at Cahokia: The Mound 34 case’, in L. Fogelin ed., Religion in the material world, Carbondale, 2008, pp. 297–318.

69

70

ro bin  a . bec k

for general consumption, the former had highly restricted access, as shown by their concentration in elite burial contexts at Middle Mississippian centers such as Nashville, Etowah, Lake Jackson, Moundville, Lake George, and Spiro. Indeed, well-connected households at some of these centers, too, began to support their own, local schools of craft specialization, producing iconographically rich objects in copper and shell that also entered exchange networks among regional elites. In sum, we can see the emergence of class or durable inequality across the Mississippian world by ad 1200, probably beginning at Cahokia and spreading out from there. Let’s look at this through the lens of Tilly’s four mechanisms. We see exploitation — the co-opting of outsiders’ labor to ends that exclude them from the full benefits of that labor — in the construction of the Central Palisades that closed off Monks Mound and the Grand Plaza, as well as in the construction of a new mound-plaza group, the Ramey Plaza. These projects enabled the physical and spiritual segregation of Tilly’s insiders from the outsiders who performed much of the work. Moreover, it seems that those who sponsored these projects also appropriated a cosmological narrative that had once been part of the communal ideology — the Morning Star or Birdman myth cycle — and transformed it into an origin narrative that legitimized segregation. We also find opportunity hoarding in the emergence of craft specialists who produced those one-of-a-kind objects in copper and shell — objects that gave material shape to the appropriation of communal cosmology. Jon Muller’s experiments with replicating engraved shell indicate that artists may have needed many hundreds of hours to produce some of these objects — using stone tools, he found that ‘continuous cutting with chert flakes in an already established groove took hours per millimeter of depth’.44 These specialists, whom I would compare with the members of craft guilds, may have controlled the exchange networks needed to get raw shells to Cahokia and the process of training new acolytes. Together, these constitute an opportunity hoarding that reinforced exploitation. Emulation, in turn, is apparent in the spread of these objects to other emerging centers, as well as in the founding there of comparable schools of craft specialists. Finally, adaptation is suggested in the expansion of — and increased dependence on — maize agriculture throughout the Mississippian world, as local communities produced the agricultural surpluses needed to finance class inequalities. How, then, did this happen? And why did it happen at Cahokia at or about ad 1200? Let me return to that fundamental question I posed a bit earlier: What historical conditions enabled a categorically defined group of actors, through an exercise of structural power, to transform segregation into class? I noted that Earle and Spriggs have recently offered an important way of thinking about the problem.45 They suggest that we focus on bottlenecks, which they refer to as ‘constriction points in commodity chains that offer an aspiring leader the opportunity to limit access, thus

44 J. Muller, Mississippian political economy, New York, 1997, p. 16. 45 T. K. Earle and M. Spriggs, op. cit. not. 19.

m ai ze , mo u nd s, and co smo s

creating ownership over resources, technologies, or knowledge’.46 They ‘stress that the development of central power and social stratification was grounded in conditions that created bottlenecks, allowing hierarchical property relations to emerge’.47 Not coincidentally, I believe that we can identify such a bottleneck in the American Bottom right around ad 1200. Recent work by geomorphologist Samuel Munoz and his colleagues shows that Cahokia rose to prominence during a time when the Central Mississippi River experienced no major flooding events.48 Starting about ad 580, the American Bottom enjoyed a long period — nearly six centuries — of relatively stable riverine development. Then at ad 1200, give or take a couple of decades, a massive spring flood bore down on Cahokia and its neighbors, with water rising to a likely height of 10 m and depositing as much as 20 cm of sediment in some areas. For much of Cahokia’s hinterland, these floodwaters would have soon retreated, and while the event itself would have undoubtedly shaken its people, they might quickly have resumed their former lives along the river but for one calamitous outcome. As James Brown points out, this deluge of water — borne south from the upper Missouri River and its tributaries — must have spread millions upon millions of cottonwood seeds across those flooded fields so carefully cleared and tended by Cahokia’s farmers.49 Clearing the fields of seeds would have been a hopeless task, such that in a very short time — just a handful of years — many of these fields would have been overtaken by stands of juvenile cottonwood, making farming in those places impractical if not impossible. Some people simply moved away and did not return. Those who remained would have focused their labor on reclearing the fields least impacted by the flood and its cottonwood cargo. For the families who owned or controlled such fields (and the maize produced there), opportunities arose for changing the social relations of labor and its rewards. For the first time, land rather than labor had become the scarce resource at Cahokia. And in taking advantage of such opportunities, I propose that these well-positioned families transformed segregation into class.

Conclusions Let me stress that I am not advocating a simplistic return to environmental determinism, to a claim that environmental factors cause inequality. On the contrary, bottlenecks are social phenomena that may be conditioned by environmental as well as social factors. The Mississippi River flood of ad 1200 changed the ecology of the American Bottom floodplain. For centuries, this floodplain had supported a large population of farmers who built a major religious center at the heart of its landscape. This intersection 46 T. K. Earle and M. Spriggs, op. cit. not. 19, p. 517. 47 T. K. Earle and M. Spriggs, op. cit. not. 19, p. 517. 48 S. E. Munoz, K. E. Gruley, A. Massie, D. A. Fike, S. Schroeder, and J. W. Williams, ‘Cahokia’s emergence and decline coincided with shifts of flood frequency on the Mississippi River’, in Proceedings of the National Academy of Science 112(20), 2016, p. 6319–6324. 49 J. A. Brown, personal communication to author, 2016.

71

72

ro bin  a . bec k

of social and ecological factors created challenges that stimulated people to change their relations of labor and production. Yet the flip side of challenge is opportunity, and among these changes was a forging of durable or class inequality. As Earle and Spriggs observe, however, people often find ways to subvert the appropriations of power that emerge from bottlenecks, and indeed, such was the case with Cahokia. Within fifty years of its transformation from a place of communal practice to one of exclusion, it was all but abandoned, its people voting with their feet to reject these new relations. What I advocate, in sum, is that we conceive of durable inequality or class as a historical, contingent phenomenon. It is never inevitable, nor does it necessarily coincide with surplus, agriculture, aggregation, or any other so-called prime mover. Indeed, people can and do reject it even while maintaining surplus, agriculture, and aggregation. In every case, inequality is something we must explain historically, not on the basis of ahistorical, universalizing types.

2

Economic Specialization, Elite Demand and Social Inequality

Sauro Gelichi

Pottery as Inequality? Systems of Production and Distribution in North Italian Societies During the Early Middle Ages

Introduction As archaeologists know, pottery is a oversized category of artifacts in the archaeological record. This is bad, because we can overestimate its role in the interpretation of social and economic processes. At the same time, however, always being present, it lends itself better to comparison than other objects. It is right that we consider how best to use it. In recent years, the approach to studies into ceramics in general, and especially into medieval pottery, has shifted from the historical/artistic or antiquarian aspects to those where more stress has been placed on chronological, distributive and functional/social evidence: what is conventionally called ‘the big three’.1 In particular, studies into the distribution of pottery have, as is well known, proved very useful in describing and shedding light on trade around the Mediterranean in Classical and Late Antiquity,2 although some believe this is not the case for the Middle Ages.3 Nevertheless, many Italian studies were conducted in this regard in the 1980s and 1990s and these have attracted the attention of some historians, including Paolo Delogu and Chris Wickham, who appreciate the value of archeology, using this data in different ways and to varying degrees.4 More specifically, Wickham was especially concerned with reflecting on the production of ceramics in Late Antiquity and the Early Middle



1 C. Orton, P. Tyers and A. Vince, Pottery in archaeology, Cambridge, 1993, pp. 23–30. 2 See, for instance: C. Panella, ‘Per lo studio dei contesti e delle merci tardo antiche’, in A. Giardina ed., Società Romana e Impero Tardoantico, III, Merci, Insediamenti, Bari, 1986, pp. 21–449; C. Panella, ‘Merci e scambi nel Mediterraneo in età tardoantica’, in Storia di Roma, III, 2, Torino, 1993, pp. 251–72. 3 P. Davey and R. Hodges, ‘Ceramics and trade: a critique of the archaeological evidence’, in P. Davey and R. Hodges eds, Ceramics & Trade. The production and distribution of late medieval pottery in NorthWest Europe, Sheffield, 1983, pp. 1–14. 4 See, for instance: C. Wickham, ‘Marx, Sherlock Holmes and the Late Roman Commerce’, in Journal of Roman Studies, 78, 1988, pp. 183–93; P. Delogu, Le origini del medioevo. Studi sul settimo secolo, Roma, 2010, where much use is also made of ceramic indicators. Sauro Gelichi  •  University Ca’ Foscari, Venice Social Inequality in Early Medieval Europe: Local Societies and Beyond, ed. by Juan Antonio Quirós Castillo, Turnhout : Brepols, 2020 (HAMA 39), pp. 75–97 © FHG10.1484/M.HAMA-EB.5.118446

76

s auro ge l ich i

Ages, drawing some sophisticated socio-economic conclusions:5 pottery is, after all, a commodity that was easily imitated and mass produced in pre-industrial societies.6 The purely procedural approach has been generally criticized by post-processual archeology applied to pottery from the 1990s onwards. As a result of justified dissatisfaction with the analytical/descriptive method for classification purposes, these practices have since been abandoned and distribution studies are now somewhat distrusted, without, at least in Italy, leading to an alternative, innovative interpretation of ceramics in the archaeological record. Moreover, later types of pottery are more suitable for this type of analysis or, put another way, more research and encouraging results have been obtained using more recent types of pottery, perhaps because these better lend themselves to comparison against the written sources.7 The current situation, therefore, is that studies of post Antiquity ceramics is no longer as central to research as it was just a couple of decades ago. This is a good thing, from a certain point of view, as the abundance of pottery in the archaeological record had created a kind of ‘monothematic’ approach that did not assist general historical studies or those focusing more specifically on ceramics. The introduction of the concept of ‘context’ led to a shift in focus to relationships and their meaning, thereby slowly taking pottery out of its ‘splendid isolation’ and returning it to within its more pertinent social space. At the same time, however, it has contributed to a loss of valuable specialist skills and has introduced quicker analysis procedures that have often become approximate, with the risk that many researchers are now able to explain relationships between various artifacts, but do not know — or know very little — about the objects they are called upon to analyze. It would be inappropriate and completely impractical to propose that taxonomic studies become once again central in the theoretical and methodological approach to research. Having said that, however, a call for more stringent philology and a more critical consideration of artifacts is not unreasonable, especially as this would help corroborate these more interesting and useful approaches. I would here like to consider a specific problem, picking up and developing a subject of importance to me over the years. It concerns a clearly defined chronological period within the Middle Ages (eighth–tenth century) and an equally well defined geographic area (North Italy). I would like to do this by drawing on some coherent, sufficiently broad knowledge and using approaches that combine different methods and tools of analysis, while at the same time using the ceramics indicator in its more specific social function.



5 More specifically, in C. Wickham, Framing the Early Middle Ages. Europe and the Mediterranean, 400–880, Oxford, 2005, in general on system of exchange pp. 693–824 and on Italy pp. 728–59. See also various comments on this book (including mine) in Storica, 34, pp. 121–72. 6 C. Wickham, ‘Overview: production, distribution and demand’, in R. Hodges and W. Bowden eds, The Sixth Century. Production, Distribution and Demand, Leiden, 1998, p. 285. 7 See, merely by way of example, the recent book by M. Ferri, C. Moine and L. Sabbionesi eds, In & Around. Ceramiche e comunità. Secondo Convegno tematico dell’AIECM3 (Faenza 2015), Firenze, 2016.

pot t e ry as i neq uali t y ?

Pottery in North Italy Studies on pottery from the early Middle Ages (eighth–tenth centuries) in northern Italy point to a strong apparent homogeneity in ceramics: in terms of their typology, function and form. Over time, this homogeneity has led to a move away from exclusively taxonomic approaches.8 As a result, pottery has been viewed as being not a particularly sensitive indicator of social variability between rural and urban communities, within the same topographical context (urban or rural areas) or even within similar social groups. But is this really so? Should we refrain from using ceramics as a possible social marker? How can pottery reflect social diversity, if indeed it can? In the past, archaeologists relied on a fairly well-defined taxonomy of pottery production in northern Italy.9 The situation outlined in the 1980s has not changed much, with just a few modifications (chronological and typological). Societies, whether urban or rural, used a small number of functional and technological types of pottery (fig. 1). This is true in the eighth to tenth centuries, whereas, in the seventh century, the distribution of ARS and Eastern Sigillata shows that some social groups still used this type of artifact as tableware. Some countries in the Mediterranean area continued to produce ceramic tableware (albeit in combination with other forms of tableware).10 This means that the decline in the production of ARS and Eastern



8 These must not, however, be seen as impediments or obstacles in more sophisticated approaches to the social aspects, as, for instance, certain attempts to interpret even very simple ceramics (such as Anglo Saxon pottery) in terms of what they can tell us about society: P. Blinkhorn, ‘Habitus, social identity and Anglo-Saxon pottery’, in C. G. Cumberpatch and P. W. Blinkhorn eds, Not so much a pot, more a way of life. Current approaches to artefact analysis in archaeology, Oxford, 1997, pp. 113–24. Others are attempting to interpret ceramics from other areas of Europe in a social key, including pottery found in the north of the Iberian Peninsula showing signs of technological and formal simplification: A. Vigil-Escalera Guidaro and J. A. Quiros Castillo eds, La cerámica de la Alta Edad Media en el Cuadrante Noroeste de la Península Ibérica (siglos V–X). Sistemas de producción, mecanismos de distribución y patrones de consumo, Bilbao, 2016. 9 On a more general note concerning pottery from the North of Italy, see G. P. Brogiolo and S. Gelichi, ‘La ceramica grezza medievale nella pianura padana’, in La ceramica medievale nel Mediterraneo Occidentale (Siena- Faenza 1984), Firenze, 1986, pp. 293–316; G. P. Brogiolo and S. Gelichi, ‘Ceramiche, tecnologia e organizzazione della produzione nell’Italia settentrionale tra VI e X secolo’, in G. Démians D’Archimbaud ed., La céramique médiévale en Méditerranée. Actes du VIe Congrès de l’AIECM2 (Aix-en-Provence 1995), Aix-en-Provence, 1997, pp. 139–45; G. P. Brogiolo and S. Gelichi, ‘La ceramica comune in Italia settentrionale tra IV e VII secolo’, in L. Saguì ed., La ceramica in Italia: VI–VII secolo, Atti del Convegno in onore di J. W. Hayes (Roma 1995), Firenze, 1998, pp. 209–26; S. Gelichi and F. Sbarra, ‘La tavola di San Gerardo. Ceramica tra X e XI secolo nel nord Italia: importazioni e produzioni locali’, in Rivista di Archeologia, XXVII, 2003, pp. 119–41. 10 There are no good studies on glazed Mediterranean ceramics from the eighth-tenth centuries. Even regional or sub-regional studies are compromised due to the doubtful reliability of the archeological documentation. Of importance here is the correct identification of the age of the artifacts and a precise indication of their origin. The techniques used have often been poorly studied, understood and described. In any case, regarding early lusterware in the Mediterranean area, see R. B. Mason 2004, Shine like the Sun: Lustre-painted and Associated Pottery from the Medieval Middle East, Costa Mesa, 2004. Useful information can still be gleaned from the classic book on Egyptian ceramics by

77

78

s auro ge l ich i

Fig. 1. The main types of pottery kilns in the North Italy during the Early Middle ages (from Gelichi and Sbarra, ‘La Tavola’, fig. 3).

H. Philon, Benaki Museum, Athens. Early Islamic Ceramics. Nine to Twelfth Century, London, 1980 and from the article by G. Scanlon, ‘Fustat Fatimid Sgraffito: Less than Lustre’, in M. Barrucand ed., L’Egypte fatimide, son art et son histoire, Paris, 1999, pp. 265–83. A better source — also because it is based on stratigraphic associations at the Fustat excavation — is the recent book on Egyptian ceramics in the 9th and 10th centuries by R. P. Gayraud and L. Vallaury, Fustat. II. Fouilles d’Istabl ‘Antar. Céramiques d’ensembles des ixe et xe siècle, Le Caire, 2017; for Egyptian ‘Fayyumi Ware’, see G. Scanlon ‘Fayyumi Pottery: a Long-lived Misnomer in Egyptian Islamic Ceramics. Type I’, in Société Archéologique d’Alexandrie 35, 1993, pp. 295–330 and G. Williams, ‘Fayyumi’ Ware: Variations, Imitations, and Importations of an Early Islamic Glazed Ceramic Type. Unpublished Thesis American University of Cairo, 2013 (http://dar.aucegypt.edu/bitstream/handle/10526/3388/Thesis%20 Fayyumi.pdf?sequence = 1). For Byzantine ‘Glazed White Ware’ see J. Vroom, Byzantine to Modern Pottery. An Introduction Field Guide, Utrecht, 2005, pp. 72–79. For Byzantine ceramics imported into Italy, see E. D’Amico, ‘Byzantine Finewares in Italy (tenth to fourteenth centuries ad): Social and Economic Contexts in the Mediterranean World’, in S. Gelichi ed., Atti del IX Congresso Internazionale sulla Ceramica Medievale nel Mediterraneo (Venezia, 2009), Firenze, 2012, pp. 473–79. For oriental glazed ceramics and their relations with Syrian production, see M. Tite, O. Watson, T. Pradell, M. Matin, G. Molina, K. Domoney and A. Bouquillon, ‘Revisiting the beginnings of tin-opacified glazes’, in Journal of Archaeological Science 57, 2015, pp. 80–91 and O. Watson, ‘Revisiting Samarra: the Rise of Islamic Glazed Pottery’, in Beiträge zur Islamischen Kunst und Archäologie 4, 2010, pp. 124–32.

pot t e ry as i neq uali t y ?

Fig. 2. “Glazed White Ware” (Polycrhome type) from Nonantola (archaeological excavations into the Abbey).

Sigillata should not be seen as the underlying reason for a change in behavior in Italic societies (especially in North Italy). We are now certain that the revival in imports from the Mediterranean did not take place before the eleventh or, more probably, the twelfth century.11 Some sherds of pottery dating from before this, would have been from exotic products and the result of occasional exchanges or gifts: e.g. “Glazed White Ware” (Polychrome type) from Nonantola (fig. 2) should be seen in the same light as samite, again from Nonantola;12 likewise, the Iranian pottery of the same period found in San Vincenzo al Volturno and in Fulda.13 In terms of technology, north Italian potters used the following methods and materials: the fast potter’s wheel (rarely, the slow wheel); clays with the addition of inclusions; vessels usually fired in reducing conditions; rarely any coating (slip or glaze); surface treatment sometimes absent. Occasionally they added incisions as decorative features, obtained with pointed instruments or created through impression; many other marks on the surface are most probably linked to the technology used (surface treatment). Very few products made from fine or partially fine clays and fired in oxidizing conditions have been discovered. Glazed pottery was produced,

11 Updated information on these matters can be found in S. Gelichi, ‘Islamic pottery in the neighbourhood of the Venetian lagoon. A contribution on the relationships between Venice and the Eastern Mediterranean during the 11th-12th century’ in T. Nowakiewicz, M. Trzeciecki, D. Błaszczyk, eds, Animos labor nutrit. Studia ofiarowane Profesorowi Andrzejowi Buko w siedemdziesiątą rocznicę urodzin, Warszawa 2018, pp. 115–128. 12 For Samite tissue, see P. Peri, ‘Antiche reliquie tessili dell’Abbazia di Nonantola (secoli VIII–XII)’, in R. Fangarezzi, P. Golinelli and A. M. Orselli eds, Sant’Anselmo di Nonantola e i santi fondatori nella tradizione monastica tra Oriente e Occidente, Roma, 2006, pp. 239–59. 13 For the fragments from San Vincenzo al Volturno, see H. Patterson, ‘The Pottery’, in J. Mitchell, I. L. Hansen and C. M. Couts eds, San Vincenzo al Volturno 3: the Finds from the 1980–86 Excavations, Spoleto, 2001, pp. 319–20, Fig. 10.134, plates 101–02. Thanks to Thomas Kind for informations on the example from Fulda.

79

80

s auro ge l ich i

but only in some specific areas in the north.14 Generally speaking, this means that the production structures must have been very simple, including open fire kilns, indicating very little investment in technology. The number of kilns excavated in north Italy is still very limited. Despite their small number, these kilns allow us to make some interesting observations.15 Concerning the technology, all the known production systems are modest in size and relatively simple; complexes with several kilns and other service facilities are rarely documented. Their archaeological visibility, especially in the case of open fire kilns,16 however, is modest, leaving almost no traces, as they are very similar to those of simple hearths.17 With regard to the location, however, kilns near or inside Roman villas or rural settlements have been discovered but mainly date back to the beginning of the early Middle Ages. Their presence has been interpreted as the terminal phase of the widely documented use of such structures in Ancient Roman times; or, better still, as the exploitation of ancient abandoned structures.18 This phenomenon

14 For early Medieval glazed ceramics in northern Italy, see Gelichi and Sbarra, ‘La tavola di San Gerardo’ and (more recently) S. Gelichi, ‘Nuove invetriate alto-medievali dalla laguna di Venezia e di Comacchio’, in S. Lusuardi Siena, C. Perassi, F. Sacchi and M. Sannazaro eds, Archeologia classica e post-classica tra Italia e Mediterraneo. Scritti in ricordo di Maria Pia Rossignani, Milano, 2016, pp. 297–317. 15 Cf. The two kilns excavated in the Capitolium area of Brescia: see A. Guglielmetti, ‘La ceramica comune fra fine VI e X secolo a Brescia nei siti di casa Pallaveri, palazzo Martinengo Cesaresco e piazza Labus’, in G. P. Brogiolo and S. Gelichi eds, La ceramica altomedievale (fine VI–X secolo) in Italia settentrionale: produzione commerci (Monte Barro — Galbiate 1995), Mantova, 1996, pp. 9–14; A. Guglielmetti, ‘Ceramica di età longobarda nell’area del Capitolium: analisi di una struttura produttiva’, in F. Rossi ed., Carta Archeologica della Lombardia. V. Brescia, la città, Modena, 1996, pp. 265–83; A. Guglielmetti, ‘Il vasellame in ceramica di età altomedievale’, in F. Rossi ed., Un luogo per gli dei. L’area del Capitolium a Brescia, Firenze, 2014, pp. 454–56, Fig. 8 (left). One of these two kilns has been dated to roughly 592 (± 160) and was used to produce glazed ‘Longobard’ ceramics and most common pottery. Another, smaller kiln has been found and excavated (also in Brescia) close to the Basilica di Santa Maria; thermoluminescence dating puts this at 610± 110 or 659 ± 150 (two specimens): Guglielmetti, ‘Il vasellame’, p. 456, note 90, Fig. 8 (right). Another kiln (ninth or tenth century) has been excavated in Libarna (AL): C. Davite and F. Filippi ‘Un forno altomedievale per la cottura della ceramica a Libarna’, in E. Giannichedda ed., Archeologia della produzione. Antichi mestieri, Genova, 1996, pp. 74–76; F. Filippi, G. Gaij and G. Pantò, ‘La produzione di una fornace altomedievale per ceramica da Libarba (AL)’, in G. Pantò ed., Produzione e circolazione dei materiali ceramici in Italia settentrionale tra VI e X secolo (Torino 2002), Mantova, 2004, pp. 57–83. 16 T. Mannoni and E. Giannichedda, Archeologia della produzione, Torino, 1996, p. 173, Fig. 33.1. 17 It is thought that the existence of an open fire kiln in the area of Carlino (UD), in a site close to Aquileia, where glazed waste kiln products have been found, can be dated to the late fourth century / early fifth century (C. Magrini and F. Sbarra, Le ceramiche invetriate di Carlino. Nuovo contributo allo studio di una produzione tardoantica, Firenze, 2005), but there are no clear traces of a kiln: see C. Magrini and F. Sbarra, ‘Nuovi dati per un vecchio ritrovamento: le ceramiche grezze e invetriate del ‘focolare’ nel sito produttivo di Carlino (UD)’ in G. Pantò ed., Produzione e circolazione dei materiali ceramici in Italia settentrionale tra VI e X secolo (Torino 2002), Mantova, 2004, pp. 247–63. 18 The kiln excavated in Libarna (AL) does not provide clear evidence in this sense: the kiln was found in a Roman cemetery, but does not appear to have any association with this, partly due to the time lapse. This area is not very far from the ancient Roman city, but it would appear to be a coincidence: Filippi, Gaij and Pantò, ‘La produzione di una fornace’ pp. 57–61 and 64 (chronology), Fig. 3 and 6.

pot t e ry as i neq uali t y ?

might be analogous with a similar documented phenomenon regarding glass kilns from the same period.19 The second aspect to note is that, contrary to what one might expect, only a small number of workshops belonging to the same period have been unearthed in urban locations (or close to a major town). The urban kilns of Brescia would appear to lie within a fiscal space (a ducal court) where people were engaged in various artisanal activities (archaeologists have found traces of bone and metal working).20 In a broader geographical context (Italy), the third consideration is that most productive structures are either found within villages (or rural areas) or in the vicinity of religious buildings (churches and monasteries) and hence are linked to the exploitation of the properties of the local soil and rock. In some places, one can also assume that these activities were under the direct control of the landowners (San Genesio, near San Miniato, in Tuscany) or the church hierarchy (as in the case of Mola di Monte Gelato in the Roman countryside).21 Subsequent studies have slightly modified this reconstruction in two senses: the first concerns the identification of some previously unknown productions using fine, well prepared clays fired in oxidizing conditions; the other regards the production of lead glazed pottery. The first category appears in the context of the eighth century. The second perhaps in the context of the ninth and tenth centuries. These ceramics require more complex kilns, a more sophisticated treatment of the clay and, finally, the use of additional materials, like color and glaze: in other words, greater investment in technology and costs. We will consider this in greater depth later on.

19 M. Ferri, ‘Consumo e produzione nella laguna in età tardo-antica: i vetri’, in S. Gelichi et al., ‘Importare, produrre e consumare nella laguna di Venezia dal IV al XII secolo. Anfore, vetri e ceramiche’, in S. Gelichi and C. Negrelli eds, Adriatico altomedievale (VI–XI secolo). Scambi, porti, produzioni, Venezia, 2017, pp. 52–57. 20 The kiln discovered near the Basilica di Santa Maria, again in Brescia, is, however, considered as being ‘dependent or controlled by ecclesiastical institutions’: Guglielmetti, ‘Il vasellame’, p. 456. 21 For San Genesio (San Miniato — Pisa), see below, but also: F. Cantini, ‘Ceramiche dai siti medievali rurali della Toscana (VIII–X secolo): una prima sintesi’, in S. Gelichi ed., Campagne medievali. Strutture materiali, economia e società nell’insediamento rurale dell’Italia settentrionale (VIII–X secolo) (Nonantola — San Giovanni in Persiceto 2003), Mantova, 2005, pp. 258–76; F. Cantini, ‘Forme, dimensioni e logiche della produzione nel medioevo: tendenze generali per l’Italia centrale tra V e XV secolo’, in A. Molinari, R. Santangeli Valenzani and L. Spera eds, L’archeologia della produzione a Roma (secoli V–XV). Atti del Convegno Internazionale di Studi (Roma 2014), Roma- Bari, 2015, p. 504. Regarding Motta di Monte Gelato, see F. Marazzi, T. Potter and A. King, ‘Mola di Monte Gelato (Mazzano Romano — VT): notizie preliminari sulle campagne di scavo 1986–1988 e considerazioni sulle origini dell’incastellamento in Etruria meridionale alla luce dei nuovi dati archeologici’, in Archeologia Medievale, XVI, 1989, p. 108, Fig. 4; T. W. Potter and A. C. King, Excavations at the Mola di Monte Gelato. A Roman and Medieval Settlement in South Etruria, Rome, 1997, pp. 92–94, Figg. 87–88. Also for Libarna (AL), see above; albeit in a rural location, we are unaware of any relationships with the kiln owners.

81

82

s auro ge l ich i

The Mode of Production Some time ago, I tried to conceptualize the modes of production system in Italy during the Middle Ages.22 An ambitious project, I do not deny it; a project that had some strengths (the prospect of a comparative and synchronic reading of the processes), but also several negative points (the weakness of the individual cases, the abstractness of the models). Then again, this is a weakness that is inherent in all attempts to understand and contain the processes in a single system. At the time, a point of reference was Peacock’s book on production in the Ancient Roman World.23 Peacock had used an ethno-anthropological approach and, it was obvious, these recognized models were specifically transferred to the Roman age. In the same manner, we could usefully transfer the same models to other historical pre-capitalistic societies.24 We basically identified three main models for the medieval age in North Italy. The use of these models goes beyond the generic definitions often used by researchers, such as ‘home production’, ‘industrial products’, ‘handcrafted products’ and so on. Model 1 (Household Production)

This is what is conventionally called ‘home production’. It has long been recognized as the basic model for the production of ceramics in northern Italy, but also, for certain periods and for certain areas, also in central and southern Italy. It would be characterized by a low level of investment in technology and a low volume of output. It would be a very simple system where the producers are family members (or, at most, members of the village community: especially women or part-time farmers). The same people would be the consumers. Despite this being the model has long been postulated for the socio-economic reality of most of the Italian peninsula, there is no direct evidence for its existence relying primarily on such comparisons according to the ethno-anthropological approach. No specific case of this kind has been recognized archaeologically. We could perhaps attribute to this type of model some productions of the tenth–eleventh centuries in southern Tuscany or the Ligurian Apennines (the production of testelli).25 22 S. Gelichi, ‘Gestione e significato sociale della produzione, della circolazione e dei consumi della ceramica nell’Italia dell’Alto Medioevo’, in G. P. Brogiolo and A. Chavarria Arnau eds, Archeologia e società tra Tardo Antico e Alto medioevo (Padova 2005), Mantova, 2007, pp. 47–69. 23 D. P. S. Peacock, Pottery in the Roman World: an Ethnoarchaeological approach, New York, 1983. Before Peacock, van der Leeuw had developed a similar structure of models: S. van der Leeuw, Studies in the Technology of Ancient Pottery, Amsterdam, 1976. 24 The Peacock model has, obviously, been used by other archeologists and historians in other geographical areas, such as Cora di Tudmir in Spain (S. Gutiérrez, La Cora di Tudmīr. De la Antigüedad Tardía al Mundo Islámico. Poblamiento y Cultura Material, Madrid — Alicante, 1996), southern Tuscany in Italy (F. Grassi, La ceramica e le vie del commercio tra VIII e XIV secolo. Il caso della Toscana meridionale, Oxford, 2010, pp. 20–24) and, more generally, throughout the Mediterranean (Wickham, Framing, pp. 704–05). 25 T. Mannoni ‘Il testo e la sua diffusione nella Liguria di levante’, in Bullettino Ligustico per la Storia e la Cultura Regionale, 17, 1965, pp. 49–64; Mannoni and Giannichedda, Archeologia, pp. 300–01.

pot t e ry as i neq uali t y ?

Model 2 (Household Industry)

This type of production is very similar to the previous kind (low investment in technology, artifacts very simple and functional). The difference lies, however, in the volume of production and the distribution facilities; finally, often (but not always) there is centralized production in a single area. There are some aspects worth pointing out here. The first concerns the craftsmen: again, in this case, they are part-time specialists, also engaged in other activities both within and outside the domestic sphere (farmers and breeders). Investment in technology is modest. At the same time, there is a concentration of production in the same place, which could only be partly due to special circumstances regarding the availability of the raw material. The concentration, therefore, can be better explained by looking at the topographical location of the site (in relation to routes of communication) and at its functions, likely to encourage trade. Many of the workshops known in northern Italy (and Italy in general) could fall into this second category. The existence of this model could explain the movement of pottery within the village community, as part of seasonal markets in the context of rural areas, but also in urban markets.26 The acceptance of this model, obviously, presupposes the existence of some form of exchange, a surplus in ceramic production and also, but not always, of middlemen for its distribution. There must, therefore, have been (more or less consistent) demand for this kind of product. This model may reflect, at least in part, the so called ‘curtense system’ in northern Italy: seen as the possible result of a centralized activity managed directly by the dominicus or, potentially, specialization built up within rural communities.27 The production in the vicinity of ecclesiastical

26 With reservations, the kiln area excavated in Libarna (AL) — already discussed — and dated to the ninth or tenth century could be associated with this category. Unfortunately, we know very little about the context of the area in which this structure lies — only the excavated section of the kiln still exist, all other data concerning the stages of its activity have been lost. Nevertheless, technological studies into the products — including kiln waste — and the remains of the kiln itself have led to a few interesting observations on the shape of the kiln, the temperatures and product standardization. Based on observations of the professional skills of the artisans, it has even been possible to estimate output: ‘The type of production is qualitatively and technologically high, but not yet industrial. The organization of kiln activities points to output capacity of between 800 and 1500 pots each month, probably sold within a relatively limited local radius of perhaps 50–70 km from the site of production’ (Filippi, Gaij and Pantò, ‘La produzione di una fornace’ p. 64). The social and economic framework within which the structure operated is entirely unknown. 27 This problem is not new. The fact that ceramic production, like other crafts, was the expression of a ‘curtense industry’ had already been suggested by Monneret de Villard: U. Monneret de Villard, ‘L’organizzazione industriale nell’Italia longobarda durante l’Alto Medioevo’, in Archivio Storico Lombardo, V, XLVI, 1919, p. 20. This theory was partly criticized by Cinzio Violante and, more recently, by Tourbert: C. Violante, La società milanese in età precomunale, Bari, 1953; P. Toubert, ‘Il sistema curtense: la produzione e lo scambio interno in Italia nei secoli VIII, IX e X,’, in G. Sergi ed., Curtis e signoria rurale: interferenze fra due strutture medievali, Torino, 1993, pp. 25–94 (re-issue of an essay dated 1983). I have already discussed these matters, suggesting this possibility, in S. Gelichi, ‘Ceramiche senza rivestimento grezze’, in S. Gelichi and N. Giordani eds, Il tesoro nel pozzo. Pozzi deposito e tesaurizzazioni nell’antica Emilia, Modena, 1994, pp. 88–95.

83

84

s auro ge l ich i

complexes may also reflect such a reality. A parallel could be the production system managed by the Church in the Roman countryside (domuscultae).28 Model 3 (Individual Workshop)

In this model, there is far greater investment in technology (needed for the decoration of the pieces and the use of lead glaze). The kilns are also more complex. Circulation is obviously greater and must be through mediation. These products were, therefore, traded in rural and town markets. Pottery decorated with red-slip (in central and southern Italy), that made from fine clay fired in oxidizing kilns (also in the north) and, finally and most especially, lead glazed pottery could all fall within this category.29 It is possible, but by no means certain, that the production, in this case, was also managed by feudal landowners (see below) or monasteries (there is, however, only indirect evidence of this at present). A few years ago, Paul Arthur and Helen Patterson suggested that the foundation of domuscultae, in the Roman countryside, might have created the necessary conditions for the emergence of “independent professional potters”. It is also possible that, early on, the Church played a role in the production and distribution of ceramics, possibly with the presence of “attached specialists”.30 The social category of potters, however, becomes important to distinguish this model from the other two models. In fact, one has to assume that the craftsmen were fulltime potters, although their existence is, at present, only indirectly documented by written sources from the tenth century only. For example, Beatrice Annis attributes the first production of lead glazed pottery in Rome to this model.31 On the basis of mineral-petrographic analysis, one can reasonably assume that workshops were located in both towns and the countryside. The production of similar pottery (lead glazed pottery), in a part of northern Italy between the ninth and eleventh centuries 28 One such case could be that of Mola di Monte Gelato (Potter and King, Excavations) which we have already mentioned, especially if it can be established that the complex belonged to the Papal State. 29 The artisan activities at Serratone near Rispescia (GR) might belong to this category, although only kiln waste has been found during surveys and magnetometry analysis. This would, in any case, only concern good quality ceramics, produced on a fast potter’s and sold down the Ombrone and Osa valleys: E. Vaccaro, Sites and Pots: Settlement and Economy in Southern Tuscany (ad 300–900), Oxford, 2011, pp. 202–15; E. Vaccaro, ‘Ceramic Production and Trade in Tuscany (3rd–mid 9th c. ad): new Evidence from the South-West’, in E. Cirelli, F. Diosono and H. Patterson eds, Le forme della crisi. Produzioni ceramiche e commerci nell’Italia centrale tra Romani e Longobardi (III–VIII sec. d. C.) (Spoleto- Campelio sul Clitumno 2012), Bologna, 2015, p. 222. 30 P. Arthur and H. Patterson, ‘Ceramics and early medieval central and southern Italy: ‘a potted history’’, in R. Francovich and G. Noyé eds, La storia dell’alto medioevo italiano (V–X secolo) alla luce dell’archeologia (Siena 1992), Firenze, 1994, p. 433. 31 B. Annis, ‘Ceramica altomedievale a vetrina pesante e ceramica medievale a vetrina sparsa dallo scavo di San Sisto Vecchio in Roma: analisi tecnologica e proposta interpretativa’, in L. Paroli ed., La ceramica invetriata tardo antica in Italia. (Certosa di Pontignano, Siena 1990), Firenze, 1992, p. 412. See also a few contributions published in a recent book on manufacturing activity in Rome during the Early Middle Ages: A. Molinari, R. Santangeli Valenzani and L. Spera eds, L’archeologia della produzione a Roma (secoli V–XV). Atti del Convegno Internazionale di Studi (Roma 2014), Roma- Bari, 2015.

pot t e ry as i neq uali t y ?

Fig. 3. Lead Glazed Pottery from Venice (nn. 1-2) and from Comacchio (nn. 3-7) (ninth century) (from Gelichi, ‘Nuove invetriate’, fig. 1).

could be the output of similar workshops. Even the pottery from the Comacchio area could be the result of a similar activity (see below). These models, however, presuppose a permanence of production, although they do not indicate the duration. We could add a further possibility, which perhaps explains the production of single-fired lead glazed pottery in northern Italy. The existence of this production is testified by archaeological findings and archaeometric data. There are two main groups. The first group (ninth century?) consists of vessels very similar to Roman ‘Forum Ware’ (fig. 3). The second group (tenth and eleventh centuries), with no applied decoration and few traces of glazing, is similar to ‘Roman Sparse Glazed’32 (fig. 4). Minero-petrographical analysis proves that these were not imported from the

32 This type of ceramics has been known (and studied) for some time: S. Gelichi and M. G. Maioli, ‘La ceramica invetriata tardo antica e alto-medievale dall’Emilia Romagna’, in L. Paroli ed., La ceramica invetriata tardo antica in Italia. (Certosa di Pontignano, Siena 1990), Firenze, 1992, pp. 215–78; Gelichi and Sbarra, ‘La tavola di San Gerardo’ and, more recently, Gelichi, ‘Nuove invetriate’, containing a summary of and an update on the matter.

85

86

s auro ge l ich i

Fig. 4. Lead Glazed Pottery from Ravenna (tenth-eleventh century) from Gelichi, Maioli, ‘La ceramica’, fig. 19).

Central Italy and were are probably produced locally. Very few examples of the first type exist, almost all concentrated in the Comacchio and Ravennna areas, and Venice. Many more examples have been found of the second type, the vessels being distributed across a far wider area. The existence of the first group can be explained by the concept of intermittent production: since specific technical skills are required, different from

pot t e ry as i neq uali t y ?

those needed for standard production in northern Italy, one must assume that qualified workers from central and southern Italy travelled north at different times in history.

From the Model to Social Relationship of Production The adoption of these conceptually functional models, however, poses some problems. The greatest problem lies in the fact that the differences between these models are not always easily recognizable at an archaeological level, especially when faced with contexts like most of those already analyzed: poorly documented and often topographically circumscribed. Moreover, it is not always easy to associate the model and the archaeological data with the political-social structure of reference. Put simply, it is not easy to translate an anthropological model into a historically interpreted framework. For example, the birth of centralized sites in southern Tuscany during the Middle Ages33 is accompanied by the emergence of “small production sites located in the vicinity of each site” and serving the needs of these new realities. However, those who have supported this approach are unsure whether it is best for model 2 (household industry) or model 3 (individual workshop).34 Furthermore, and always regarding southern Tuscany, a couple of sites — including Roccastrada — have been interpreted as belonging to a higher level (a nucleated workshop), where production was characterized by the use of a fast potter’s wheel, stable equipment, vitreous glaze and, especially, a far wider radius of distribution.35 But even in this case, the social dimension of production is still difficult to establish, both as regards the typology of the artisans (part-time, attached specialists?) and the economic context. One example that might help us translate the material data into a social context of production is the site of San Genesio (PI), Tuscany. Archaeological studies over many years have uncovered the existence of a production center inside the settlement itself that manufactured good quality ceramics, fired in an oxidizing environment and sometimes decorated with red slip. Work is still ongoing and archaeologists have dated these kilns to the ninth century using archeomagnetism.36 This activity would, most probably, have been conducted within a single feudal area: the lands owned

33 Regarding the question of the concentration of dwellings in medieval Tuscany and its evolution over the years, it is still worth a classic text: R. Francovich and R. Hodges, Villa to village. The Transformation of the Roman Countryside in Italy c. 400–1000, London, 2003. 34 Grassi, La ceramica, p. 22. 35 Grassi, La ceramica p. 23. In reality, the difference between a household industry and a nucleated workshop lies not so much in the complexity and sophistication of the productive structure, but rather in the social organization of the location, especially when one considers the number of workshops to be found in the same place. 36 For the excavations at San Genesio, see note 21 and F. Cantini, ‘Per un’archeologia dei vici tra Tarda Antichità e Medioevo. Il caso di Vicus Wallari-Burgus Sancti Genesi ’, in P. Galetti ed., Paesaggi, comunità, villaggi medievali (Bologna 2010), Centro Italiano di Studi sull’Alto Medioevo, Spoleto, 2012, pp. 511–23. For specific details of the kiln and its products, see F. Cantini, ‘Forme, dimensioni’, p. 505. The kiln here produced pots decorated with red slip and was a large vertical kiln (2.50 × 1.50 m). This kiln was capable of producing anything up to 200 items per firing. Thanks to Federico Cantini for these information.

87

88

s auro ge l ich i

by the Marquis of Tuscia Adalberto ‘il ricco’.37 What is more, the proximity of this site and an olive press lead us to believe that the kiln also produced vessels for the oil obtained from the manorial lord’s lands (the curtes). The relationship between taxable property (money exacted by the king, a marquis or a count) and productive activities in Tuscia between the ninth and tenth centuries is now becoming much clearer, thanks also to recent archaeological evidence.38 The different types of ownership (a fiscal curtes or a private curtes) would influence the way the settlements were structured and whether or not they would include specialized production activities. Furthermore, certain productive activities can be differentiated within a fiscal curtes. In the specific case of ceramics, there is another site of interest: as well as San Genesio, Donoratico (LI) was a center of glazed ceramic production in the ninth and tenth centuries.39 However, the pottery produced here would appear to have enjoyed limited distribution (just within the surrounding area), suggesting that the workshops served specific, perhaps even distant, markets. If we were to transfer this archaeological reality to the aforementioned models, the workshops in both Donoratico and, perhaps, San Genesio, could have been individual workshops (or nucleated workshops). Nevertheless, the examples we have mentioned qualify in a specific way, to the extent that we can place them in a specific area of social and political relationships. If we now return to our original models, we must recognize that most case-histories concern not only the parameters defining each model (as Peacock has already pointed out),40 but also the differences in what are called “production relations”. Even with such reservations, I believe that all these models are still useful, at least as tools. Although they may appear to be schematic (from the simplest to the most complex) and may not be used to explain the way production is organized and has evolved, they still allow us to describe the functional aspects and material structures

37 F. Cantini, ‘La gestione della produzione fra curtes fiscali e curtes private in età carolingia’, in G. Bianchi, C. La Rocca and T. Lazzari eds, Spazio pubblico e spazio privato tra storia e archeologia (secoli VI–X) (Bologna 2014), Turnhout, 2018, pp. 261–291. 38 G. Bianchi, ‘Recenti ricerche nelle colline metallifere ed alcune riflessioni sul modello toscano’, in Archeologia Medievale, XLII, 2015, p. 18. For a specific look at this problem and updates, see the forthcoming report by G. Bianchi and S. Collavini, ‘Beni fiscali e strategie economiche nell’Alto medioevo toscano: verso una nuova lettura?’, at the workshop entitled Origins of new economic union (7th-12th centuries), Resources, landscapes and political strategies in Mediterranean Region, Siena, 12 April 2017. Firenze 2018, pp. 223–291. 39 Material still under analysis. The results of the first digs have been published in G. Bianchi ed., Castello di Donoratico. I risultati delle prime campagne di scavo (2000–2002), Firenze, 2004. Specific references to ceramics (albeit preliminary) can be found in Grassi, La ceramica, pp. 114–17. Local production of these ceramics is likely on the basis of archeological evidence and archeometric analysis, but no kiln has, as yet, been found. Thanks to Giovanna Bianchi for this information. More recently A. Briano and E. Sibilia, ‘Progetto nEU-Med. Nuove analisi archeologiche e archeometriche sulla ceramica a vetrina sparsa dal castello di Donoratico (LI): i risultati della Termoluminescenza (LI)’, in Archeologia Medievale , XLV, 2018, pp. 357–365. 40 Peacock, Pottery, p. 8: ‘However, it must be remembered that we are attempting to impose a conceptual framework upon a situation that in practice may be almost infinitely variable, with many examples falling between rather than within the modes here defined, but it is only when the rules have been made that the exceptions can be recognized’.

pot t e ry as i neq uali t y ?

of a given productive reality. However, as has been rightly observed and as we have already noted, these models basically define the concept of mode of production in technological and behavioral terms, while the social aspects are merely implied. The modes of production should, however, be redefined as “a configuration of relationships which link the social and economic sphere”.41 I therefore find it extremely useful to introduce at this point some other parameters, within these conceptual categories, that let us better define the modes of production. In other words, those elements that make up each mode of production and which, according to Andrews, would be the following: Raw material; Labour; Technology; Output. These components are key to all modes of production because, without any one of these, production simply cannot take place. Certain variables will also affect these components, such as the division and intensity of work, or the time needed to obtain and process the raw material. As a result, our models can also be re-formulated and re-conceptualized42 and once again become applicable.

Consumption: Use and Distinction Some time ago I noticed that the quality of ceramic production was worse in the north of the Italian peninsula compared to that from central and southern Italy. Once again, this finding seemed unrelated to any relationship with the city’s elite, since the phenomenon of urbanism appeared to be stronger in the north than in the south or center of Italy.43 I deduced, therefore, that there was not always a direct relation between pottery consumption and the urban elite, either on account of the different wealth of the elite in the north, center and south, or because they had different attitudes to such objects. There is a rather different situation in Tuscany, where the quality of ceramics produced along the Arno valley is far is better than in the south of this region.44 Here, one should note that ‘the city’ survived quite well in the north (Lucca, Pisa and Florence), but not so in the south of the region; it could, therefore, be argued that this difference can be associated with the wealth of the city’s elite, for whom these products were intended. However, this approach (processualistic in theory) is not truly satisfactory. Specific realities are more complex. Other tools and other conceptual means are needed in order to analyze this phenomenon better and so avoid a very general but unsatisfactory interpretation. Coarse ware, at the moment, is not subject to this kind of analysis: in fact, the situation depends on the way this kind of ceramics is studied. Archaeologists have, until now, adopted traditional methods and theories. They generally describe the form of the vessel, the color of the fabric. Chronology is their main topic. Some years ago 41 See K. Andrews, ‘From ceramic finishes to modes of production: Iron Age finewares from central France’, in C. G. Cumberpatch and P. W. Blinkhorn eds, Not so much a pot, more a way of life. Current approaches to artefact analysis in archaeology, Oxford, 1997, pp. 60–61. 42 Andrews, ‘From ceramic’, pp. 61–62, Table 1 and 2. 43 S. Gelichi, ‘Ceramic production and distribution in the early medieval Mediterranean basin (seventh to tenth centuries ad): between town and countryside’, in G. P. Brogiolo, N. Gauthier and N. Christie eds, Towns and their Territories between Late Antiquity and the Early Middle Ages, TRW 9, Leiden, pp. 115–39. 44 F. Cantini, ‘Ceramiche’, pp. 258–76.

89

90

s auro ge l ich i

Tiziano Mannoni in Italy, and more recently Alan Vince in England, have independently demonstrated the value of adopting some other specific archaeological approaches, such as the analysis, for example, of easily visible characteristics, like the form of base type, the decoration or the surface treatment (combined with detailed analysis of the material components).45 This could lead, at a micro local level, to the presence of different pottery traditions (there have recently been some good applications: for example, in a recent issue of the ‘Medieval Archaeology’ journal regarding Anglo-Saxon pottery in Torksey, Lincolnshire, by Gareth Perry).46 But no area of Italy (north or south) has been subject to such indepth analysis involving the same number of examples (whether places of production or places of consumption).47 So, our ability to understand the levels of the production and consumption of ceramics on a local scale in order to highlight typology and diversity in the social consumption is not yet possible. However, I would like to attempt social consumption analysis using the data currently available to us. In line with the guiding idea of this publication, here are a few reflections on two phenomena typically found in northern Italy. The first phenomenon concerns the production of ceramics without lead glaze, but with fine, well prepared clay, usually fired in oxidizing conditions. Its presence was recently noted into two types of context linked to different forms of settlement: the first are specialized centers that we could call ‘emporia’ (like Venice and Comacchio); the second are ancient cities (like Ravenna, Rimini and Cesena). These offer two different perspectives of interpretation. Since the pottery found in Comacchio (and maybe even in Venice) was not imported from outside (central and southern Italy, the Mediterranean), it must be inferred that it may have been produced locally.48 Minero-petrographic analysis does not exclude

45 A. Vince, ‘Forms, functions, and manufacturing techniques of late 9th- and 10th century wheel thrown pottery in England and their origins’, in D. Piton ed., Travaux de Groupe de Recherches et D’ètudes sur la céramique dans le Nord-Pas-De-Calais, Groupe de Recherches et d’études sur la céramique dans le Nord/Pas-de-Calais, Saint-Josse-sur-Mer, 1993, pp. 151–64. A. Vince, ‘Ceramic petrology and the study of Anglo-Saxon and later medieval ceramics’, in Medieval Archaeology, 49, 2005, pp. 219–45. See Mannoni, for his ground-breaking research into clays and traces of working: T. Mannoni, ‘Analisi mineralogiche e tecnologiche delle ceramiche medievali. Nota II’, in Atti del V Convegno Internazionale della Ceramica, Albisola, 1972, pp. 107–28; T. Mannoni, La ceramica medievale a Genova e nella Liguria, Bordighera, 1975; M. G. Magi, T. Mannoni, ‘Analisi mineralogiche di ceramiche mediterranee. Nota IV’, in Atti dell’VIII Convegno Internazionale della Ceramica, Albisola, 1975, pp. 155–66. 46 G. Perry, ‘Pottery Production in Anglo-Scandinavian Torksey (Lincolnshire): Reconstructing and Contextualising the Chaîme Opératoire’, in Medieval Archaeology, 60, 2016, pp. 72–114. 47 The site excavated in Loc. Crocetta, Possessione Canale, near Sant’Agata Bolognese (BO) could be an exception here. This settlement (tenth–eleventh century) produced a considerable amount of ceramics, with some theories as to its consumption: F. Sbarra, ‘La ceramica di un villaggio di X secolo nell’area padana: produzione e circolazione’, in R. Curina and C. Negrelli eds, 1° Incontro di Studio sulle Ceramiche Tardoantiche e Alto medievali Cer.Am.Is (Manerba 1998), Mantova, 2002, pp. 95–95-124, Tav. 11. The weakness of this interpretation, however, lies in the empirical division of the material, in turn based on the distance of diffusion of given types. 48 Regarding this category of products, see C. Negrelli, ‘Produzione, circolazione e consumo tra VI e IX secolo: dal territorio del Padovetere a Comacchio’, in F. Berti, M. Bollini, S. Gelichi and J. Ortalli eds, Genti nel Delta da Spina a Comacchio. Uomini, territorio e culto dall’Antichità all’Alto medioevo,

pot t e ry as i neq uali t y ?

Fig. 5. Pottery without glaze with two handles from Comacchio (small local amphora?) (eight–ninth century) (courtesy Claudio Negrelli).

this hypothesis.49 The variety of shape of the pottery — very standardized types and decorations — is extremely limited: only closed forms with a flat bottom and almost always with two handles are currently documented (fig. 5). It is logical to think that these were products with a specific function (perhaps small amphorae). They might also have been used to divide the contents of the globular amphora and transfer them into smaller containers that were perhaps more functional when transporting goods on the lagoons and rivers where a different type of craft was needed50 (fig. 6). Ferrara, 2007, pp. 444–54; C. Negrelli, ‘Towards a definition of early medieval pottery: amphorae and other vessels in the northern Adriatic between the 7th and the 8th centuries’, in S. Gelichi and R. Hodges eds, From one Sea to Another. Trading Places in the European and Mediterranean Early Middle Ages (Comacchio 2009), Turnhout, 2012, pp. 49–413. 49 For a preliminary report, see C. Capelli, ‘6. Analisi archeometriche’, in S. Gelichi et al., ‘I materiali da Comacchio’, in F. Berti, M. Bollini, S. Gelichi and J. Ortalli eds, Genti nel Delta da Spina a Comacchio. Uomini, territorio e culto dall’Antichità all’Alto medioevo, Ferrara, 2007, p. 643. 50 S. Gelichi, ‘Societies at the edge: new Cities in the Adriatic Sea during the Early Middle Ages (8th-9th centuries)’, in S. Gelichi and R. Hodges, New Directions in Early Medieval European Archaeology: Spain and Italy compared. Essays for Riccardo Francovich, Turnhout, 2015, pp. 293–97, Figg. 5–7.

91

92

s auro ge l ich i

Fig. 6. Example of small local amphora (n. 3) from Comacchio in comparison with imported amphorae (eight–ninth century) (nn. 1-2).

The presence of this type of object in contemporary monastic or urban contexts inland51 could support this interpretation. This pottery was, therefore, an element of distinction, but not necessarily one of inequality: its presence is closely related to the functions of the places where it was produced. Indeed, it might represent an element of differentiation, not in terms of the ceramic objects themselves, but their contents. In recent years, however, in some urban contexts during the seventh and eighth centuries (and, in some cases, also the ninth–tenth centuries) pottery made using fine, well-prepared clay would not necessarily survive. We refer, in particular, a series of urban contexts of Ravenna (such as Piazza Traversari, Via Cavour and Piazza Anita Garibaldi)52 (fig. 7). Some significant examples of this type of pottery have, however, been found here: both similar to that coming from Comacchio (i.e., the small jar with two handles) and also other types of closed or open shapes (for example, single-handled bottles and basins) (fig. 8). However, a kiln producing this type of ceramics has recently been discovered in Classe, in the suburbs of Ravenna, on the site of the Petriana Basilica.53 This might point to a readily identifiable consumer, whereas the case of early medieval deposits of accumulation discovered on Via Cavour could be related to a proper commercial workshop. One may suppose that the local economy was vibrant, at least until the eighth century, perhaps indirectly related, through flumisellum Padennae, to 51 For example, ceramics of this type have been found during excavation of the monastery of Nonantola (Modena) from the late eighth century: see anticipated details in Gelichi ‘Societies’, Fig. 7. 52 An anticipation of the results of these excavations with an accurate analysis of the ceramics so far discovered can be found in the recently published C. Guarnieri, G. Montevecchi and C. Negrelli, ‘Ravenna, una città in declino? Contesti altomedievali in ambito urbano’, in S. Gelichi and C. Negrelli eds, Adriatico altomedievale (VI–XI secolo). Scambi, porti, produzioni, Venezia, 2017, pp. 115–58. 53 E. Cirelli, ‘Material culture in Ravenna and its hinterland between the 8th and the 10th century’, in V. West-Harling ed., Three Empires, three Cities: Identity, Material Culture and Legitimacy in Venice, Ravenna and Rome, 750–1000, Turnhout, 2015, pp. 116–19, Figg. 12–13.

pot t e ry as i neq uali t y ?

Fig. 7. Urban context from Ravenna discussed in the paper 1. Via Cavour 60. 2. Piazza Anita Garibaldi. 3. Via Guaccimanni, via G. Boccaccio, largo Firenze. 4. Via P. Traversari (from Guarnieri, Montevecchi, Negrelli, ‘Ravenna una città in declino’?, fig. 1).

the River Po, one of the most important infrastructures of the medieval city. In the case of the excavation of Piazza Anita Garibaldi (in the east of the city), a domus — of imperial Roman tradition — continued to exist until the seventh century, when the whole area was converted and, perhaps, a building (a small chapel?) was built near the Plateia Maior. Piazza Anita Garibaldi sat in an important position within the city during Late Antiquity and the early Middle Ages: close the church of San Giovanni Evangelista and the imperial palace. Finally, also the site of Piazza Traversari, in the north west of Ravenna, is of great importance. In this case, a series of houses from the seventh to the ninth century is well preserved. Not particularly large, they are rectangular in shape, divided into various rooms and built in bricks on strong foundations. This context must still be attributed to the new type of medium and high standard buildings that were constructed in many cities in the Carolingian and post-Carolingian eras, especially in central Italy. From a more general economic point of view, Ravenna at this time would have been undergoing a shift from a redistribution center to a center of consumption. These contexts are not only of the urban type, but also characterized by medium and high social content, both in terms of the quality of the residential structures and their topographical position within the urban context. These contexts are significant

93

94

s auro ge l ich i

Fig. 8. Pottery without glaze from Ravenna, via Traversari (courtesy C. Negrelli).

not only for the presence of these ceramics, but also for the overall composition of household products: special imports (the globular amphorae with their valuable content) and items of particular value, such as some glazed or lead glazed pottery. A relatively similar situation also seems to be emerging from other areas in the Po Valley: particularly the cases of Rimini (the excavation of Piazza Ferrari) and, perhaps, Cesena (the Garampo hill). In the case of Rimini, the plausible, but by no means certain, reconstruction of the passages of ownership of the urban space between Late Antiquity and the Early Middle Ages — namely, from the aristocracy (sixth century) to monasteries (seventh century) and then back to the nobles (eighth century) — represents reasonably well social levels of consumption and can, therefore, explain the quality of the artifacts and the nature of the associations.54 In the case of Cesena, however, the most likely context is that of military use — the materials come from an environment up against fortified walls built during Late Antiquity — and so

54 C. Negrelli, Rimini Capitale. Strutture insediative, sociali ed economiche tra V e VIII secolo, Firenze, 2008, pp. 48–102.

pot t e ry as i neq uali t y ?

it would be the specific social category that determined the nature of the contexts and, in this case, the associations.55 In these cases, one might, therefore, be faced with a situation whereby pottery (distributed in a different manner within the same context, in this case urban) may constitute an element of distinction. This could mean that the elite used certain goods, but also objects, as a mark of distinction. Within a framework of apparent homogenization, even a glazed jug or simply a fine ceramic bowl might have made a difference. This is true for two main reasons. Firstly, the rarity or difficulty of access of an item would entail a higher cost, thus creating inequality by its very acquisition. Its use may also indicate a different degree of sophistication in daily life, with diverse consequences: both in the type of diet and style of conviviality. But there is another aspect which I think it is interesting to note. Distinction is not created so much (or only) by individual types of pottery, but by the character of the associations of these objects. In essence, it is the type of combinations that characterize the domestic environment and so constitutes a factor of inequality. So, one should analyze the recurrence of these associations, rather than the individual types. For me, the examples discovered in Ravenna seem to point in this direction and so should help orientate future research. Lastly, on a more general level, it should be noted that such associations are currently being documented in both new emerging centers (such as Venice and Comacchio) and in ancient centers with a strong and established aristocracy. It might be tempting to use the old distinction of Lombard/Byzantine areas, but this might prove to be a weak heuristic tool. Yet it is very likely that such an affinity, if any, should be sought in what has been termed ‘Italic Byzantium’: a much more promising concept and one that could be linked to a different form of access to property (at least until the eighth century) and a different system of political relationships, all within the sphere of Italian power in the Carolingian and post-Carolingian ages. However, generalizations are always dangerous, especially during a stage of research that has yet to construct good contexts for comparison. If a city/countryside comparison produces unsatisfactory overall results in terms of the social context that these represent, a city/city comparison is also likely to be disappointing at present, in that we currently only have modest representative urban social nuclei. Furthermore, instead of comparing ‘Lombard’ and ‘Byzantine’ cities, why should we not compare cities along rivers and those along the ancient Roman consular roads? Perhaps such a novel comparison could give rise to new insights into ceramic consumption.

Conclusion: The Value of Pottery Over Time A few words in conclusion. Peter Davey and Richard Hodges published (during the processual era), a table of the values of imported objects56 (fig. 9). But now, in

55 C. Negrelli, F. Bracci and A. A. Rucco, ‘1.2.2 I materiali ceramici dalla tarda Antichità alla prima età altomedievale’, in M. Miari and C. Negrelli eds, Ritmi di transizione. Dal Garampo al Foro Annonario: ricerche archeologiche 2009–2013, Firenze, 2016, pp. 93–130. 56 Davey and Hodges, ‘Ceramics and Trade’, pp. 6–11.

95

96

s auro ge l ich i

Fig. 9. Value of pottery in comparison with other typology of commodities according to a processualistic (from Davey, Hodges, Ceramics and trade, p. 7).

Fig. 10. Value of pottery according to a post-processualist point of view (Gutierrez, Mediterranean, Fig. 1.2).

this post-processual era awaiting new Positivism, we are well aware that the value of pottery can be defined as ‘unstable’57 (fig. 10). Therefore, in this contribution, we have tried to find instances of such instability within apparent homogeneity. We have tested the value of this instability and translated it, when possible, as a social value. It is not easy to measure this instability or translate it, when and where possible, in terms of social disparity: firstly, it presupposes the conceptual reconstruction of the modes of production and the mechanisms of redistribution and exchange, plus an association and comparison with the variability of social contexts. Results of some historical significance are only possible by working within these parameters. Not only must we conceptualize our heuristic approaches better — all too often theoretically weak, if not generic — but also, and above all, must we return to rigorous philological approaches to the contexts, from the analysis of an individual artifact to an understanding of how they were formed. I believe that there is much to do in the future, mainly in two directions. The first concerns the creation of targeted archaeometric projects, without which any

57 A. Gutierrez, Mediterranean pottery in Wessex households (13th to 17th centuries), Oxford, 2000, fig. 1.2.

pot t e ry as i neq uali t y ?

Fig. 11. Example of combinations of objects in some coastal settlements in NorthEastern Italy (eight–ninth century), like Venice, Comacchio, Ravenna: imported amphorae, coarse pottery, local refined pottery, lead glazed pottery, glasses.

interpretation of production and distribution systems will continue to be merely generic. The second regards consumption: here, we should focus more on comparing the associations and various combinations of objects in different contexts, rather than on the presence of any single type of pottery (fig. 11). On a more general note, it is important that any project should be carefully planned and have clearly defined targets.

97

Francesca Grassi

Social Complexity in Peripheral Areas in the Light of Pottery Production Between Sixth and Tenth Centuries (Alava, Basque Country, Spain)

Introduction The study of pottery has led to numerous conceptual models being drawn up regarding the relationship between the craft industry and social complexity. The difference between the many approaches to the issue lies in the differing emphasis given to technological or environmental aspects (the typological or ecological approach), rather than to social or cultural aspects (the social and cultural chaîne operatoire, and behavioural approach).1 The holistic approach put forward in 2007 by Shimada and Wagner, based on several different components, is the one which best reflects the need, which we agree with, to address the study of pottery on a regional, multi-site and diachronic basis. This model also involves a study of production sites, and relies on integration with archaeometric disciplines, and the disciplines of ethnoarcheology and experimentation, highlighting the fact that an interdisciplinary approach is the predominant factor to achieve a result which aims to extract historical depth from the findings.2 The holistic model, proposed by Duistermatt, also offers the possibility of an approach directed at all the nuances, summing up and combining the differing academic traditions.3 Our intention in this study is to give consideration to an all-inclusive and multidisciplinary approach, so as to describe: – the people connected to the pottery artefact, in all the roles that they may occupy (owner of production facility, producer, seller, intermediary in sale, consumer, mender);

1 A recent overview can be found in K. Duistermaat, ‘The organization of pottery production: Toward a relational approach’ in A. Hunt (ed), The Oxford Handbook of Archaeological Ceramics Analysis, online publication Jan 2017, DOI: 10.1093/oxfordhb/9780199681532.013.9. 2 I. Shimada, U. Wagner, ‘A Holistic Approach to Pre-Hispanic Craft Production’ In J. M. Skibo, M. W. Graves, M. T. Stark, (eds), Archaeological Anthropology. Perspectives on Method and Theory, Tuscon, 2007, pp. 163–97. 3 K. Duistermaat, The organization of pottery production… op. cit., pp. 14–25. Francesca Grassi  •  University of Siena, University of the Basque Country Social Inequality in Early Medieval Europe: Local Societies and Beyond, ed. by Juan Antonio Quirós Castillo, Turnhout : Brepols, 2020 (HAMA 39), pp. 99–129 © FHG10.1484/M.HAMA-EB.5.118447

1 00

f r a n ce s ca g r as s i

– the object in itself, in its technological, functional and sometimes symbolic dimension; — the activity as such, namely the act of making pottery, in the following dimensions: practical and organizational (how it is organized); economic (what role it has in the economic fabric in the area being studied); and political (what are the social and political dynamics which allow its existence, and what are the political links which the production highlights). There is also the social and identity-related dimension (what links exist with specific social groups). The assumption which we set out from is that pottery distribution and production systems may be regarded as a direct reflection of the demand of societies, and may show the presence of social stratifications, and suggest the coexistence of several levels of complexity, bound to demand and to the social status of all the people involved in the process.4 Studies of this sort have been successfully conducted also in the case of other areas in Europe, comparing Mediterranean or northern European regions.5 The whole descriptive process is explored in this article by setting out from these premises, based on a study of the links between the social structure of a given area (defined complexity) and the development of the craft industry itself, and the importance of areas regarded as peripheral, with a view to politics and geography.6 The study of the pottery industry would thus allow us to create a link with the ‘complexity’ of early medieval society, and to highlight its stratifications and composition.7 In particular, a social characterization of communities, also in line with the model put into practice on a wide scale by Costin, is possible by means of a study of the phase of consumption and demand which would in part reveal social









4 C. L. Costin, ‘Craft specialization: Issues in defining, documenting, and explaining the organization of production’, in M. B. Schiffer (ed), Archaeological Method and Theory, Volume 3, Tucson, 1991, pp. 1–56; C. L. Costin, ‘Craft production systems’ in G. Feinman, T. Douglas Price (eds), Archaeology at the millennium: A source-book, New York, 2001, pp. 273–327. 5 F. Grassi, A. Vigil Escalera Guirado, ‘La ceramica come indicatore di complessità economica e sociale: un confronto tra due regioni di Italia e Spagna (600–800 d.C.)’, in Archeologia Medievale XLIV, 2017, pp. 271–91; B. Jervis, ‘A Patchwork of People, Pots and Places. Material engagements and the construction of the social’, in Journal of Social Archaeology, 11 (3), 2011, pp. 239–65; P. Reynolds, ‘Material culture and the economy in the Age of Saint Isidore of Seville (6th and 7th centuries)’, in Antiquité Tardive, 23, 2015, pp. 163–210; J. Wroom,‘From One Coast to Another: early medieval ceramics in the southern Adriatic region’, in S. Gelichi, R. Hodges (eds), From one sea to another. Trading places in the European and Mediterranean Early Middle Ages, Turnhout, 2012, pp. 353–91. 6 The research findings come from a large-scale European project conducted in north-west Spain (Basque Country) which envisaged the study of pottery production, distribution and consumption between the sixth and tenth century (Earmedcastile, Horizon 2020 Research and Innovation Programme funds, Marie Sklodowska Curie Action, funding contract no. 656540. www.http:// earmedcastile.blogspot.com); F. Grassi, J. A. Quirós Castillo, A. Alonso Olazabal, L. A. Ortega, C. Fornacelli, ‘State formation in Early Medieval Castile: craft production and social complexity’, in Antiquity, Project Gallery, Vol. 91, issue 358, 2017. 7 C. Wickham, ‘Production, distribution and demand II’, in I. L. Hansen, C. Wickham (eds), The long eighth century, Leiden-Boston-Koln, 2000, pp. 345–79; C. Wickham, Framing the Early Middle Ages. Europe and the Mediterranean, 400–800, Oxford, 2005.

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

Fig. 1. Map of the sites investigated in the European project and in the box the 4 sites in the Alava mentioned in this article.

identities and relations with producers, thereby establishing a link to the technological aspects of production itself, so as to complete the life-cycle of the object and of the people who are involved in it.8 Obviously this reconstruction becomes effective at the territorial level when several sites can be compared with each other.9 For this reason, we have applied this method to a European area with certain characteristics, which can be summed up by the term ‘peripheral’. A peripheral area, in our case, does not mean one that is devoid of complexity, but one primarily without central places such as monasteries, emporia or towns of Roman foundation.10 Our peripheral area of study is the Alava plain, situated in the Basque Country: on top of the physical absence of such places, the area investigated by us also has an absence documented by written sources, in which there is mention of a “no-man’s land” (fig. 1).11 The absence of a central power that directly controls the area encouraged

8 C. L. Costin, ‘Craft Production,’ in H. Maschner (ed), Handbook of Methods in Archaeology, 2005, New York, pp. 1032–1105. 9 I. Santos Salazar, J. A. Quirós Castillo, ‘Territorios sin ciudades y complejidad social. El Cantábrico oriental en la Alta Edad Media’, in F. Sabaté, J. Brufal (eds), La ciutat Medieval i Arqueologia. VI Curs Internacional d’Arqueologia Medieval, Lleida, 2015, pp. 139–74. 10 Iruña Veleia is the only Roman city in the area, but its role in Late Antiquity and the early Middle Ages is unknown (E. Gil, ‘Ciudad de Iruña/Veleia (Iruña de Oca)’, Arkeoikuska 96, 1996, pp. 175–80). 11 S. Castellanos, I. Martín Viso, ‘The local articulation of central power in the North of the Iberian Peninsula (500–1000)’, in Early Medieval Europe, 13 (1), 2005, pp. 1–42.

101

102

f r a n ce s ca g r as s i

experimentation for many centuries.12 As of the sixth century, the territory under investigation was used as a frontier between the Visigoth kingdom of Toledo to the south, and the Merovingian kingdom to the north, but neither of the two kingdoms took direct control over it.13 Moreover, peripheral means that this area, already in the mid-sixth century, would have been disconnected from the trading circuits which distributed Roman fineware, although this phenomenon did not have a homogeneous diffusion from the Atlantic coast to the hinterland.14 In this apparent “vacuum” of power, the possibility emerged clearly of seeing a full set of minor local powers and relationships which had the effect, for many centuries, of enabling settlements and the economy to develop in structure, with the management of small local communities, and links and relations that included political relations.15 The absence of large catalyzing poles seems to have allowed this complexity to flourish, and paradoxically also modern studies, not just this research project, have had more freedom to direct themselves towards a reconstruction of all the territory, in a “horizontal” view, as regards geography and chronology, rather than focus on the reconstruction of an individual site, seen as a “main engine”.16 The structure of the article is as follows: a general overview of the Alava plain from the geographical, historical and archaeological point of view, explaining its main characteristics, and those of the four archaeological sites which are the subject of the research (Aldaieta, Zornoztegi, Aistra, Zaballa) (1), the state of play as regards the issue of pottery studies for the early medieval period in the Basque Country (2), an analysis of the systems of production, distribution and consumption of pottery, as they emerge from the new research findings (3), and two final paragraphs of discussion of the findings, with a conclusion (4–5).

12 In the second half of the fifth century ‘…la zona vascona se convirtió en un espacio al margen del control romano y en un área de paso — y también de control de los germanos hacia la Península Ibérica……estratégicamente importante y desvinculada de un control costante por un poder central…’, I. Martín Viso, ‘La configuración de un espacio de frontera: propuestas sobre la Vasconia tardoantigua’ in U. Espinosa, S. Castellanos (eds), Comunidades locales y dinámicas de poder en el norte de la Península Ibérica durante la Antigüedad tardía, Logroño, 2006, p. 101–139. 13 I. Martín Viso, La configuración de un espacio de frontera… op. cit, pp. 110–11. 14 For example, very different dynamics are seen between the Atlantic area of Gipuskoa (site of Iruaxpe) and the hinterland, see A. Azkarate Garai-Olaun, J. Nuñez, J. L. Solaun Bustinza, ‘Materiales y contextos cerámicos de los siglos VI al X en el País Vasco’, in L. Caballero, P. Mateos, M. Retuerce (eds), Cerámicas tardorromanas y altomedievales en la Península Ibérica. Ruptura y continuidad, Anejos de AEspA, XXVIII, Madrid, 2003, pp. 321–71. 15 A. Vigil Escalera Guirado, J. A. Quirós Castillo, ‘Early medieval rural societies in Northwestern Spain: archaeological reflections of fragmentation and convergence’, in J. Escalona, A. Reynolds (eds), Scale and Scale Change in the Early Middle Ages. Exploring Landscape, Local Society, and the World Beyond, Turnhout, 2010, pp. 33–61. 16 J. A. Quirós Castillo, ‘Early medieval landscapes in north-west Spain: local powers and communities (fifth–tenth centuries)’, in Early Medieval Europe, 19 (3), 2011, pp. 285–311; J. A. Quirós Castillo, (ed), Social Complexity in Early Medieval Rural Communities. The north-western Iberia archaeological record, Oxford, 2016.

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

The Research Area The Alava plain, or Core Alava, lies to the south of the Basque Country region, and has a surface area measuring 556 km2. It is a fertile agricultural plain, interspersed with several hills, in the centre of which today stands the city of Vitoria-Gasteiz. The rivers in the area, in the plain itself (Zadorra), or in the neighbouring zone (Bayas and Omecillo), all flow into the Ebro, which runs parallel to the Basque Country border, 100 km to the south. Of the sites which are the subject of the study, one is situated in the west (Aldaieta), near the large artificial reservoir built on the Zadorra, and two are in the east (Zornoztegi and Aistra), close to the Sierra de Aizkorri-Aratz and to the mountains of Iturrieta and Enzia, and finally there is one last site to the south (Zaballa), on the slopes of the Vitoria mountains, which create a physical barrier with the Miranda di Ebro basin.17 In the last ten years, the intensification of archaeological investigations18 has begun to offer the possibility of developing a new approach to the study of early medieval rural societies.19 In detail, the sites used for this article have fairly differing characteristics. Aldaieta is a necropolis with 116 burials. It was investigated thanks to a chance find made close to a dew pond, created at the end of the 1950s.20 The circumstances of the site’s conservation have not allowed dwelling-places to be found, but only burials;21 many of these had grave goods, and some of the objects recovered also come from surrounding zones. The chronology of the necropolis is situated within the sixth–seventh century, on the basis of stylistic and formal parallels for the objects, although C14 dates circumscribe with precision the time period only to the end of the seventh century.22 The interpretation of the necropolis is not unambiguous: some people claim that the burials have a Merovingian-style aristocratic profile, on the basis of the grave goods, and others use other archaeological data, in particular 17 A. Sáenz de Olazagoitia Blanco, ‘Espacios de interés geomorfológico en Álava’, in Azkoaga, 12, 2004, pp. 247–79. 18 J. A. Quirós Castillo (ed), The Archaeology of villages in the early middle ages, Bilbao, 2009; A. Azkarate Garai-Olaun, L. Solaun Bustinza (eds), Arqueología y Historia de una ciudad. Los orígenes de VitoriaGasteiz, Vitoria-Gasteiz, 2013; J. A. Quirós Castillo (ed), Arqueología del campesinado medieval: la aldea de Zaballa, Bilbao, 2011; J. A. Quirós Castillo (ed), Vasconia en la Alta Edad Media (450–1000). Poderes y communidades rurales en el Norte Peninsular, Bilbao, 2011. 19 J. A. Quirós Castillo, (ed), Social Complexity… op. cit. 20 A. Azkarate Garai-Olaun, Necropolis tardoantigua de Aldaieta (Nanclares de Gamboa, Alava), Vitoria Gasteiz, 1999. 21 J. A. Quirós Castillo (ed), A. Azkarate Garai-Olaun, R. Bohigas Roldán, I. García Camino, Á. L. Palomino Lázaro, J. M. Tejado Sebastián, ‘Arqueología de la Alta Edad Media en el Cantábrico Oriental’, in Medio Siglo De Arqueología En El Cantábrico Oriental Y Su Entorno, 449–500, VitoriaGasteiz, 2009, pp. 449–501. 22 Burial B14 dated to seventh century (1 sigma 647–89, 65.3 %), in A. Azkarate Garai-Olaun, J. Nuñez, J. L. Solaun Bustinza, Materiales y contextos cerámicos…op. cit, p. 359; A. Azkarate Garai-Olaun, ‘The Western Pyrenees during the Late Antiquity. Reflections for a reconsideration of the issue’, in G. P. Brogiolo, L. Castelletti (eds), Il territorio tra tardoantico e altomedioevo. Metodi di indagine e risultati, Firenze, 1992, pp. 179–91.

103

104

f r a n ce s ca g r as s i

the paleo-diet and grave goods, to describe a low social level, without any apparent hierarchization, identifiable by means of a diet based on a high consumption of plants.23 The three remaining sites are villages located in fertile spots at the bottom of the valley, in low-lying flat areas or in a hilly position. This is a particular model of occupation and exploitation of local resources, with a vocation for farming and herding animals, as shown by the presence of agricultural terraces around the villages. Zornoztegi and Aistra are located in the eastern part of the Alava plain, in the area of the municipality of Salvatierra. Although the villages are very near to each other, their material and social characteristics are very different. The Zornoztegi site existed between the fourth and fourteenth century ad. The excavation, currently in press, has made it possible to investigate an initial phase of settlement datable to the late Roman period (fourth–fifth century), in which a number of houses were identified, built using mixed materials, which formed the nucleus of a rural farm. As of the sixth century, a residential nucleus developed, of which a house and surrounding cultivated land have been identified. But it is only as of the eighth century that a village took shape, home to a community with social differentiation, with a maximum of 10 individual dwellings connected to storage silos for cereals. This village did not have spaces for collective or community use, or a cemetery; only in the twelfth century was a church built, dedicated to St Mary.24 The Aistra settlement site is today identifiable by its rural church, still standing, built in the tenth century.25 The excavation has brought to light, around the church, a settlement with occupation phases dating from the sixth to the eleventh century. The whole medieval settlement was probably founded near a Roman settlement, as borne witness to by a large amount of pottery found in the early medieval domestic structures. The early medieval sequence began in the fifth century, when agricultural terraces were created, and a village rose up, featuring dwellings made of perishable materials, a well, and a number of silos. This village, datable to between the sixth and seventh century, was formed by a number of domestic dwellings which

23 Most recently, J. A. Quirós Castillo, ‘Ceremonial funds, gifts and peasant societies in the light of North Iberia archaeological evidence, 5th-8th centuries’, paper presented in the conference Peasant households in the landscape. Peasant production and consumption patterns in Early Medieval Europe, Oxford, on 15-16 2016 July. 24 J. A. Quirós Castillo, (ed.), Arqueología de una comunidad campesina medieval: Zornoztegi (Alava), Bilbao, 2019; for pottery, F. Grassi, ‘La cerámica medieval’, in J. A. Quirós Castillo, (ed.), Arqueología de una comunidad campesina medieval: Zornoztegi (Alava), Bilbao, 2019, pp. 263–93. 25 J. A. Quirós Castillo, ‘Archaeology of power and hierarchies in early medieval villages in Northern Spain’, in J. Klápště, (ed.), Hierarchies in rural settlements, Turnhout, 2013, pp. 199–212; A. Reynolds, J. A. Quirós Castillo, ‘Despoblado de Aistra (Zalduondo)’ Arkeoikuska, 2007–2010. On pottery, in a comparison with Zornoztegi, see also F. Grassi, ‘La complessità sociale e politica analizzata attraverso la ceramica: i villaggi di Aistra e Zornoztegi nella pianura di Alava nei secoli VI e VII d. C. (Álava)’, in I. Martin Viso, P. Fuentes Melgar, J. C. Sastre Blanco, R. Catalan Ramos, (eds), Cerámicas altomedievales en Hispania y su entorno (V–VIII siglos), Valladolid, 2018, pp. 116–33.

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

also comprised a longhouse-style building.26 In the eighth century the site was transformed, with the creation of new homes and a cemetery area with a church. This site, unlike Zornoztegi, seems to display the direct presence of elites, in the shape of the two longhouses, and as a result of a study of meat consumption, and a number of metal finds.27 Finally, the village of Zaballa has been the subject of a research project already published.28 The settlement history of Zaballa began between the sixth and seventh century, the date given to an initial occupation in the form of a small rural farm. As of the eighth century evidence of a village has been recognised, composed of huts, with large farmland spaces situated in the bottom of the valley and systems for the storage of grain, with raised granaries and underground silos. In the tenth century, the construction of the church of San Tirso, together with a new spatial organization, led to the destruction of many dwellings in the previous village, and a substantial reorganization of the whole site.

The Current State of Pottery Studies in the Basque Country: The Early Medieval Period Studies of early medieval pottery in the Basque Country have addressed the following issues: classification of pottery wares, analysis of production sites, and a reconstruction of socio-economic models of production, distribution and consumption. As regards pottery classification, three distinct volumes giving overviews cover this area,29 and show primarily a high degree of isolation as regards the circuits of Mediterranean trade, active in the Iberian Peninsula up until the seventh century and, as a secondary aspect, models of consumption which are very different from

26 On possible interpretations of the ‘longhouses’ in southern Europe, see most recently J. A. Quirós Castillo, ‘Longhouses, biografía de la casa y complejidad social en el noroeste peninsular en la alta edad media’, in Arqueología de la Arquitectura, 14: e059, doi: http://dx.doi.org/10.3989/arq.arqt.2017.019. 27 I. Grau Sologestoa, The zooarchaeology of medieval Alava in its Iberian context, Oxford, 2015, pp. 75–87; J. A. Quirós Castillo, ‘Archaeology of power and hierarchies in early medieval villages in Northern Spain’, in J. Klápštē (ed), Hierarchies in rural settlements, Ruralia IX, Turnhout, 2013, pp. 199–212. 28 J. A. Quirós Castillo (ed), Arqueología del campesinado medieval… op. cit.; L. Elorza, S. Gobbato, ‘Estudio de la cerámica del despoblado de Zaballa’, in J. A. Quirós Castillo (ed), Arqueología del campesinado medieval… op. cit., 2011, pp. 236–89. 29 F. Sáenz de Urturi Rodríguez, ‘La cerámica medieval no esmaltada en yacimientos alaveses’, in J. A. Gutierrez González, R. Bohigas Roldan (eds), La cerámica medieval en el Norte y Noroeste de la península Ibérica. Aproximación a su estudio, Leon, 1989, pp. 53–87; A. Azkarate Garai-Olaun, J. Nuñez, J. L. Solaun Bustinza, Materiales y contextos cerámicos… op. cit, pp. 321–71; A. Azkarate Garai-Olaun, J. L. Solaun Bustinza, ‘La cerámica altomedieval en el País Vasco (siglos V–X d.C.): producciones, modelos productivos y patrones de consumo’, in J. A. Quirós Castillo, A. Vigil-Escalera Guirado (eds), La cerámica de la Alta Edad Media en el cuadrante noroeste de la Península Ibérica (siglos V–X). Sistemas de producción, mecanismos de distribución y patrones de consumo, Bilbao, 2016, pp. 193–229.

105

1 06

f r a n ce s ca g r as s i

the surrounding territory, such as the Duero basin or the Toledo region, in terms both of the repertoire of shapes and the amount of pottery.30 In addition to these general overviews there have been two monographs providing a complete picture of pottery wares produced.31 These two works chronologically leave a gap in discussing the sixth–eighth century, with the former dealing with the first–fifth century ad and the latter with the eighth–fourteenth century ad. These overviews were drawn up on the basis of research doctorates, based on a geographical division linked to the modern-day administrative territory of the Basque Autonomous Community.32 Both studies provide an extensive pottery classification, and a distribution map of forms and fabrics, also subjected to petrographic and x-ray scattering analyses.33 The general model that takes shape from these studies, viewing the pottery of the early medieval centuries specifically, is as follows: between the end of the fifth century and the mid-sixth century, we see the disappearance of Roman finewares for the table (THST, DSP),34 and a new production of locally-made jugs with very coarse fabrics, made by hand/slow wheel. Thus we see a total transformation of domestic wares, composed only of pottery for cooking and marked by the disappearance of table wares and transportation vessels. The two following centuries (eighth-tenth century), are described with an increase in the number of functional and technological types: among cooking vessels there is an expansion in the morphological and decorative repertoire, and production remains limited to shaping by hand, or use of the slow wheel. Also, new wares make up the pottery assemblage, including a type of levigated pottery characterized by fine clay fired in an oxidizing atmosphere, and a number of levigated red-slip wares, referred to as of “non-local” production. In the case of this time period, too, there are no known production centres, and suggestions as to areas of production derive from archeometric analyses of the fabrics.35

30 C. Tejerizo Garcia, Arqueología de las sociedades campesinas en la cuenca del Duero durante la Primera Alta Edad Media, Bilbao, 2017; F. Grassi, A. Vigil Escalera Guirado, La ceramica come indicatore di complessità economica e sociale… op. cit., pp. 271–91. 31 A. Martínez Salcedo, La cerámica común de época romana en el País Vasco, Vitoria- Gasteiz, 2004; J. L. Solaun Bustinza, La cerámica medieval en el Pais Vasco (siglos VIII–XIII), Vitoria- Gasteiz, 2005. 32 In J. L. Solaun Bustinza, La cerámica medieval… op. cit., a comparative analysis is offered of around 18,000 sherds from 13 sites; in A. Martínez Salcedo, La cerámica común… op. cit., the work is based on only 1,500 sherds from 3 sites (including the Roman town of Iruña Veleia). 33 L. A. Ortega, M. C. Zuluaga M. C., ‘Resultados de los análisis arqueométricos de laboratorio’, in J. L. Solaun Bustinza, La cerámica medieval…op. cit., pp. 277–305. 34 TSHT and DSP (Terra Sigillata Hispanica Tardía and Derivate Sigillate Paleocristiane, or grey sigillatas) are the two main ‘families’ of fine tableware present in northern Spain. The former was produced throughout Spain, and the latter is of Gallic production. 35 Three areas are identified as zones of pottery production, on the basis of the fabrics analysed: the first linked to Trias diapir deposits, the second to the Bayas river basin, and the third to the floodplain of the river Zadorra, between Vitoria and Miranda di Ebro ( J. L. Solaun Bustinza, La ceramica medieval… op. cit., pp. 147, 153, 172, 207).

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

As regards the current state of research on production centres and production models which derive from them, the reconstruction we can put forward is very sparse, owing to the almost total lack of workshops that have been found. A recent summary36 records a total of 19 early medieval kilns in northern Spain, all situated in the western part of the peninsula, and active only up until the eighth century. In the Basque Country, specifically, there are three possible production models, put forward on the basis of the remains of two archaeological kilns, and on the basis of written sources: 1. Itinerant domestic production. This model was drawn up on the basis of the production facility at the site of Bagoeta, 10 km north of Vitoria-Gasteiz.37 In the village an oval-shaped, rock-cut structure was investigated, consisting in a single chamber. It was abandoned in the tenth–eleventh century. In the fill, abundant remains were found of a kind of jug made on the slow wheel, and several pieces of kiln waste.38 Hence its interpretation as a “seasonal” pottery kiln, active in the tenth century, and used to make the type of jug found in the abandonment layers. This type of productive structure was associated with a “domestic and itinerant” form of production, given the highly representative nature of this type of jug throughout the Alava plain, and the fact it was produced with at least three different fabrics.39 2. Specialized workshop. This model was drawn up on the basis of research at the site of Rivabellosa (Ribera Baja, Alava). Here, archaeological investigations did not yield production facilities, but two silos filled with 65 jugs, formally very similar to each other. It is suggested that these are part of the production of a distinct specialized workshop. The fill was dated by C14 as being between the eighth and tenth century, and the chronology was used to date the jugs themselves, the circulation of which was recognized in a radius of 20–30 km, on the basis of other finds in the area.40 3. Multiple workshops. Despite the absence of archaeological finds, it is suggested that there was another clearly defined form of production in the shape of multiple workshops, located in villages with a vocation for pottery-making. This hypothesis is based on the existence of toponyms and written documents referring to the activity of the potter, and cites the example of the village of “Olleros”, cited in a document dating to ad 871.41

36 L. C. Juan Tovar, ‘Talleres y hornos cerámicos tardoantiguos y altomedievales en el noroeste peninsular: estructuras y tecnología’, in J. A. Quirós Castillo, A. Vigil Escalera Guirado (eds), La cerámica de la Alta Edad Media en el cuadrante noroeste de la Península Ibérica (siglos V–X). Sistemas de producción, mecanismos de distribución y patrones de consumo, Bilbao, 2016, pp. 339–63. 37 Bagoeta was a village inhabited as of the seventh century, specializing in iron production, a hypothesis formulated on the basis of the amount of slag recovered during field research: A. Azkarate GaraiOlaun, J. M. Martínez, J. L. Solaun Bustinza, ‘Metalurgia y hábitat en el País Vasco de época medieval: el asentamiento ferrón de Bagoeta, Álava (ss. VII–XIV d.C.)’, in Arqueología y Territorio Medieval, 18, 2011, pp. 71–89. 38 A. Azkarate Garai-Olaun, J. L. Solaun Bustinza, La cerámica altomedieval… op. cit., footnote 42. 39 A. Azkarate Garai-Olaun, J. L. Solaun Bustinza, La cerámica altomedieval… op. cit., fig.7.8, pp. 223–25. 40 J. L. Solaun Bustinza, La cerámica medieval… op. cit., pp. 131–32, pp. 357–59, p. 381. 41 J. L. Solaun Bustinza, La cerámica medieval… op. cit., pp. 360–61.

107

1 08

f r a n ce s ca g r as s i

Site

Name

Sherds

Weight (Kg)

NVR

Rim

Bottom

Extension excavation

Village Village Village Necropolis Total

Aistra Zornoztegi Zaballa Aldaieta

4429 2218 1732 885 9264

19,162 9,881 12,2 2,905 44,148

515 195 160 28 898

279 148 134 75 636

244 135 129 63 571

2.581,55 m2 4.299,2 m2 33.641,1 m2 116 graves

Fig. 2. Complete pottery assemblages from the Alava sites, sixth–tenth century (number of sherds, weight, number of vessels represented).

Analysis and Interpretation of the Archaeological Evidence: New Data from the Pottery Record Quantitative Findings

The five centuries covered by this research study are subdivided into two separate blocks, mainly the result of the settlement sequences of the sites referred to: sixth– seventh century; and eighth–tenth century. Looking at the pottery, the sixth–seventh chronological block has raised many problems, and forced us to find solutions: the pottery kits in this 200-year long period underwent radical change. In the sixth century we see a survival in the use of pottery forms produced in the fifth century, although this is a very dishomogeneous phenomenon that is clearly evident only in Aistra. By contrast, in the seventh century these survivals no longer exist, and the kit is composed of a new type of early medieval pottery. However, the very uniformity of these two centuries in terms of sites — although not in terms of pottery itself — has not always allowed us to make this distinction as we would have liked. Setting out from a purely quantitative analysis, the data compiled with the project for the Alava plain corresponds to 9,264 sherds, namely 44 kg and 898 minimum forms (fig. 2). From the initial quantitative data, one sees the three villages (Aistra, Zornoztegi and Zaballa) display a number of sherds that is inversely proportional to the size of the excavated area. This is partly the result of chance, and partly the result of factors due to the destruction of the archaeological deposits. In the case of the necropolis, it also seems significant to us to include the number of burials investigated (116), and to relate it to the low number of individual vessels (28), a sign that the use of pottery in grave goods was not at all deeply-rooted. To assess the type of deposit and its formation, one profitable method was to calculate the Index of Fragmentation (IF), which can be derived from the ratio between weight and fragments. Low numbers indicate a high index of fragmentation, while on the other hand a high number indicates a low index of fragmentation. Concretely, as regards our research, we note that the Alava Plain settlements are associated with high IF perhaps caused by the Plain’s recent history, which saw the removal of deposits, to cultivate fields. The only exception in this area is the village of Zaballa during the sixth–seventh century (fig. 3).

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

Site 1 2 3 4

Aldaieta Aistra Zornoztegi Zaballa

VI-VII

VIII-IX

X

3,28 4,57 4,16 18,5

4,26

4,26 4,42

5,14

6,31

Fig. 3. Comparison of fragmentation indices for the sixth and seventh century at the sites (sherds/weight).

Site

VIII-IX

Aistra Zornoztegi Zaballa

15%

X 30% 2%

4%

5%

Fig. 4. Percentage of residuality from the Alava sites, eighth–tenth century (sherds/%).

Finally, using the quantitative figures, we calculated the percentage of residuality on the basis of the number of fragments for the eighth–tenth century, except for the site of Aldaieta, where there are no stratigraphies datable to these time periods (fig. 4). In general, the residual presence of Roman pottery material, easily identifiable by form, ware, and fabric, is a tiny phenomenon, with the exception of Aistra. As regards the seventh century, we are unable to provide data, because the site sequences do not allow us to distinguish it chronologically from the sixth century, as stated earlier. Production Systems

The almost total absence of identified production facilities forces us to discuss producton indirectly, referring to the technology of the pottery that has been found, a description of fabrics, and making suggestions as to their possible provenance on the basis of archaeometric analyses. 1. Production technologies.42 Among the pottery (fig. 5), we see that 30% of pottery wares in the sixth and seventh century consists of pottery of Late Antique production (on the double wheel, on the single wheel, and mould-made), while early medieval pottery, corresponding to 70%, is predominantly produced on the single wheel (58%), with a minority made by hand (12%).43 For the eighth century, the findings show the disappearance of pottery made on the double wheel, and handmade pottery, mainly in the Roman tradition, and the 42 By production technologies, we mean pottery processing techniques, distinguished between handmade, single wheel, double wheel, and mould-made (P. M. Rice, Pottery Analysis. A sourcebook, Second Edition, Chicago, 2015, pp. 135–46). 43 The quantitative data is grouped for the four sites, but it should be noted that, in the sixth–seventh century, the site of Aistra is the site that has the largest amount of Late Antique pottery, compared to Zaballa, Zornoztegi and Aldaieta, where this type of pottery appears in percentages of less than 5%.

109

110

f r a n ce s ca g r as s i

Fig. 5. Production technologies present in sixth–seventh century (sherds/%).

Fig. 6. Production technologies present in eighth–tenth century (sherds/%).

absolute prevalence of modelling on the single wheel, and thus with a radicalization of this production technology (fig. 6). 2. As regards fabric characteristics as derived from archaeometric analyses, in the sixth and seventh century the majority are constituted of mould-made pottery with

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

Fig. 7. Characteristics of pottery fabrics in sixth–seventh century (sherds/%).

large inclusions (71%), designed to be exposed to flames for cooking. The remaining 29% of the pottery has levigated fabrics (including TSHT fineware — Terra Sigillata Hispanica Tardia), or with finely ground inclusions (fine coarseware, for the preparation of food and for storage) (fig. 7). In the following two centuries there is a change: 55% of the pottery kit consists of pottery with levigated fabrics, while the remaining 48% consists of pottery with coarse fabrics, although generally of the fine coarseware type (fig. 8). Let us dwell for a moment on the contemporaneous presence of coarseware with more or less coarse fabrics, called here coarseware and fine coarseware, for convenience. Both have heat-resistant features, and thus fall within the category of wares for cooking or preparing food, but the degree of difference in the clay mix, together with the macroscopic observation of traces of fire on the outside, leads us to think of a use for different types of cooking. The coarseware pots have large refractory inclusions, and clear signs of blackening over the entire surface, which is why it has been suggested that they were placed in burning embers, in direct contact with fire. By contrast, pottery with fine/coarse fabrics are characterized by very small micaceous inclusions (which give the surfaces an iridescent appearance), and traces of fire only on the edge and on one side of the pot. Accordingly, they do not seem to have been used for cooking in direct contact with fire, but only near flames. This specialization is one factor which could be linked to diet, and we will return to it in the section on eating (below, 3.4). This fact shows a clear selection on the part of workshops in choosing raw materials on the basis of functionality.

111

112

f r a n ce s ca g r as s i

Fig. 8. Characteristics of pottery fabrics in eighth–tenth century (sherds/%).

3. Finally, as regards provenance, we can sum up by saying that the raw materials of the pottery examined (clays and filler) all come from in and around the central Alava plain, but with some specific features.44 First, pottery called coarse pottery is always characterized by large inclusions from areas of Trias break-up, located in certain specific areas, not near any of the sites. Meanwhile, the pottery called fine coarseware comes from the zones of the rivers Zadorra and Bayas, and partly contains the same detritus as the Trias, but it is smooth, having been river-borne. Accordingly, these clays maintain the characteristic of being heat-resistant, thanks to the inclusions, but they are much more refined. Finally, the pottery referred to as levigated pottery are of two types, one with a clay mix similar to that of fine coarseware, but even more refined, and another that is very levigated, and present above all at southern sites (Zaballa), showing affinities with pottery made using clays from the river Ebro area. A differentiation needs to be made for the sixth century. In this period, Late Antique pottery forming part of table wares and cooking wares seems to have been produced mainly at production centres outside the Alava plain, as shown by the fabrics, absent in the later century. Also, as regards fine tableware (TSHT), we know from literature that the production centers which supplied the Basque Country were situated mainly in the Rioja area, and near Burgos, connected to the area of the Ebro.45

44 L. A. Ortega, A. Alonso Olazabal, J. A. Quirós Castillo, M. Cruz Zuluaga, ‘Arqueometria de la cerámica comun altomedieval enÁlava: resultados para los yacimientos de Zaballa y Zornoztegi’, in F. Grassi, Quirós Castillo J. A. (eds), Arqueometría de los materiales cerámicos en la Península Ibérica: métodos y estrategias para el futuro, Bilbao, 2018, pp. 111–33. 45 J. A. Paz Peralta, ‘Las producciones de terra sigillata hispanica intermedia y tardia’, in D. Bernal Csasola, A. Ribeira i Lacomba (eds), Ceramicas hispanorromanas. Un estado de la cuestion, Cadiz, 2008, pp. 497–541. R. Morais, A. Fernandez Fernandez, ‘Difusion y comercio de la TSHT. Nuevos

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

Thus, from a synthesis of this information, we can suggest the following hypotheses for the early medieval production facilities in Alava: – early medieval workshops active in the sixth century introduce their new products alongside others made in the fifth century, which certainly continue distribution, at least to some sites. The sixth century is the moment when pottery repertoires are rich in products with technological standards from the previous century (fast wheel/mould, non-local clays), and at the same time with early medieval pottery with characteristics that are in part new (slow wheel, local clays). This combination creates an interesting factor as regards form and function: In the sixth century one seems to note that the output of the new workshops finds a niche that is not supplied by the older wares, producing only objects for the kitchen and food storage vessels. Among the sites analysed, Aistra is certainly the one which has the highest percentage of this type of “mixed” repertoire (37% of all the pottery in the sixth century is composed of fifth century pottery), allowing us to put forward this hypothesis. – in the new workshops, a preference becomes established for shaping on the slow wheel. In our view, this decision is not linked to a loss of technological skill compared to the use of more serialized means of production (fast wheel, mould). Indeed, the pottery products are characterized by thin walls, frequent incised decoration, and regularity of form. What is more, observing the fact that, in the following centuries, we see a continuation of this use of technology, we can suggest that the choice was connected to the actual way in which local potters made pottery.46 – the few wares made by hand, without the help of equipment to rotate the pot, could be the result of domestic production, at individual settlement nuclei, following a generalized interpretation of this kind of pottery. But it seems to us that two factors argue against this hypothesis: first, the use of the same clays as those found in pottery made on the slow wheel; second, the use of the same forms, and their homogenous distribution across the local area. Thus it is easier to suppose that handmade wares are also the product of artisanal (small-scale) workshops. – the workshops were apparently located in and around the central Alava (Core Alava), but it is difficult to take this hypothesis further, on the basis of mere indirect evidence; – the workshops in these centuries show a formal identity in the products they made, with peculiar choices as regards function, as we have seen for the presence of coarseware and fine coarseware side by side. – all the workshops seem to make the same products, thus we cannot give quantitative information. In other words: was there just one producer centre? Are there several centralized producer centres, or several centres scattered across the area?47 What yacimientos, estructura comercial y evolucion de la misma, expansion y contraccion de mercados’, in Ex Officina Hispana, Quadernos de la Secah, I/2013, pp. 47–64. 46 J. L. Solaun Bustinza, La ceramica medieval… op. cit., pp. 322–38. 47 These production-related features have also prompted a suggested itinerant production model (A. Azkarate Garai-Olaun, I., Garcia Gomez, J. L. Solaun Bustinza, ‘El asentamiento altomedieval de Gasteiz’, in A. Azkarate Garai-Olaun, (ed.), Arqueologia y Historia de una ciudad. Los origenes de Vitoria-Gasteiz, Vitoria-Gasteiz, 2013, pp. 399–446).

113

114

f r a n ce s ca g r as s i

seems certain is that all the workshops look to the same clay extraction areas (Trias deposits, the river Zadorra and the Bayas); – finally, these producer centres do not seem permeable to technological/decorative changes and/or innovations, such as painted decoration, slips, or the use of the fast wheel, although present in neighbouring areas. Indeed, the presence in the eighth century of the first painted pottery in the study contexts does not show technological innovation, it shows the first imports from the Ebro area. The same is true as regards the use of the fast wheel, a form of production used in the Ebro area itself, but not in Alava. These factors may lead us to describe the Alava plain workshops as relatively “closed” production sites, at which deeply-rooted technological processes were implemented, without an input from outside. Pottery Distribution Networks

Moving on to pottery distribution, in Alava we can note that, between the sixth and the tenth century, only some forms of pottery circulation can be identified. Before describing them, let us define a number of terms that we will use in this paragraph. First, Alava will be regarded as a region, although not in an administrative sense, but only in a geographical and cultural sense. Accordingly, in referring to extra-regional trade, we will mean trade with other areas or regions, both neighbouring (short-range), but separated by mountains or rivers, as the Ebro zone to the south could be, or Biscay to the north, and distant (long-range), such as the Mediterranean or Atlantic areas. The term regional trade will be used to refer to trade that took place within Alava. In this case, too, we will make a distinction between long-range (around 40–50 km) and short-range or local, being around 10–15 km. Thus, there are presently just two forms of trade identifiable in Alava, on the basis of data from pottery: Extra-regional, short-range trade. One product of this type of trade is red-painted pottery which, as of the eighth century, comes from the neighbouring area of the Ebro, along with levigated pottery, but without slips and decoration, recognisable from the fabric. This trade is very rare, and remains at about 1% throughout the plain, although it is totally absent in some cases, such as at the site of Zornoztegi and Aistra.48 The arrival of TSHT, and common and levigated coarseware with Late Antique fabrics, in the sixth century, may equally be attributed to this type of trade, although this is only a working hypothesis connected to the only production sites that we know, those in the Ebro area, and which would presuppose continuity of use of Roman pottery distribution networks throughout the sixth century.49

48 F. Grassi, C. Fornacelli, ‘New archaeometrical data on Red Painted pottery: case study of Northern Iberia’, in F. Grassi, J. A. Quirós Castillo (eds), Arqueometría de los materiales cerámicos en la Península Ibérica: métodos y estrategias para el futuro, Bilbao, pp. 133–63. 49 R. Moiras, A. Fernandez Fernandez, 2013, Difusion y comercio… op. cit., pp. 49–50.

s o c i al co m p l exi t y i n pe ri phe ral are as LATE ANTIQUE FORMS

MEDIEVAL FORMS Jug/Flask

Jug/Jar

Dolium

Cooking Pot Fine ware

TSHT

Cooking Pot-Lid

Fig. 9. Chronotypological table of ceramic forms in sixth–seventh century.

Regional long-/short-range trade. Basically, all the pottery trade in Alava falls within the regional mode, although we are unable to distinguish between short range and long range, insofar as the very similar geology throughout the plain, and around it, together with the absence of finds of production centres, does not allow us to locate the workshops with precision. Consumption at Sites

As regards consumption, we have tried to reconstruct this via analysis of the forms, although the highly fragmentary nature of the pottery remains has hindered this part of the study. Indeed, there are 728 fragments (around 12% of the total) which are still undetermined, since they are small pieces of walls, or bases of pots, for which it has proved difficult even to determine whether they come from open or closed vessels. Thus, here we give a summary table, by chronology and type, for centuries examined (figs 9 and 10). For the sixth and seventh century, before describing the identifiable repertoire, a division needs to be made between forms made with medieval fabrics and forms made with Late Antique fabrics, given that their presence side by side is a feature of consumption in this period, especially at the site of Aistra (fig. 11). Moving on to details of the forms, open vessels are in a marked minority, and many are made with Late Antique fabrics. They are highly recognisable: they consist in fragments of TSHT, and cooking and storage vessels already present in assemblages, at least since the fifth century.

115

116

f r a n ce s ca g r as s i

Fig. 10. Chronotypological table of ceramic forms in eighth–tenth century.

Fig. 11. Percentage quantification of repertoires with Medieval and Late Antique fabrics (sixth–seventh century) (sherds).

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

FORMA

Medieval fabrics

Late Antique fabrics

Cooking Pot Jug/Jar Flask dolium Closed wessel bowl plate Little bowl Open vessel Lid Indeterminate TOTAL

584 169 0 0 354 40 0 0 66 0 294 1507

0 14 4 3 110 0 3 4 58 1 77 274

Fig. 12. Quantification and division of forms, on the basis of fabric type, in sixth and seventh century (sherds).

Fig. 13. Summary: sixth–seventh century (sherds/%).

As regards closed vessels, cooking pots and jugs/jars are almost always associated with medieval coarse fabrics, but in this case too the number of fragments with Late Antique fabrics is nevertheless high. Finally, as regards lids, only one type of form is present (fig. 12). To sum up (fig. 13), sixth and seventh century pottery repertoires are composed of a majority of closed vessels made with medieval fabrics, belonging to tableware containers, such as the jug/jar, together with pottery for cooking, such as cooking pots (accounting for 85% of all closed vessels).

117

118

f r a n ce s ca g r as s i

Shape

Sherds

Open vessel Closed vessel Cooking Pot Jug Jar Bowl Lid Indeterminate TOTALE

3 2412 712 609 68 66 3 351 4224

Fig. 14. Quantification of forms in eighth–tenth century (sherds).

Fig. 15. Summary: eighth–tenth century (sherds/%).

As regards open vessels, on the other hand, their presence in the assemblage mainly takes the form of Late Antique wares, often slip-decorated or plain and undecorated, although they can all be identified as storage vessels. In these two centuries there is no open vessel for cooking food. Moreover, levigated pottery for food preparation (large bowls, basins, small bowls) is also almost absent. Moving on to the eighth-tenth century (Figs. 14–15), the repertoire is mainly composed of closed vessels, especially cooking pots, and jugs/jars. Open vessels are still rare. In terms of specialization of forms, not much has changed from the previous centuries, apart from the disappearance of all the repertoire from the Late Antique tradition. This means that, together with this disappearance, early medieval workshops do not begin to produce a greater variety of forms, and in fact there continues to emerge the absence of lids and open vessels for cooking, which

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

Fig. 16. Percentage summary of technological wares making up the sixth–seventh century pottery repertoire: above, with Medieval fabrics; below with Late Antique fabrics (sherds/%).

are present in the same centuries in neighbouring regions. In short, the domestic pottery kit is composed of high-walled pots for cooking, and small jugs, larger jugs and jars for the kitchen and for storage. As for consumption, we propose one last illustration showing the percentage presence of the individual technological wares:50 coarse pottery, fine coarseware, levigated pottery, and, as of the eighth century, red-painted pottery. In addition, we have taken into account Terra Sigillata Hispanica Tardia for sixth century contexts (figs 16–17). 50 By ‘technological class’ - CT - we mean a group of objects displaying similar technological and functional characteristics, namely fabric type and features connected to use. Contained within it are pottery forms and types (M. Milanese, ‘Le classi ceramiche nell’archeologia medievale tra

119

1 20

f r a n ce s ca g r as s i

Fig. 17. Percentage summary of technological wares making up the eighth–tenth century pottery repertoire (sherds/%).

In the sixth-seventh century, in the part of the assemblage made up of medieval pottery, there is consumption of the three pottery classes, with an imbalance weighted towards the function of cooking foods (coarse pottery and fine coarseware, 60% in all), at the expense of levigated pottery for food storage (conserving food for the short term) and for consumption. Moreover, the part of the assemblage composed of Late Antique pottery (produced in the fifth century) allows us to note the presence of fineware (TSHT and levigated pottery), also accounting for the function of tableware. In the eighth–tenth century, when the pottery range is composed only of medieval pottery, we note the following: an increase in levigated pottery for food storage; the disappearance of vessels made with coarse fabrics, in favour of vessels with more refined coarse fabrics; the disappearance of all tableware; and finally the introduction of pottery painted with red slip designs. Regarding the increase of levigated pottery in these centuries, it is possible that this is in part a reflection of the disappearance of the tableware repertoire from the Late Antique tradition. Regarding the use of fine coarseware appearing alongside coarse pottery, we can make some further considerations, addressing the issue of types of diet. In the sixth and seventh century the two types of cooking vessels exist side by side, but as of the eighth century we note the clear disappearance of coarse pottery in favour

terminologie, archeometria e tecnologia’, in Le classi ceramiche. Situazione degli studi — Atti della 10a Giornata di Archeometria della Ceramica, Bari, 2009, pp. 47–55; T. Mannoni, La ceramica medievale a Genova e nella Liguria, Genova, 1975, pp. 164–81).

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

of fine coarseware. On the back of these findings, our hypothesis is that the change must have also limited the possibility of cooking in these pots. Indeed, based on the traces left on the surfaces, we can safely say that they were not used in direct contact with hot embers, as stated in paragraph 3.2. Thus, it seemed interesting to observe whether there were any changes in eating habits at the same time, visible as of the eighth century itself. Going back to a study of the taxa relating to the bone remains of animals from three of these sites (Zaballa, Zornoztegi, Aistra), we see that, in the centuries from the eighth to the tenth C, there is a predominance of sheep and/or goats, also used for meat, as well as for wool and milk.51 The second taxon present is cattle, but used mainly for traction, and the third would be pigs, used for meat or for reproduction, and the fourth is chickens, for eggs and meat.52 Thus, as far as meat is concerned, the type most often cooked seems to be lamb/goat, which, at least in traditional cooking, needed very lengthy cooking, on direct flames or in a wood fire, often in metal containers (in Italy a traditional cooking vessel is the copper cauldron, especially in the central Apennines).53 Looking at the kind of meat that was mostly consumed, we have to think that these kitchen vessels with fine coarse fabrics and few traces of blackening must have served to cook other types of food, such as cereals or vegetables, given that they do not display signs of having been in direct contact with fire. It would seem that they were most likely used as vessels to heat up food near the hearth, rather than to cook (fig. 18). Finally, as regards the red-painted pottery that appears as of the eighth century, certain considerations can be made. First, regarding the fact they exist in such small numbers (1% of the total of the assemblage), and also regarding their extra-regional provenance, and finally the fact that Zaballa is the only site in our study that features this pottery, in the form of two fragments.54 It was very important that we were able to establish, in the course of our research, thanks to chemical analyses, that the place where the raw materials were extracted was along the river Ebro: thus, red-painted pottery is an example of extra-regional trade, and this could have been a factor in its low presence.55 In the discussion below, we will also reflect on the possible social value of this pottery, placing it within the general context.

51 I. Grau Sologestoa, The zooarchaeology… op. cit., pp. 129–31. 52 I. Grau Sologestoa, The zooarchaeology… op. cit., pp. 131–33; figure 13.10, p. 130; figures 13.17–13.18, p. 136. 53 M. Weiss Adamson, Food in Medieval Times, Grenwood, Connecticut-London, 2004, pp. 55–83. 54 In the Alava plain we only find two other instances of red-painted pottery: at the sites of Gasteiz and Dulantzi, where 2 sherds were found for each site (A. Azkarate Garai Olaun, J. L. Solaun Bustinza, La cerámica altomedieval… op. cit., p. 209). 55 The new archeometric analyses of red-painted pottery found in north-western Spain are published in in C. Fornacelli, F. Grassi, New archaeometrical data on Red Painted Pottery… op. cit.

121

122

f r a n ce s ca g r as s i

Fig. 18. Rim, handle and base of a jag with fine coarse fabric (Zornoztegi, tenth century).

Discussion In line with a tradition in archaeological studies, finding in a given territory a specialized artisanal organization would correspond to a complexity in the social context.56 If we apply this paradigm to our area, and to the characteristics of production, we would have to consider the society of Alava as devoid of complexity. Indeed, pottery craft production has shown a level of subsistence, visible from the absence of the use of a wheel, the absence of pottery with slip decoration, and the absence of any outside technological contribution. As a result, we would have to suggest the absence of elite groups; clearly, if present, these would have directed artisanal production in a different form. In this line of reasoning there also emerges another very frequent correspondence, namely between the periphery and economic subsistence. Here, by contrast, in discussing the findings made, we intend to set out from pottery to shed new light on the concepts of periphery and complexity, showing that periphery does not always mean subsistence and social simplification, but involves dynamics that are much more complex, also when we try to reconstruct them.57 First, it is important to introduce into the discussion the fact that pottery itself takes on meaning, which can be appraised also in social terms, if we place it in relation to other materials, whether they be objects (metal, glass etc), or particular practices 56 C. L. Costin, Craft Production… op. cit., p. 1035. 57 J. A. Quirós Castillo, ‘Dalla periferia: archeometallurgia del ferro nella Spagna nord-occidentale nell’alto e pieno Medioevo’, in A. Molinari, R. Santangeli Valenzani, L. Spera (eds), L’archeologia della produzione a Roma (secoli V–XV), Atti del Convegno Internazionale di Studi, Bari, 2016, pp. 597–613.

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

(food consumption) or techniques (other activities linked to fire) or written sources themselves.58 For this reason, the analysis of all relations between men and things takes on a meaning that can be interpreted historically by means of a parallel study of every type of link and material source.59 Unfortunately, this is not facilitated by the tradition of archaeological studies, which keeps every specialist separated in the study of their own class of material. In the spirit of this intention, namely to reconstruct the political and social relations in a given area, we will try to take as broad a view as possible, and we will go over the main findings made by correlating them to each other. In the first chronological period analysed, the sixth and seventh century, the central fact is that differing traditions of production and consumption existed side by side, accompanied by a substantial degree of instability at the political level, which leads to a contraction towards local resources, and to a certain degree of experimentation. The sphere of production, although the new findings come only from pottery and not from an investigation of production centres, can be described by the generalised use of the single wheel, a careful selection of local clays, and low variation in form and function. Products with differing characteristics — use of the double wheel/mould, use of slips — are typical of workshops outside the Alava region, and these products may have already been in use in pottery repertoires as of the previous century. Thus they would not bear witness to an active trading network, but to a phenomenon of second use, or maintenance of household pottery.60 What emerges with certainty from the present study is that, even as late as the sixth century, TSHT products continued to form part of household pottery, also in the area of the plain, as shown by the case of Aistra61 (fig. 19). In one case, Aldaieta, this is a minority presence, but the context from which the finds come must be taken into account, since this is closely linked to burials.62 As of the seventh century the contribution of these wares disappears entirely, and this involves a change in early medieval workshops, certainly in the amount of products made (which increased). On the other hand, as regards the variety of products, we

58 C. L. Costin, Craft Production… op. cit., pp. 1039–1040; B. Jervis, A Patchwork of people… op. cit. 59 B. Jervis, Pottery and Social Life in Medieval England, Oxford, 2014, pp. 33–70. 60 T. J. Peña, Roman Pottery in the Archaeological Record, Cambridge University Press, Cambridge, 2007, p. 209. The latest chronology currently accepted for the production of TSHT is ad 450, and as yet we do not have any new findings to depart from this date, such as continuity of production by these workshops after that date (A. Vigil-Escalera Guirado, Los primeros paisajes altomedievales en el interior de Hespania. Registros campesinos del siglo quinto d.C., Bilbao, 2015). 61 A different view is taken, especially for the hinterland, by A. Azkarate Garai-Olaun, J. Nuñez, J. L. Solaun Bustinza, Materiales y contextos cerámicos… op. cit., final pages, in which this model of coexistence is set out only for areas overlooking the Atlantic. 62 On the other hand, there are neighbouring contexts dated to the sixth century in which Late Antique pottery has apparently alreadytotally disappeared, such as Las Gobas (Treviño, Burgos) and San Martin (Dulantzi, Alava) A. Azkarate Garai-Olaun, J. Nuñez, J. L. Solaun Bustinza, Materiales y contextos cerámicos… op. cit.; A. Azkarate Garai-Olaun, J. L. Solaun Bustinza, La cerámica altomedieval … op cit., pp. 195–99.

123

1 24

f r a n ce s ca g r as s i

Fig. 19. TSHT bowl with a repair hole from the village of Zornoztegi, fifth century.

note that the pottery made in these workshops never catered for tableware, perhaps owing to changes in consumption models, as we will argue later on. Turning to comparisons with other findings, in the artisanal production of iron, for example, as of the seventh century we see a number of sites in the Alava area with production facilities for transformation and processing (such as Gasteiz and Bagoeta), situated far from metal-bearing ore veins, and thus marked by complexity not just as regards getting hold of supplies, but also as regards the production cycle itself. The consumption of iron objects in rural communities, including some sites such as Zaballa, also shows the use of multiple objects, and the common practice of repairing these objects (at Zornoztegi), revealing dense inter-regional trade.63 Thus, experimentation seems active not only in pottery, with the coexistence of differing traditions — medieval and Late Antique — but also in metallurgy. Going back to pottery, as regards consumption, there was thus, on the one hand, the prolonged use of Late Antique pottery, especially for tableware, and on the other hand, among the new products, the predominance of cooking pots with coarse

63 J. A. Quirós Castillo, Dalla periferia: archeometallurgia del ferro… op. cit., pp. 607–09.

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

fabrics that are local and devoid of embellishment. This domestic kit was a response partly to production needs, and partly to social and cultural needs, in a transition phase between a world that was no longer Roman and a “new” world. In the Basque Country we do not know how far this “new” world was influenced politically and culturally by the realm of the Visigoths. While fineware for the table seems a pure expression of the way of producing and consuming food at table that were typical of the Roman Empire, in our view the coarse cooking pots were the product of the form of pottery-making found in the local rural context, rooted in the local territory, and the use of its resources, and the serial repetition of a practical and functional piece of kitchen equipment, and thus the expression of a social identity that had not been entirely acculturated by the dynamics of the Empire.64 These coarse cooking pots are today the most important witness (also in quantitative terms) to the vessel-type produced by early medieval workshops in the Alava plain in the sixth and seventh century, as well as the only form present in the most important necropolis in the plain itself, the Aldaieta necropolis. In a sense, the pots in the grave goods in the Aldaieta burials display the same discovery provided by an analysis of the paleo-diet, namely the presence of interred burials from the local substrate.65 It is interesting to remember that, also in the late Imperial period, at least from the fourth century ad onwards, the manufacture of pottery in the Alava plain showed clear signs of a local tradition, which in our view is very clear from the discoveries made at the Roman city of Iruña Veleia, where no less than 49% of fine pottery for cooking was made without the aid of a wheel.66 Looking at the cooking pots of the fourth century ad found here, there are many similarities with those of the sixth century ad, albeit with differing fabrics, and thus coming from different production centres. Equally, in this case, we are to glimpse the absence of acculturation to the dynamics for identification with the Roman Empire, based also on objects in daily use, such as pottery. For distribution, we have noted that pottery circulated mainly within the Alava region, having suggested, on the basis of clays, that the production sites were located all in the same area. Thus, we have to conclude that the geographical area investigated

64 As regards pottery, social identity and acculturation, see the example of the eastern Mediterranean in the Crusader era, analysed by Vroom ( J. Vroom, ‘The Morea and its Links with Southern Italy after ad 1204: Ceramics and Identity’, in Archeologia Medievale XXXVIII, 2011, pp. 409–30). In the sixth century this phenomenon is seen in Spain also outside the area analysed, as in the area of Burgos (C. Alonso Fernandez, J. Jimenez Echevarria, El yacimiento arqueologico Alto del Mural / Camino de los Aguanares (Cogollos, Burgos), Burgos, 2010) or in the Visigoth villages of the Meseta Centrale (F. Grassi, A. Vigil Escalera-Guirado, La ceramica come indicatore di complessità economica e sociale… op. cit.), or in Galicia, in the area of Vigo (A. Fernandez Fernandez, R. Bartolomé Abraira, ‘Ceramicas tardoantiguas en el Noroeste de la Peninsula (Galicia y norte del Portugal): entre la importacion y el artesanato local/regional (ss. V–VII)’, in J. A. Quirós Castillo, A. Vigil Escalera Guirado (eds), La cerámica de la Alta Edad Media en el cuadrante noroeste de la Península Ibérica (siglos V–X). Sistemas de producción, mecanismos de distribución y patrones de consumo, Bilbao, 2016, pp. 69–111. 65 An analysis of diet itself appears to show that the individuals buried in the graves were characterized by a poor, plant-based diet, see J. A. Quirós Castillo, Ceremonial funds… op. cit. 66 A. Martinez Salcedo, La ceramica comun… op. cit., p. 396, figure 26.

125

1 26

f r a n ce s ca g r as s i

was very active as regards distribution, and well connected internally. Metal itself, in the specific case of iron, shows active trade at the regional level between the production areas (to the north, the area of Bilbao and Biscay) and transformation sites, not always nearby, and a very homogenous economic area.67 As of the eighth century, the phase of adaptation to the new fabric of sites, already active with the seventh century experimentations, seems to become consolidated, and at the same time initial contacts start to be opened up with neighbouring territories, such as the area of the Ebro and the county of Castile.68 In production, workshops consolidate the manufacture of increasingly specialized products, although with a limited formal repertoire. The production of cooking vessels with highly selected clays (fine coarseware), and the contemporary disappearance of coarseware with fabrics with larger inclusions, denote a careful choice of raw materials linked to the needs of consumption. There is also an increase in the amount of pottery that we find at sites, probably along with a growth in productive output. The use of raw materials from just three river areas, that of the Bayas, the Trias and the Zadorra, shows organization and management of deposits for extracting clay. Can we put forward a possible production model where workshops were also in some sense centralized? To date, archaeological research has not found sites that can be described as centralized production centres, and thus we must accept the fact that either this represents a lacuna in research itself, or else these sites did not exist. We can add that the information derivable from the products (being thus indirect information regarding the sphere of production) certainly shows that the necessary skills — namely the technical knowhow of the potters — were very homogenous throughout the territory examined. Thus, rather than look for a single physical place of production, we should think in terms of a collectivity of craftsmen working with the same skills across the entire territory, perhaps on a seasonal and/or itinerant basis. I believe it to be obvious that, having only indirect findings available, we cannot take our hypotheses further. The fact that the local micro-scale was very active as of the eighth century is also shown by pottery circulation: the same products present at all the sites, although in differing amounts. The development of a local, well-functioning communications network is undeniable, and almost all the trade and exchange occurs at the regional level. Were there intermediaries? What was the basis for the relationship between producer and consumer? If we pursue the hypothesis of itinerant craft-workers, the phase of exchange is eliminated by production in situ, at sites themselves. Otherwise, one may think in terms of intermediate centres for exchange and trade, but here, once again, we do not have archaeological evidence for this. The town of Iruña-Veleia was not an example of this, although we do not 67 J. A. Quirós Castillo, Dalla periferia: archeometallurgia del ferro… op. cit., p. 608. 68 From the historical point of view, part of this area, as of the ninth century, was contained within, or bordered on, the ancient County of Castile, and there are several mentions of this, written between 816 and ad 931, and the Counts came under the authority of the Kingdom of Asturias (C. Estepa Díez, ‘La Castilla primitiva (750–931): condes, territorios y villas’, Territorio, Sociedad y Poder, Anejo N. 2, 2009, pp. 261–78).

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

know, in these centuries, its social and political value, nor is the village of Gasteiz, which it is unlikely can be defined as a central place on the basis of an analysis of the pottery products from the archaeological excavation of the Cathedral area.69 Thus, in the absence of intermediate exchange centres, we would be looking at a direct market linking producer and consumer, and this would certainly have encouraged short-range, regional supply. Non-regional pottery from the Ebro valley (red-painted pottery) arrived as of the eighth century and in very low percentages (1% of the total), but, in this context of pottery production for the primary purposes of cooking and the domestic sphere, this figure stands out somewhat prominently. It would seem that this presence, found only in minimal amounts at the three sites of Gasteiz, Dulantzi and Zaballa, is to be interpreted as a social indicator of the ability to attract particular goods, and thus also a political ability to gain access to distribution and production networks associated with areas outside the immediate region. Furthermore, in Zaballa, the presence of pottery produced in the area of the Ebro is not only restricted to painted pottery: levigated, undecorated pottery is also present, with the same characteristics of fabric. Acquiring and consuming this pottery brings the communities closer to those of the Ebro area, belonging to the old Castile county, at least as of the ninth century, and thus designed to create a sense of belonging to this area of power, coming under the kingdom of the Asturias. Moving on to the ambit of specific consumption, the factors that emerge as of the eighth century once again include the lack of tableware, little variability in form, and the almost total absence of open vessels. As regards tableware, as already stated, the fineware from the Roman tradition would not be replaced by any new form, and this is an important indicator for stating that the custom of consuming food with particular vessels was introduced with the Roman Empire, and it disappeared with its demise, also due to the fact that, to make a certain type of pottery (such as Terra Sigillata), a totally different manufacturing system was needed, beginning with the type of production facilities themselves. Regarding the scant formal variability, which in a sense goes hand-in-hand with the lack of open vessels, it is hard to give an economic or social assessment. In our view, this factor is closely connected to diet, given that it is unthinkable that the pottery workshops in the Alava plain did not have the possibility of producing bowls (large and small), cooking pans, or flasks or lamps, made and used in other parts

69 On the basis of archeological excavations, the hill of Gasteiz, as of the end of the seventh century, was the site of a village which it has been suggested had a vocation for metallurgy, on the basis of the discovery of a forge (A. Azkarate Garai-Olaun, I., Garcia Gomez, J. L. Solaun Bustinza, El asentamiento altomedieval de Gasteiz… op. cit., pp. 399–446). As regards pottery, it is thought that the wares were in part locally-made, produced in the village itself, and in part the work of itinerant craft-workers, operating also at the regional level. The only imports from surrounding areas date to the eighth century, in the form of red-painted pottery from the Ebro valley (A. Azkarate Garai-Olaun, I., Garcia Gomez, J. L. Solaun Bustinza, El asentamiento altomedieval de Gasteiz op. cit., p. 417). Thus, a type of consumption similar to that which can be outlined for Zaballa.

127

1 28

f r a n ce s ca g r as s i

of Spain in the eighth and ninth century.70 What we are analysing is an extremely functional repertoire which meets criteria of use, and thus production was based on local dietary models, reconstructed also thanks to analyses of the paleo-diet of the people buried in the cemeteries of Zornoztegi, Aistra, Zaballa and Dulantzi.71 These dietary models, aside from what people actually ate, envisaged the following pottery vessels: high-sided pots for cooking (also in line with the abundant use of plants); and jugs and jars to contain and preserve liquids and cereals. There was also an absence of clay vessels for the consumption of food, a function perhaps met by metal or wood. This model is found equally in all the communities, that seem to be identified with these practices in a homogenous way, thereby reinforcing their concept of their own identity by means of the domestic material culture.72 At the same time, the choice of these eating habits also seems to unite all the social groups analysed, reinforcing the identity of the communities that we are describing through pottery. We can suggest that, at this historical moment, from the sixth century, it was particularly important for the population to discover, or rediscover, their identity, also engaging in practices of social recognition by means of household objects.

Conclusions In this study, our aim was to use pottery to describe the complexities of a territory in its historical dimension, looking at social and political aspects. Having reached the end, we can conclude that pottery has undoubtedly proved to be an excellent and formidable material indicator, allowing the emergence, in this specific geographical area, of many political and social complexities, and a mosaic of differing realities that was previously invisible. However, in our view, the contribution made by this holistic study of pottery, or relational study, is much more wide-ranging.73 By means of the relations that are created between an object itself and people, pottery is to be seen as a real window onto ancient society, in the relations of each object with people, and not just a material indicator on a par with documents, the landscape, or buildings. However, to achieve this objective, research cannot move in one direction only. Only by addressing all the directions that an object points us in, meaning the routes of production, distribution and consumption, can we 70 For example in Madrid, Toledo and Guadalajara, see E. Serrano Herrero, M. Torra Perez, R. Catalan Ramos, A. Vigil Escalera Guirado, ‘La ceramica de los siglos VIII–IX en Madrid, Toledo y Guadalajara’, in J. A. Quirós Castillo, A. Vigil Escalera Guirado (eds), La cerámica de la Alta Edad Media en el cuadrante noroeste de la Península Ibérica (siglos V–X). Sistemas de producción, mecanismos de distribución y patrones de consumo, Bilbao, 2016, pp. 279–313. 71 J. A. Quirós Castillo, P. Ricci, C. Sirignano, C. Lubritto, ‘Paleodieta e società rurali altomedievali dei Paesi Baschi alla luce dei marcatori isotopici di C e N (secoli V–XI)’, in Archeologia Medievale, XXXIX, 2012, pp. 87–92. 72 B. J. Mills, ‘Communities of Consumption. Cuisines as Constellated Networks of Situated Practices’, in A. Roddick, A. Stahl (eds), Knowledge in Motion: Constellations of Learning Across Time and Place, Arizona, 2016, pp. 247–70; B. Jervis, Pottery and Social Life… op. cit., pp. 70–103. 73 In this connection, see also B. Jervis, Pottery and Social Life… op. cit., pp. 131–54.

S o c i al Co m p l e xi t y i n Pe ri phe ral Are as

expand the range of our interpretative possibilities, and finally converge on an overall reconstruction. Had we only looked at the presence of long-distance goods, the peripheral area of Alava would have appeared isolated and closed off. Instead, the combined study of distribution and consumption, in other words the analysis of what was circulating and what each site actually bought and used, has revealed a social context which wholly describes to us early medieval communities. This economic and social dimension also brings with it a political aspect, and at some of these sites, by means of the index of differentiated consumption, it has been possible to hypothesize the presence of local control, while at other sites this local political level did not appear. The study of production itself has shown that some goods can express an economic strength on the part of a certain area, for example red-painted pottery in the area of the Ebro, and that they can have a social value, not just an economic value, wherever they happen to be present at a great distance from the areas where they were made. Each sign, each trace, and each clue has a value only if contextualized, geographically and politically. In the social dimension, pottery has, on the one hand, displayed an important role in maintaining community identities, in the same way as diet, and on the other hand its use is an attempt to identify with the emerging political powers. Indeed, the political situation of great instability, and the fact of having focused attention on a peripheral, border area, in the sense illustrated in our introductory section, are what have allowed us to completely use all these possibilities which pottery offers us, and to put them to the test. Obviously, the final prompting that stems from this research is to address the reconstruction of ancient societies with an approach that is as inter-disciplinary as possible, in which each specialist, starting with ceramics experts themselves, has an awareness that, in order to describe a historical era, it is necessary to always take a very broad view.

129

Edith Pey tremann

Indications of an Estate Economy from a Renewed Analysis of Sites of Rural Settlements

Interestingly, the question of peasant production patterns in the early Middle Ages involves specifically historians using written sources and drawing inspiration from models developed by economists.1 Attempting to respond to them from archaeological data, it is straightforward to admit that these production methods, whatever they may have been, are likely to have left material traces that can be detected by the archaeology. However, while the production is at least are partially discernible through archaeology, the status of the producer is more difficult to grasp from the vestiges alone. The same is true for the identification of possible surpluses, even though quantification methods make it possible to distinguish productions that exceed the needs of a habitat, especially in the case of metallurgy2 or, more rarely, pottery production. Archaeology can, however, provide information on the framework in which the goods were produced and their nature. It is now about thirty years that rural sites have been the subject of archaeological excavations, often as part of preventive archaeology. A novelty in the 1980s, this type of site has become commonplace. When the term “rural settlement of the early Middle Ages” is mentioned, it immediately evoked a picture of sunken-featured buildings, posthole buildings, pits, silos and sometimes ovens and wells. The discovery of one or more of these structures is generally sufficient for the site to be categorised as a rural settlement, precluding any further analysis, especially from a socio-economic perspective on production frameworks in the countryside. Therefore, this contribution aims first of all and try to define the nature of the rural sites excavated, then to focus more specifically on a type of site, the specific activity areas from an example at the



1 Witold Kula, Théorie économique du système féodal: pour un modèle de l’économie polonaise 16e–18e siècles, Paris, 1970; Jean-Pierre Devroey, Économie rurale et société dans l’Europe franque (vie-ixe siècles).T 1 Fondements matériels, échanges et lien social, Paris 2003. 2 Vincent Serneels, ‘La chaîne opératoire de la sidérurgie du fer’, in Recherches sur l’économie du fer en Méditerranée nord-occidentale, ed. by Michel Feugère and Vincent Serneels, Monographie Instrumentum 4, Montagnac, 1998, pp. 7–44. Edith Peytremann  •  Inrap-UMR6273 CRAHAM Caen Social Inequality in Early Medieval Europe: Local Societies and Beyond, ed. by Juan Antonio Quirós Castillo, Turnhout : Brepols, 2020 (HAMA 39), pp. 131–148 © FHG10.1484/M.HAMA-EB.5.118448

13 2

e di th p e ytr e m an n

site of Sermersheim (Alsace, France), as well as a few others in France and Germany. Finally, an interpretative reading from a socio-economic perspective will be proposed.

Definitions of Archaeological Situations Before defining a number of archaeological situations, it is necessary to understand the contribution of the archaeological source, to define the archaeology. For this we use the definition the archaeology offered by a French archaeologist, Laurent Olivier:3 “The purpose of archaeology − of all archaeology in fact − is not the past, but the vestiges of the past which are preserved in the present time. Archaeology does not deal so much with what happened in the past, as history aims to do that, but rather about what happened to material things, how they were affected by the events of the past, each in their own particular situation and configuration. Archaeology is a discipline of the memory of material things. It does not say the same things about the past as history; in way it does not speak the same language. To look at the past, the archaeology begins from the present, in which all the past of humanity is physically preserved.” This definition allows us to remind ourselves that the intellectual construction that follows and the hypothesis that will be proposed have been constructed solely on the basis of what has been preserved in our present and, of course, from the cultural perspective of the archaeologist who proposes this construction. The large corpus of rural sites currently compiled for the northern half of France shows their compositional and topographical diversity. It is mainly these two elements that make it possible to distinguish a number of entities. One of the first entities defined by archaeologists working on rural sites is that of a farmstead. This is the group of buildings needed by a farmer to live in and to run the farm operation. This building complex covers an area ranging from 240 m2 to 15,000 m2. Its composition varies in the number of buildings and facilities and can sometimes accommodate burials. It may or may not be defined in area by enclosures or ditches. Its organisation is also variable: around a courtyard, perpendicular to an axis of movement, bipartite (residential part/agricultural part) or without any discernible organisation.4 The social level of the farming unit, which can be estimated according to its size, the archaeological finds and the quality of the construction,





3 Laurent Olivier, ‘I Can Get No Satisfaction. Pour Une archéologie du passé contemporain’, in Clashes of Times.The contemporary Past as a Challenge for Archaeology Actes Du Colloque Du Musée de La Cour d’Or, Metz, 29–31 Oct. 2014, ed. by Jean-Marie Blaising, Jan Driessen, Jean-Pierre Legendre and Laurent Olivier (presented at ‘Un passé factice ? L’étude des vestiges contemporains comme outil critique de l’archéologie’, Louvain-la-Neuve, 2017, pp. 11–22). 4 Édith Peytremann, ‘Unité d’habitation, unité d’exploitation agricole : l’unité de référence dans l’habitat rural carolingien de la moitié septentrionale de la France ?’, in Cinquante années d’études médiévales. À la confluence de nos disciplines. Actes du colloque organisé à l’occasion du cinquantenaire du CESCM. Poitiers, 1er-4 Septembre 2003, ed. by Claude Arrignon and others, Turnhout, 2006), pp. 481–94.

Indications of an Estate Economy from a Renewed Analysis of Sites

also vary. However, it is has not been possible until now to establish a hierarchical scale to, for example, distinguish the elite from a prosperous peasant or a prosperous peasant from a more modest one. Estimating the amount of land that belonged to a farmstead has largely been inaccessible to the archaeologist.5 At one site, the number of farm units varies from few to a dozen. Obviously, the composition, topography and social level vary according to the chronology. Although for a long time the farm unit was considered to be the main entity in the composition of a rural settlement, since at least the mid-1990s, other entities have been defined, but without being studied carefully.6 The specific activity areas is a group of constructions and facilities intended for agricultural activities or/and craftwork. Their sizes vary from 350 m² to 30,000 m². The most frequent structures are the silos, often organised in batteries, pits, and, depending on the regions, ovens, sunken-featured buildings in large numbers and wells. More rarely, some areas include post constructions such as granaries or outbuildings. In rare cases, burials or groups of burials are also present.7 The most common activities identified from structures and archaeological finds are silage-making, cereal-processing (thermal treatment of the grain before threshing and threshing), textile-making and baking. Certain areas can accumulate many activities.

The Specific Activitiy Areas The Example of Sermersheim (Bas-Rhin-Alsace)

The village of Sermersheim is located on the Alsace Plain, on the left bank of the river Ill, about thirty kilometres south-west of Strasbourg. The 1.73-hectares excavated area is located 400 m east of the centre of the present-day village is marked by a church dating from the nineteenth century, although it was rebuilt on the site of an older structure, perhaps from the eleventh century. The site is located on the edge of two geological formations (fig. 1): to the east, the floodplain of the River Ill with alluvial deposits, to the west the beginning of the loess terrace. The archaeological structures are mostly located on this terrace. This site is particularly characterised by the extreme density of its occupation between the Neolithic and the twelfth century without there being any confirmed continuity. In the absence of an associated dating element, almost three quarters of the structures (73%) could not be directly dated. The main occupation of the excavated area began





5 Archaeogeographic and historical studies can sometimes make it possible to reconstruct the boundaries of a domain and its associated lands. 6 Édith Peytremann, ‘Pour une approche renouvelée des habitats ruraux du haut Moyen Âge dans la moitié nord de la France’, in Pratiques de l’espace. Archéologie et histoire des territoires médiévaux. Actes de la table ronde organisée à l’occasion du cinquantenaire du CRAHAM, Caen 29 Février-1er Mars 2008, Caen, forthcoming. 7 Édith Peytremann, ‘Pour une approche renouvelée des habitats ruraux du haut moyen âge dans la moitié nord de la France’.

133

986 850

986 800

986 750

e di th p e ytr e m an n

85 100

85 100

N E

O S

2310

B1

85 050

85 050

B

St 2377 St

2505

B

A

85 000

85 000

St 3080

B

1564B

St 3079

1500A

1599B

A

1596A

3037

A

84 950

84 950

St 3107 St 3106

St 1249

84 900

84 900

Early and high medieval structures

C

1854

Gallo-Roman structures

1093

1861

Néolithic and protohistoric structures Néolithic structures Undated structures Unexcavated structures Geomorphological soundings

84 850

986 850

986 800

84 850

986 750

134

0

20 m

Fig. 1. Diachronic plan of the Sermersheim (Bas-Rhin) site. Topographic measurements: N. Saadi and J.-L. Wuttmann, CAD: P. Girard and É. Peytremann.

Indications of an Estate Economy from a Renewed Analysis of Sites N E

O

A1

B1

S

C1

B2

A2

C2

A3

B3 C3

B B4 A4

C4 A5

HMA 1

HMA 2

HMA 3 0

Légend : well

burial well dated from early Middle Ages, unspecified phase structure dated from early Ages, unspecified phase

délimiting of sets according to dates strutures ditch

lœss terrace

unmeasured structure located structure

50 m

A1

floodplain numbering of sets

Fig. 2. Topographical development of the occupation of the early Middle Ages from the plans of localization of the dated structures. CAD: P. Girard and É. Peytremann.

in the sixth century ad and ended in the twelfth century. Thanks to joint studies of the stratigraphy and pottery, as well as dendrochronological and radiocarbon dating, it has been possible to distinguish three phases of occupation that overlap due to the diversity of dating elements. The first phase (HMA 1) extends from the beginning of the sixth to the beginning of the eighth centuries. The second (HMA 2) begins at the end of the seventh century and ends in the tenth century. The final phase (HMA 3) is from the beginning of the tenth century to the twelfth century (fig. 2). For each of these phases, four to five topographic sets were isolated from the topographic distribution of the vestiges. Their surface area varies from 1500 to 3265 m². Each of these complexes includes sunken-featured buildings (fig. 3), pits, silos, sometimes

135

1 36

e di th p e ytr e m an n

Fig. 3. Example of feature sunken building with two post-holes.

granaries or a construction on posts or sill beam and water wells. These attest several construction techniques. The most frequent technique is the use of a hollow oak trunk of centuries-long (fig. 4a), but the techniques of corner-timbering (Figure 04b) or with staves (Figure 04c) were also used. One of the singularities of this site is the large number of wells for the early and late Middle Ages, as no less than forty-six have been discovered. The number of vestiges varies according to the phases; they are the most abundant during Phase 2 (HMA 2: end of the seventh to the tenth centuries). It is also during this phase that a funeral area consisting of two groups accounting for a minimum of seventy-two burials is in service in Complex B2 (fig. 2). These two burial groups are located at the western end of the buildings and are bordered by a ditch that lasts throughout during the medieval occupation. Unfortunately, the boundaries of the excavation did not make it possible to explore more precisely whether there were remains beyond this ditch. Some burials scattered in the extent of the excavation, isolated or in clusters, also date from this phase. The eastern boundary of the site corresponds more or less to the natural boundary formed by the loess terrace. The northern and southern limits of the site have not been reached. Thanks to the numerous wells, it was possible to collect a large quantity of moist samples mainly intended for carpological analysis. In addition to climatic data, the study of the wood from the well casings has yielded no less than nineteen dendrochronological dates. A complete study of fauna, archaeological finds and metallurgical remains was also carried out, thus offering a relatively comprehensive approach to the site. Several arguments derived from the analysis of the excavation data make it possible to interpret this site as a specific activities zone devoid of residential functions. The two main arguments put forward for this interpretation correspond to the lack of any building that could have served as a dwelling and the nature of the waste. The archaeozoological study showed that the faunal remains were

Indications of an Estate Economy from a Renewed Analysis of Sites

a

b

c Fig. 4. Wells a: made with a hollow oak trunk, b: made with corner-timbering, c: made with staves.

137

138

e di th p e ytr e m an n

Fig. 5. Clay loom weights of Sermersheim (Bas-Rhin), © F. Schneikert, Inrap.

mostly large sizes, which usually indicates a remoteness of the dwelling houses.8 On the other hand, pottery not very abundant and absent from the well fillings, which tends to indicate remoteness from the habitat. The analysis of the structures and the archaeological finds also made it possible to recognize several craft and agricultural activities at the site. Firstly, a textile activity was identified, thanks not only to sunken-featured buildings and the importance of the clay loom weights (fig. 5) but also to the carpological analysis which attests the cultivation of hemp, flax and the presence of dye plants. The abundance of wells at the site can be linked to this textile activity, which would have required water for fibre washing, fibre moistening, dyeing and bleaching of the web. Secondly, metallurgical forging has been recognised from the abundance of slag, scaling and objects being manufactured. The palaeometallurgical study9 showed that it was a recycling forge and manufacturing of small objects using at least two different types of hearth. This craftwork would also have required having water nearby. Among the agricultural activities, animal husbandry appears to have



8 Gaetan Jouanin and Jean-Hervé Yvinec, ‘L’élevage : étude archéozoologique’, in En marge du village : la zone d’activités spécifiques et les groupes funéraires de Sermersheim (Bas-Rhin) Du vie au xiie siècle, by Édith Peytremann, Supplément à la RAE, Revue archéologique de l’Est, Dijon, 2018, pp. 263–72. 9 Marc Leroy and Paul Merluzzo, ‘Activité métallurgique : étude des déchets paléométallurgiques de l’habitat du haut Moyen Âge de Sermersheim “Hintere Buen” (Bas-Rhin)’, in En marge du village : la zone d’activités spécifiques et les groupes funéraires de Sermersheim (Bas-Rhin) Du vie Au xiie siècle, by Édith Peytremann, Supplément à la Revue archéologique de l’Est, Dijon, 2018, pp. 201–15.

Indications of an Estate Economy from a Renewed Analysis of Sites

been orientated towards cattle and equidae. Cultivated crops did not only include wheat, fruit trees10 and vines, but also condimental and/or medicinal and textile plants. The final recognised agricultural activity recognised is the storage of cereals in silos and granaries. It is likely the cereals were processed prior to storage, but it this has not been clearly identified. The Sermersheim site thus appears to correspond to a specific activities zone that can be described as mixed in the sense that both agricultural and craft work was carried out there. A Few Other Examples in France and Germany

A review of the corpus of rural sites11 shows that some have or correspond to specific activity zones. Two types of activities zones can be distinguished depending on the size, the activities and the topographical situation. The first are small areas ( 5000 m²), which can comprise sunken-featured buildings, storage structures, pits and sometimes wells or ovens and extend over a large area. They are usually at the fringes of the settlement. For example, such areas have been identified on the Marlenheim “Hoffstatt” site in the Grand-Est, Marines “Les Carreaux” in Ile-de-France or Wehringen in Germany (Bavaria). Marlenheim’s specific activities sector, dating from the eighth to twelfth centuries, measures almost 8,000 m² (fig. 7). It includes a hundred sunken-featured buildings,

10 The tree peach cultivation is notably attested at the end of the seventh century. 11 The complete takeover of the corpus to identify specific activities areas is entirely to be done. Here we present some of the best-known examples. 12 Edith Peytremann, ‘L’habitat déserté de Gungling à Grosbliederstroff (Moselle) IXe-début XVIe Siècle’, Archéologie Médiévale, 2006, pp. 57–113. 13 Nathalie Buchez, ‘Un habitat du haut Moyen Âge à Bussy-Saint-Georges, Les Dix-Huit Arpents (Seine-et-Marne)’, in L’habitat rural du haut Moyen Âge (France, Pays-Bas, Danemark et GrandeBretagne), Actes Des XIVe Journées internationales d’Archéologie mérovingienne, Guiry-en-Vexin et Paris, 4–8 février 1993, ed. by Claude Lorren and Patrick Périn, Mémoires de l’Association française d’archéologie mérovingienne, t. VI, Rouen, 1995, pp. 109–12.

1 39

e di th p e ytr e m an n

N Burials

Specific activities area

Potter’s oven

Burials

Burials

Grosbliederstroff-1998 Gungling

I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I

I I I I I

I

I

I

I

I

I II

I

I

I

I

I I I I I I I I I I I I I I I I I I II I

modern sandpit

II I I I I I I

I

I

I

I

roasting structures, earth

I I I I I I I I I I I I I I

140

0

10

20 m

Fig. 6. Location of the specific activities area on the plan of the Gungling site at Grosbliederstroff (Moselle) in the ninth to twelth century. CAD: J. Dolata and É. Peytremann.

Indications of an Estate Economy from a Renewed Analysis of Sites

Residential plots (6th-7th century) 0

domestic and craft structure burial

25 m

Specific activities area (8th - 12th century)

Fig. 7. Temporary plan of sixth–twelfth /thirteenth century structures of the Marlenheim ‘Hofstatt’.

pits and a well. The activities carried out in this sector were mainly to do with textiles (spinning, weaving and sewing) recognised thanks to the discovery of spindle whorls, loom weights and needles, metallurgy (iron ingots, slag, tools, furnace walls, etc.) and possibly brickworks.14 The presence of certain building materials (painted plaster, window glass, terracotta slabs, etc.) in the discharge position suggests the proximity of luxury residential buildings. The specific activities area of Marines “Les Carreaux”, dating from the end of the seventh to the end of the ninth centuries, covers at least 25,000 m². It has about twenty sunken-featured buildings, about forty ovens, as well as pits and silos. It has the peculiarity of including some twenty burials. The activities recognised in this sector

14 Madeleine Châtelet, ‘Marlenheim en Alsace : Une résidence royale et un centre domanial des périodes mérovingienne et carolingienne’, in Des fleuves et des hommes à l’époque mérovingienne : territoire fluvial et société au premier Moyen Âge (ve-xiie siècle), ed. by Édith Peytremann, Mémoires de l’association française d’Archéologie mérovingienne, T. XXXII ; Supplément à la RAE, 42 (SaintGermain-en-Laye: AFAM, 2016), pp. 245–54.

141

142

e di th p e ytr e m an n

correspond to cereal storage, processing (milling) and baking.15 As at the previous site, the presence of finds such as a spur, a glass window and door frames, suggests the possibility of a residential building of a certain standing nearby. The Wehringen site in Germany, excavated over one hectare, is likely to include a specific activities area with about one hundred sunken-featured buildings, six wells and pits and silos.16 Textile activities (spinning, loom and tablet weaving), metallurgy (a forge) and leather work were identified. Two burials are associated with this area. Examples of specific activities areas17 of a significant size are relatively numerous in the northern half of France.18

Interpretation In this section, we will examine the significance of these specific activities zones in the economic context of the eighth to the eleventh centuries. For this, the Sermersheim activities area will be used as the main thread for the interpretation. The historical context of the Sermersheim site, based on the written sources, allows us to make a number of deductions. Sermersheim is mentioned several times in documents dating to between the second half of the seventh century and the end of the ninth century and then in documents attributed to the tenth to the thirteenth centuries.19 From these different references it can be deduced that the abbeys of Ebersmunster and Hohenburg, situated respectively 7 and 30 km from Sermersheim, organised their property in Sermersheim in the form of estates (curtis dominica) with a church. In the tenth century, another estate belonging to Seltz Abbey (located

15 Christophe Devals, ‘Un site du haut Moyen Âge du Vexin français. Marines — Les Carreaux (Vald’Oise)’, in Villes et Campagnes En Neustrie. Sociétés — Économies — Territoires — Christianisation, Actes des XXVe Journées internationales d’Archéologie mérovingienne, Tournai, 17–20 Juin 2004, ed. by Laurent Verslype (Montagnac: Editions Monique Mergoil, 2007), pp. 138–55. 16 V. Babucke, ‘Grubenhaus und Brettchenweber. Das frühmittelalterliche Handwerkerdorf von Wehringen’, in Grubenhaus und Brettchenweber. Archäologische Entdeckungen in Wehringen, by V. Babucke and others, Archäologie in Bayerisch-Schwaben 1, Friedberg, 2005, pp. 43–57. 17 Among those published, it can be mentioned Tournedos-sur-Seine in Normandy (Florence Carré and others, ‘The Rural Site of Porte-Joie (Tournedos-Sur-Seine / Val de Reuil, Eure, France). Des espaces particuliers au sein de l’habitat rural du haut Moyen Âge’, in On the Road again. Thème 2: Archaeology and Rural Landscape: Rural settlements in Their Natural, Economical and Social Environment, Proceedings of the Fourth International Congress of Medieval and Modern Archeology, Medieval Europe, (Paris: 3–8 September 2007), by Isabelle Catteddu, Paolo De Vingo and Anne NissenJaubert, Genoa, 2011, pp. 103–23) and Torcy in the Grand-Est (Murielle Georges-Leroy, ‘Torcy-LePetit (Aube),’ The Way of Arcis’, in The Archaeological Excavations on the Route of the Highway A 26 Châlons-Sur-Marne-Troyes, Mémoires de la Société d’Agriculture, Commerce, Sciences et Art de la Marne 110, Châlon-en-Champagne, 1995, pp. 115–21). 18 Edith Peytremann, En marge du village : La zone d’activités spécifiques et les groupes funéraires de Sermersheim (Bas-Rhin) du vie Au xiie siècle, Supplément à la Revue archéologique de l’Est, Dijon, 2018, pp 294. 19 Edith Peytremann, En marge du village, cit.

Indications of an Estate Economy from a Renewed Analysis of Sites

about 80 km north of Alsace) is also attested at Sermersheim. Moreover, at least three chapels are mentioned in the seventeenth century in the village. During the period from the sixth to the beginning of the eighth centuries, at least four sunken-featured buildings out of the eleven attributed to this period were interpreted as textile workshops, insofar as they were associated with loom weights. The workshops are divided into three sets and are associated with wells. In the following period, the number of workshops increased to at least eight. In total, thirty-one sunken-featured buildings were counted for this period. The number of workshops increased to seven in the final period (tenth-twelfth centuries). These figures should be taken as estimates made solely on the basis of the preserved sunken-featured buildings dated and associated with loom weights. By generations, taking approximately 40 years as the life of a sunken-featured building,20 we obtain the minimum number of 1 (almost 2). This does not take into account undated sunken-featured buildings or those without loom weights. Here, it is indeed the concentration of workshops in the same place and over a long time that leads to the question and consideration of the existence of a textile production unit depending on an estate. This hypothesis is valid if it is admitted that the remains discovered belong to one of the “estate centres” (curtis dominica) of one of the first two abbeys mentioned above. In practical terms, if the number of women who can work in these workshops is estimated to be one or two,21 based on the size of the huts, a number of eight to sixteen is obtained, comparable to the figures given in the texts.22 This interpretation of an estate textile production unit leads to more caution in the identification of the excavated sector at Sermersheim as, at least from the end of the seventh century, the agricultural and craft zone of an estate centre. There are other arguments in favour of estate management, starting with the longevity of the economic functions of the sector — nearly five centuries. The archaeozoological study also showed a great stability in the choice of livestock, with particularly a strong equine presence. This stability can also be found in the textile activity, which was undertaken for five centuries, using the same weaving technique with a vertical warp-weighted loom, while the use of the vertical loom with a double beam is attested, for example, at Île-de-France in sunken-featured buildings dependent on a ninth-century farmstead.23 The crops — vineyards, wheat, peaches, condimental and/or medicinal plants — are comparable to those mentioned in both the Villis 20 Maruska Federici-Schenardi and Robert Fellner, Develier-Courtételle, un habitat rural mérovingien. 1, Structures et matériaux de construction, Porrentruy, 2004, p. 242. 21 P. Grimm, ‘Zwei Bemerkenswerte Gebäude in der Pfalz Tilleda. Eine zweite Tuchmacherei’, Prähistorische Zeitschrift, 1963, pp. 62–82; S. Dufrenne, Les illustrations du psautier d’Utrecht. Sources et apport carolingien, Paris, 1978. 22 Jean-Pierre Devroey, ‘Femmes au miroir des polyptyques : une approche des rapports du couple dans l’exploitation rurale dépendante entre Seine et Rhin Au ixe siècle’, in Femmes et pouvoirs des Femmes à Byzance et en Occident (VIe-XIe siècle), ed. by Stéphane Lebecq and others, Villeneuce d’Ascq, 1999. 23 François Gentili, ‘Typologie et analyse spatiale des fonds de cabane du haut Moyen Âge en Pays de France. L’exemple de Villiers-Le-Sec et des sites environnants : Baillet-En-France, Le Mesnil-Aubry, Louvres et Villiers-Le-Bel’, in L’habitat rural du haut Moyen Âge en Île-de-France. Programme Collectif de Recherche. Bilan 2004–2006, by François Gentili and Annie Lefèvre, Supplément au Bulletin Archéologique du Vexin Français et du Val-d’Oise 2, Guiry-en-Vexin, 2009, pp. 31–59.

143

144

e di th p e ytr e m an n

Fig. 8. Indented sickle of Sermersheim. © F. Schneickert, Inrap.

Capitulary24 and the Brevium exempla.25 The introduction of new tools reflecting development, such as the indented sickle (fig. 8), appears to be a feature of estate investment. This type of tool, whose morphology reflects a sawing rather than a cutting gesture, is indeed rare in France for the period between the tenth and the twelfth centuries and the known examples from the sites of La Grande Paroisse in 24 Élisabeth Magnou-Nortier, ‘Capitulaire ‘De Villis et Curtis Imperialibus’ (Vers 810–13) Texte, traduction et commentaire’, Revue Historique, 1998, pp. 643-90. 25 A. Boretius, ‘Brevium Exempla Ad describendas res ecclesiasticas et fiscales, Env. 810’, in Monumenta Germaniae Historica. Legum Section II, Capitularia Regum Francorum T. I, Hanovre, 1883, pp. 250–56.

Indications of an Estate Economy from a Renewed Analysis of Sites

Ile-de-France26 or Pinheuil in New-Aquitaine27 are smaller than those of Sermersheim. The existence of a manufacturing and not just a recycling forge is still characteristic of an estate infrastructure. The large number of wells built using centuries-old trees indicates the possibility of access to forest resources, particularly primary forest, which are generally attached to an estate. Finally, the construction and maintenance of weaving workshops, wells and silos shows an ability to invest in production and storage tools, to maintain them and to deploy the labour necessary for infrastructure construction and the production. In contrast, in the light of archaeological discoveries, it is possible to suggest a collective economic sector that was part of a settlement. Although this hypothesis can be effectively formulated, there are several arguments that tend to contradict it. One of the first concerns the decision to allocate land for the pooling of a number of economic functions over a period of five centuries. Furthermore, horse-breeding is out the reach of small peasant operations, as is the right to logging. There is also no need to group the weavers of the village in the same place and no more recent example is known, in contrast to the threshing, storage and milling activities that can be carried out collectively in the same place. Moreover, the importance of the textile production revealed by the archaeological remains presupposes a network for the diffusion of production, which is probably more important than a local market accessible to peasants. To think of such a stable collective management over five centuries by a village community which, during the period concerned, had no legal status, is virtually impossible or rather a Marxist ideal. Unless we challenge the proposed interpretation of the archaeological remains as vestiges of a specific activities area and see it as a settlement… If we preserve the hypothesis of an economic sector dependent on an estate centre, the question of its materiality, composition and organisation arises in the light of archaeological discoveries. In the excavated sector at Sermersheim, no proprietor’s residence was discovered and no element of material culture (pottery, fauna, metal, etc.) that would reveal the presence of an elite. The proximity of at least two abbeys perhaps makes the presence of an estate residence unnecessary. It may have been on the other side of the ditch, in the present-day village, but where are the houses of the people who worked in the area? If the hypothesis that it was part of an estate centre is maintained, it is interesting to note that the different co-existing activities resulted in a mix in the work area of textile production unit, a storage area and a forge in the same sector. The dead are accommodated next to the living, also without it being possible from the archaeo-anthropological study to identify the population buried on the boundary of the activities zone. There is nothing to distinguish the population buried at Sermersheim from the contemporary regional population in the cemeteries located near a church. It is

26 Michel Petit, L’habitat carolingien des Sureaux à La Grande-Paroisse (Seine-et-Marne). Une communauté villageoise à l’aube de l’an Mil, Nemours, 2009, p. 128. 27 Nicolas Portet, ‘Le mobilier métallique’, in Pineuilh (33) ‘La Mothe’, Rapport Final d’Opération, by Frédéric Prodéo, Pessac, 2007, pp. 725–79 (pp. 736–38).

145

146

e di th p e ytr e m an n

therefore not possible to speak of slaves in this population, as this legal status is not identifiable from bones. The abandonment of the sector is probably linked to the cessation of textile production in the area when there was a shift towards domestic production. This would have meant a modification in the production framework in the sense that the weaver worked at home, rather than in workshops dependent on the estate. Looking again at the above mentioned examples of specialised activities zones for which there are generally few or no studies or written documentation, with the exception of the Marlenheim site, it is possible to very cautiously link the sites of Marlenheim, Tournedos-sur-Seine, Marines and, to a lesser extent Wehringen, to part of an estate on which the economic activities were animal husbandry, farming and crafts. The main arguments in favour of estate management are the large size of the activities zone, which contains a large number of identical structures (sunken-featured buildings, ovens, silos, etc.); the presence, in the form of waste, of architectural elements indicating the proximity of a high-status building (window glass, a door frame, painted plaster, tile fragments, lime mortar, etc.), and a link, according to the written documentation, with a palace such as Marlenheim, a royal villa such as Tournedos-sur-Seine, an estate centre such as Torcy-le-Petit or, more imprecisely, an abbey or a territory dependent on the tax authorities.28

Conclusion This reading of rural domestic sites offers new avenues of reflection that touch both archaeology and historical interpretation. From an archaeological point of view, it is time to scrutinise more carefully the excavated rural sites from the early Middle Ages and to display their singularity. Although all the sites are classified as “rural settlements”, it is clear that, for example, those of Montours in Brittany,29 Villiers-le-Sec in Île de France30 and Sermersheim are very different from each other, even though two of them have sunken-featured buildings and three post-hole constructions. It is this difference that makes sense and can help us to better understand the organisation of the countryside. As has been seen, the identification of specific activities areas raises many questions, not least of which is the significance of a place in which identical structures (pits, silos, sunken-featured buildings, ovens, wells, etc.), generally linked to an agricultural and/ or craft activity, were installed, frequently over long periods. The structures generally 28 Edith Peytremann, En marge du Village, cit. 29 Isabelle Catteddu, Les habitats carolingiens de Montours et La Chapelle-Saint-Aubert (Ille-et-Vilaine), Paris, 2001. 30 François Gentili, ‘L’organisation spatiale des habitats ruraux du haut Moyen Âge : L’apport des grandes fouilles préventives. Deux exemples franciliens : Serris “Les Ruelles” (Seine-et-Marne) et Villiers-le-Sec (Val-d’Oise)’, in Trente Ans d’archéologie médiévale en France. Un bilan pour un avenir, ed. by Jean Chapelot, Caen, 2010), pp. 122–26.

Indications of an Estate Economy from a Renewed Analysis of Sites

do not have any originality and can also be found in small numbers or standing alone within the farming units. What is intriguing is the phenomenon of accumulation, which refers to activities and lasting topography that contrasts with the dynamic observed for the early Middle Ages in the present-day villages which shows regular changes in the allocation of spaces.31 This difference in the rate of change or mobility of spaces also corresponds almost certainly to the management of economic activities in different settings. We can also wonder about the significance of specific small activities areas in relation to large areas. Do the specific small activities areas indicate smallholders who are enough rich to invest in a number of infrastructures, mainly agricultural: grain processing and storage or do they reflect the presence of a village community that pools its mainly agricultural infrastructure on communal land, as suggested by C. Loveluck for the Saleux site.32 Do the large specific activities areas belong, as has been previously mentioned, to a large property-owner who concentrates in the one place, a number of agricultural and craft activities connected to the functioning of their estate? Could the importance of the specific activities zone be the quantitative reflection of the production, as well as of an economic type that would have been adapted to the needs of the rural estate or in excess of them? The particular hypothesis put forward for the Sermersheim site leads to us wonder about the materiality and functioning of the “estate centre” (curtis dominica) perceptible from archaeology. In more concrete terms, the repeated finds of sunken-featured buildings, pits, wells, silos, ovens, etc. in these zones raise the question of whether there was a stereotyped organisation characteristic of an economic management type or, in contrast, the organisation of these “centres” was adaptable to estate productions or owners’ needs that would have changed depending on whether they were laypersons or ecclesiastics. Extending the argument, could not we wonder — but can archaeologist do it? — whether the large specific activities areas do not indicate part of a demesne? The example of Sermersheim raises another valuable question, in this case chronological. From when is it guaranteed to have a textile production unit with several workshops operating simultaneously and therefore to depend on the estate centre? As early as the sixth century? This dating questions the functioning of the production frameworks of domains and the reason for the long survival of some of them. The interpretation of the specific activities areas revealed at the sites clearly opens up new avenues of reflection regarding the economy of the countryside and the management of production frameworks. Archaeologists must nevertheless take the problematic seriously by re-examining the sites likely to be or to possess specific

31 Édith Peytremann, ‘Archaeology in the Village in France. Back Twenty Five Years of Experience’, in Archaeology of Medieval Villages Currently Inhabited in Europe, ed. by Jesús Fernández Fernández, Oxford, 2019, pp. 35-51. 32 Christopher Loveluck, Northwest Europe in the Early Middle Ages, C. ad 600–1150: A Comparative Archaeology, Cambridge, 2013, p. 60.

147

1 48

e di th p e ytr e m an n

activities zones, in other words, those comprising a large accumulation of identical structures, paying particularly attention to the chronology, in order to understand how all those identical structures were accumulated, the excavated area and the location of the area throughout the site. The nature of the archaeological finds, including the plant macro-remains and documentary data, if any, in relation to these sites must also be carefully studied. It is only on this basis that it will be possible to validate or not the previously proposed hypotheses.

Catarina Tente

Social Complexity in Local Communities During the Tenth Century in Central-Northern Portugal Negotiation and Opposition

Framework The territory under study is located in the highlands of the central-northern part of Portugal, between the Duero and the Mondego river basins, corresponding to the modern-day districts of Guarda and Viseu. These highlands are bordered by a series of mountain ranges, of which Estrela, to the east, is the highest summit, at 1,993 metres a.s.l. To the north, this territory is delimited by the left banks of the Douro river valley. It is also separated from the coastal plains, to the west, by the Montemuro and Caramulo mountain chains. Two important river basins — the Mondego and the Vouga — drain directly to the Atlantic. Within these vast plateaux, the city of Viseu is remarkable for its centrality. Viseu was capital of a Roman civitas, thus centralizing the road system that connected the northern, southern and inland sectors of Iberia with the nearest Atlantic coast. After the fall of Rome, Viseu became an episcopal centre, as documented at least from the sixth century onwards. From the eighth to the eleventh centuries, these territories became a frontier land between the north dominated by Asturians and the south controlled by Muslims. Among the several sites that have been under excavation in recent years, this work will focus on those for which there is presently more robust information — Penedo dos Mouros, São Gens, Soida and Senhora do Barrocal (fig. 1). A few more sites are known in the region but these have not been excavated or did not provide relevant information. The above-cited settlements have a number of features in common. First, they all share the same chronology. Indeed, a project of systematic radiocarbon dating of short-lived samples was carried out to obtain detailed absolute chronologies, given the lack of well-structured pottery typologies or type-fossils.1



1 C. Tente, A. F. Carvalho, 2011, ‘The establishment of radiocarbon chronologies for early medieval sites: a case study from the upper Mondego Valley (Guarda, Portugal)’, Munibe, 62 (2011), pp. 461–68. Catarina Tente  •  IEM | NOVA-FCSH Social Inequality in Early Medieval Europe: Local Societies and Beyond, ed. by Juan Antonio Quirós Castillo, Turnhout : Brepols, 2020 (HAMA 39), pp. 149–164 © FHG10.1484/M.HAMA-EB.5.118449

150

cata r i n a te nt e

Fig. 1. Map with the location of sites mentioned in the text.

Site

Provenance

Sector I, SU22 Penedo dos Sector I, SU22 Mouros São Gens

Sac-1950

Sector 10, SU9 Wk-25175 (collapsed palisade) Sector 10, SU8 Wk-27455 (hearth)

Sector II, SU5 (hearth) Sector 1, UE9 Senhora (fire level) do Sector 1, UE125 Barrocal (fire level)

Soida

Lab Number Sac-1947

Sample

Date BP

cal ad

Vicia faba L. var. minuta Vicia faba L. var. minuta Quercus sp. (cork) Quercus pyrenaica

1070 F0B1 45 876–1036 (95.4%) 1060 F0B1 40 892–1028 (95.4%) 1161 F0B1 30 773–906 (72.8%), 916–968 (22.6%) 1136 F0B1 30 777–793 (4.6%), 802–847 (9.0%), 856–985 (81.8%) 1098 F0B1 30 888–1013 (95.4%)

Sorbus aucuparia Wk-40079 Vicia faba 1040 F0B1 21 974–1025 (95.4%) (broad bean) Beta - 46513 Secale cerale 1170 F0B1 30 771–903 (80.8%), (rye) 918–965 (14.6%)

Wk-27454

Table 1: Radiocarbon determinations for the four sites mentioned in the text. Calibrations were23 obtained using Version 3.10 of the OxCal Program1 and based on the IntCal09 curve.2



2 C. B Bronk-Ramsey, ‘Bayesian analysis of radiocarbon dates’, Radiocarbon, 51: 1, 2009, pp. 337–60. 3 P.J Reimer, E. Bard, A. Bayliss, J. W. Beck, P. G. Blackwell, C. Bronk Ramsey, C. E. Buck, H. Cheng, R. L. Edwards, M. Friedrich, P. M. Grootes, T. P. Guilderson, H. Haflidason, I. Hajdas, C. Hatté, T. J. Heaton, D. L. Hoffmann, A. G. Hogg, K. A. Hughen, K. F. Kaiser, B. Kromer, S. W. Manning,

Social Complexity in Local Communities During the Tenth Century

Most of the sites consist of a single occupation level, which was destroyed by a fire that allowed the preservation of burnt organic matter. Therefore, most of the available radiocarbon determinations refer to the fire events and are dating the end of the occupation at those sites (tab. 1). All sites are located in discreet spots in the landscape. Clearly, the aim was in most cases to obtain visual control over the territory whilst remaining invisible in this landscape characterized by granitic outcrops. This strategy can be observed in valley bottom sites, such as São Gens, or in mountain sites, such as at Soida. In this particular case, the site is located on a narrow plateau in the northern section of the Estrela mountain range, and from there it is possible to overlook the whole surrounding landscape, namely the upper Mondego valley and the important castle of Trancoso. However, the chosen location was not the most prominent in the mountain landscape. To see but not to be seen was strategic to all sites. All sites were also surrounded by palisades and/or stone walls. In the majority of the cases, these structures were formed by a thick stone (or stone and earth) platform, with a height of less than one metre (fig. 2). On top of these platforms, palisades were built, usually with oak or chestnut trunks. The most remarkable case was identified at São Gens, where excavations observed that the trunks of the palisade collapsed (during the fire that destroyed the site) invariably towards the inside of the settlement, thus indicating that the outer side of the palisade’s base was structured with a stone buttress. The use of perishable materials was indeed a common feature to all sites. At Penedo dos Mouros it was possible to identify different testimonies of a wooden building resting on the eastern flank of a big granite boulder. This structure, which was built with timber from chestnut and oak trees,4 would have had at least two stores above the ground floor through which access to the inside and to the top of the building was possible. Several domestic artefacts and hundreds of broad beans, which were preserved due to carbonization, were stored inside. In all these sites it was possible to identify the building materials of the huts where people lived. Usually, each hut had a central hearth around which the domestic space was organized. Ethnographic parallels (fig. 3) show that these huts were completely built with perishable materials without the use of posts or any special preparation of hut floors.5 Domestic units were also excavated at Senhora do Barrocal, but unfortunately it was not possible to identify the central hearths. Here, houses were built with oak and chestnut timber and covered with ceramic tiles.6 M. Niu, R. W. Reimer, D. A. Richards, E. M. Scott, J. R. Southon, R. A. Staff, C. S. M. Turney, J. Van Der Plicht, ‘IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50,000 Years cal BP’, Radiocarbon, Vol. 55, No 4, 2013. 4 P. Queiroz, Novos dados arqueobotânicos sobre o Penedo dos Mouros (Gouveia). Terra Scenica-Território Antigo Relatórios 11, Lisboa, 2009. 5 C. Tente, ‘Etnoarqueologia das estruturas habitacionais no alto Mondego durante a Alta Idade Média’, in R. V. Gomes, C. Tente, M. V. Gomes (coords.), Cristãos e Muçulmanos na Idade Média Peninsular. Encontros e Desencontros, Aljezur/Lisboa, 2011, pp. 209–18. 6 C. Tente, H. Baptista, J. P. Tereso, M. Cércio, J. L. Veloso, C. Oliveira, L. Seabra, C. Meira, G. Souza, T. Cordero-Ruiz, M. L. Real, ‘A Senhora do Barrocal (Sátão) na viragem do milénio. Primeira abordagem à sequência ocupacional, cultura material e função do sítio arqueológico’, in C. Tente ed., Do Império ao Reino. Viseu e o território entre os séculos IV a XII, Viseu, pp. 263–96.

15 1

152

cata r i n a te nt e

Fig. 2. Aspects of the 2011 archaeological season at São Gens. A: the wall/palisade that delimited the settlement; B: one of the hearths/huts that were identified.

Fig. 3. Ethnographic parallels with Early Medieval habitation structures. 1: “Choupana” at Prime (Viseu); 2: “Choço” at Idanha (Castelo Branco); 3: “Choço” at Marvão (Portalegre). See E. V. Oliveira, F. Galhano, B. Pereira, Construções primitivas em Portugal, Lisboa, 1969, p. 53, 107, 110.

Social Complexity in Local Communities During the Tenth Century

Fig. 4. Pottery from São Gens.

Pottery productions were very similar. All sites revealed very homogeneous assemblages, essentially composed by jars, pans, and large bowls.7 Some variability is naturally seen between settlements but basic shapes, sizes and decorations are the same (fig. 4). Despite their small number, metal artefacts are found in every settlement and show very diversified morphologies: knives, axes, nails, rings, buttons and needles are the most common tool types (fig. 5). The recurrent finding of iron slags also indicates that metalworking was taking place in the settlements. At Senhora do Barrocal it was possible to identify two forge bases associated with a gridding stone that was reused as a working ‘table’, probably in connection with



7 C. Tente, P. Prieto, O. Lantes, ‘A produção cerâmica dos séculos IX a XI da região do Alto Mondego (Portugal)’, in A. De Man, C. Tente, (eds), Estudos de Cerâmica Medieval. O Norte e o Centro de Portugal. Séculos IX a XII, Lisboa, 2014, pp. 110–39. See also G. Souza, O sítio alto-medieval de São Gens (Celorico da Beira). Contributo para o estudo da produção cerâmica identificada nas campanhas realizadas entre 2011 e 2013, PhD thesis presented to FCSH-UNL, Lisbon, 2016, photocopied.

15 3

154

cata r i n a te nt e

Fig. 5. Metal artefacts from São Gens.

metalworking activities.8 All this evidence suggests that production of iron tools was not centralised. The fire events that all sites were subjected to allowed the preservation of organic matter, which includes not only building materials or fuel for hearths, but also edible plants. These are represented mostly by cultivated species that were stored in the settlements, such as cereals and pulses. Cereals are represented by rye (Secale cereal), oats (Avena sp. and Avena sativa/strigosa), proso millet (Panicum miliaceum), barley (Hordeum vulgare) and in lesser percentages, by wheat (Triticum aestivum/durum) and foxtail millet (Setaria italica). Pulses comprise broad beans (Vicia faba), peas (Pisum sativum), red/white peas (Lathyrus cicera/sativus) and lentils (Lens sp.). Fruits are also recorded, although in much smaller quantities: chestnut (Castanea sativia), pine-nut (Pinus pinea), arbutus (Arbutus unedo), cherry (Prunus avium), grape (Vitis vinifera), raspberry (Rubus cf. idaeus) and blackberry (Rubus type ulmifolius).9 Some

8 C. Tente, H. Baptista, J. P. Tereso, M. Cércio, J. L. Veloso, C. Oliveira, L. Seabra, C. Meira, G. Souza, T. Cordero-Ruiz, M. L. Real, ‘A Senhora do Barrocal (Sátão)…’, cit. 9 J. P. Tereso, C. Tente, H. Baptista, ‘O sítio da Senhora do Barrocal (Sátão, Viseu): vestígios das práticas agrícolas e de exploração dos recursos agrários no século X’, in International Conference Old and New Worlds: the global challenges of Rural History, Lisbon, 2016, pp. 1–14. See also C. Oliveira, A. Jesus, C. Tente, J. P. Tereso, ‘Estudo arqueobotânico do povoado alto-medieval de S. Gens: perspetivas sobre a exploração de recursos lenhosos e agrícolas’, in II Congresso da Associação dos Arqueólogos Portugueses — Arqueologia em Portugal. 2017 O Estado da questão, Lisbon, 2017, pp. 1463– 1476. And also C. Tente, H. Baptista, J. P. Tereso, M. Cércio, J. L. Veloso, C. Oliveira, L. Seabra, C. Meira, G. Souza, T. Cordero-Ruiz, M. L. Real, ‘A Senhora do Barrocal (Sátão)…cit.’.

Social Complexity in Local Communities During the Tenth Century

would have been cultivated while others were surely gathered in the surrounding woods and forests. So far, the largest assemblage was found at Senhora do Barrocal, where thousands of seeds were recovered under the collapsed roofs. Storage pottery vessels or storage pits are completely unknown. The evidence from Penedo dos Mouros or Senhora do Barrocal suggests that crops were stored in bags or baskets hanging from the ceiling or in boxes of wood or cork. At Senhora do Barrocal accumulations of flax (Linum sp.) seeds were also found, therefore indicating the growth and use, likely for textiles, of this plant.10 With the exception of Soida, cloth-making activities are also attested in the settlements by the presence of ceramic and stone spindle whorls. Also wool must have been weaved, though the presence of sheep was only identified in the faunal assemblage at São Gens.11 None of the sites would have been inhabited by more than three to six families, occupying therefore small areas. At São Gens, excavated areas have revealed evidence for three domestic units/huts from which we estimate the simultaneous presence of four to six families inside the enclosure. Here the enclosure is associated with a necropolis of 54 rock-cut graves likely representing a community of around three households over three or four generations.12 Even if graves were reused several times, one would have to multiply upwards by a factor of three or four burials in each grave to imagine a living community of more than a 100. It is therefore most likely that the socio-economic units represented were formed by just a few households. A similar size seems probable at the other sites. Despite sharing a large spectrum of similar features, some sites display evidence suggesting that these were single-purpose sites, or that some very specialized activities were taking place along with other, daily activities. This is, respectively, the case of Soida and São Gens.Unlike the other sites, at Soida there are no rock-cut tombs in its vicinities. Weaving and milling activities have not been recorded there either. Given its location on a high mountain plateau, Soida probably specialised in the seasonal mountain grazing of sheep and goat, and was perhaps functionally dependent on lowland settlements.13 Unfortunately, the lack of faunal remains prevents further conclusions. But at São Gens, where animal bones were preserved, a rather unexpected scenario was uncovered. The preliminary zooarchaeological analyses indicate the absolute dominance of hunted species, with 66% of the total fauna. The species of wild fauna represented are red deer (Cervus elaphus), rabbit (Oryctolagus cuniculus), roe deer (Capreolus capreolus) and wild boar (Sus scrofa). Domestic species, represented

10 C. Tente, H. Baptista, J. P. Tereso, M. Cércio, J. L. Veloso, C. Oliveira, L. Seabra, C. Meira, G. Souza, T. Cordero-Ruiz, M. L. Real, ‘A Senhora do Barrocal (Sátão)…cit.’. 11 C. Tente, S. Prata, S. Brookes, M. Moreno-García, G. Souza, F. Cuésta, ‘Povoamento e modos de vida no limite oriental do território viseense durante o século X. O Povoado de S. Gens’, in C. Tente ed., Do Império ao Reino. Viseu e o território entre os séculos IV a XII, Viseu, 2018, pp. 197–230. 12 S. Brookes, C. Tente, S. Prata, ‘Interpreting Rock-Cut Grave Cemeteries: the early medieval necropolis and enclosure of São Gens’, Medieval Archaeology, 61/2, 2017, pp. 215–38. 13 M. Fernández-Mier, C. Tente, C., ‘Transhumant herding systems in Iberia’, in E. Costello, E. Svensson (eds), Historical archaeologies of transhumance across Europe, London, 2018, pp. 219–32.

15 5

156

cata r i n a te nt e

by cattle (Bos taurus), pig (Sus domesticus), sheep/goats (Ovis aries/Capra hircus), and one horse (Equus sp.), are in the minority. The dominance of red deer, along with roe deer and wild boar — species that inhabit forests, scrubland and also cultivated fields — indicate the daily exploitation of woodlands that characterized the surrounding landscape. Here, pastures for the grazing of ovicaprines, and of sheep in particular, were perhaps scarce.14 It is important to mention that the red deer are mainly represented by adult individuals. Horse and red deer metapodials show signs of attrition suggesting their use as tools for fur processing, an activity whose presence is also deduced from the unbalanced representation of the animal’s body parts. The archaeological record from these settlements shows a broad trend of social equality and small-scale subsistence strategies. This is visible in the diversified composition of botanic (cereals and pulses) and faunal assemblages, as recorded at Senhora do Barrocal and São Gens, respectively. At the latter site, the age profiles of the domestic animals are compatible with a self-sufficient subsistence economy exploiting their secondary products (wool, milk, traction and manure). These communities also reveal a parsimonious management of wild resources, as evidenced in the preferable hunting15 of male deer. However, a more detailed analysis reveals a distinct reality. Some evidence indicates forms of social, and maybe political, inequality between settlements. The specialization in fur processing carried out by the São Gens people and the pastoralist practices apparent at Soida seem to testify some sort of economic complementarity. In a different sphere, Senhora do Barrocal reveals other forms of social inequality between communities.16 Indeed, despite some similarities, this site is not in full accordance with features observed at other sites known in the region. One of the contrasting differences is the presence of roof tiles in at least two of its houses. Moreover, these building materials were associated with exogenous pottery vessels, of Islamic production (fig. 7), which previously had only been found in the count’s castle of Trancoso.17 An iron spur must be also highlighted (fig. 8), a unique find in the region despite the presence of a horse among the São Gens fauna. Another exceptionality of Senhora do Barrocal is the presence of a nearby church, as attested by an inscription. This was the first time in this region and historical period that a relation between a settlement and a church could be established. All this seems to suggest that the local community, or part of it, belonged to higher social ranks, enabling them to reproduce signs of power typical of Early Medieval elites.

14 C. Tente, S. Prata, S. Brookes, M. Moreno-García, G. Souza, F. Cuésta, ‘Povoamento e modos de vida… cit.’. 15 C. Tente, S. Prata, S. Brookes, M. Moreno-García, G. Souza, F. Cuésta, ‘Povoamento e modos de vida… cit.’. 16 C. Tente, ‘Social inequality in Early Medieval rural settlements: The case of central-northern Portugal in the 10th century’, in J. A. Quirós Castillo ed., Archaeology of Social Inequality in Northwestern Iberia during the Early Medieval period, Oxford, 2016, pp. 35–46. 17 M. C. Ferreira, J. C. Lobão, H. Catarino, ‘Cerâmicas altomedievais do castelo de Trancoso: uma primeira abordagem’, Arqueologia Medieval, 12, 2012, pp. 15–31.

Social Complexity in Local Communities During the Tenth Century

Fig. 6. Plan of Senhora do Barrocal (light grey: granitic boulders; dark grey: wall). A: rock-cut grave and modern church; B: settlement.

Fig. 7. Islamic pottery from Senhora do Barrocal. A. potsherd of a green and manganese decorated bowl; B. potsherd of a glazed honey-coloured bowl decorated with manganese lines; C. sherd of a painted black and white vessel; D. green-glazed bottle.

15 7

1 58

cata r i n a te nt e

Fig. 8. Archaeological context at Senhora do Barrocal where an iron spur, locally-made pottery and an Islamic green-glazed bottle can be seen in close association.

Negotiation The establishment of strong community ties and the eventual rise of a social group is usually a slow process. However, all the above-described settlements had a very short temporal duration. They were all founded in the ninth century, which implies that these communities were already formed beforehand. We do not know yet where these communities came from, nor if they had inhabited the region since the end of the Roman Empire and changed their settlement locations for still unknown reasons. Howsoever, during most of the ninth century their archaeological record testimonies stability, both in settlement and in the people’s daily lives. However, this would become a politically unstable region, in particular between the end of the ninth and the middle of the eleventh centuries. This was a moment of expansion and regression of Asturian and Muslim interests. In this context, the equilibrium of the local communities in the region could only have been reached through a complex system of negotiations, not only among themselves, but also with those external powers. Two cases stand out regarding these communities’ ability to negotiate. One is the case of Senhora do Barrocal, where — as pointed out above — Islamic pottery, a spur (fig. 8) and a church were identified. The Islamic vessels are exceptional pieces, not only in this specific northern context, but also in any Islamic site, a fact that allows them to be classed as prestige goods. Likewise, the church is a strong component of the representation of power in the Early Middle Ages.

Social Complexity in Local Communities During the Tenth Century

With all likelihood Senhora do Barrocal is the archaeological manifestation of local elites directly or indirectly connected to Islamic powers in a moment when the region was integrated in the Kingdom of Asturias after the annexation of Muslim lands (‘presúrias’) by King Alfonso III.18 Thus, it shows the ability of these groups to establish relations not only within their territory but also beyond it. Describing the attack on Santiago de Compostela in 997 by Almansor, Ibn Idhārī explains that he gathered previously in Viseu with a large number of counts who recognized his authority and joined his army.19 Would these be people like those that lived in settlements such as Senhora do Barrocal? It should be also mentioned that, after 997, the region would have been in principle under the control of the Caliphate. However, even before the military campaigns by Fernando I, Islamic attacks may have taken place in the region, indicating that effective control by Islamic authorities was lacking. Therefore, the most likely scenario is of a region generally occupied and controlled by local communities during this period of political instability, among which some social groups would have risen. These would have acquired sufficient relevance and ability to act outside their local spaces and be related — through strategies still unknown to us in their real amplitude — with dominant powers on both sides, Asturian and Caliphal. The second case showing these communities’ ability to negotiate is the Lafões episode described by Al-Muwa’īnī. At an undetermined date, but probably between 1026 and 1030, the Sevillian al-Mu’tadid may have conducted a military campaign to the area north of Viseu during which a couple of castles known by the name al-Aḫawayn (“The Two Brothers”) may have been sieged. This site name corresponds to Lafões.20 Both castles were conquered — surely after negotiation — and 300 of their defenders participated in al-Mu’tadid’s army, likely as mercenaries.21 Despite the limitations inherent to such a late mythical account, it is possible that the episode refers to one of the pacts established between local populations and Islamic power. In this 18 This proves that the territorial control by the Asturian kingdom in the region was restricted to some specific geographic areas, as would be the case of the city of Viseu or the Lafões area, or even of the villages associated with the counts’ castles mentioned in the donation of Châmoa Rodrigues to the Guimarães monastery in 960. See A. Herculano, Portugalia Monumenta Historica. Diplomata et chartae, Lisboa, Academia das Ciências, LXXXI, 1867, doc. LXXXI. 19 P. Sénac, Almanzor. El azote de año mil, València, 2011, p. 94. 20 C. Aillet, Les Mozarabes. Christianisme, ismalisation et arabisation en Péninsule Ibérique (ixe-xiie siècle), Madrid, 2010, p. xv. 21 See F. Maillo Salgado, Acerca de la conquista árabe de Hispania. Imprecisiones, equívocos y patrañas, Madrid, 2016, pp. 29–29, 130–31. In this episode it is also mentioned that the Christians inhabiting Lafões were descendants of Ğabala b. al-Āyham al-Ġassānī, and that their ancestors had made a pact with Mūsa b. Nusayr. As explicitly stated by C. Aillet, this association is mythological and cannot be taken as a sound historical fact. As the author puts it, ‘Cet épisode guerrier anodin, qui s’etait déroulé dans une marge obscure de l’Islam, se transforme ainsi en référence littéraire, établissaint un paralèle entre le geste des ‘Abbādīdes de Sèville et celle de leurs plus glorieux ancêtres orientaux. La capture des ces Ġassānides fictifs ne peut en effet que renvoyer à l’un des postulats centraux de l’idéologie des maître de Séville, à savoir leur affiliation supposée avec les Laḫmīdes de Hira, dynastie arabe de la Ğahiliyya qui justement luttait contre les Ğassānides de Syrie pour le compte de la Perse’. C. Aillet, Les Mozarabes. Christianisme…cit., p. 300.

1 59

1 60

cata r i n a te nt e

context, Islamic pottery such as the vessels found at Senhora do Barrocal may have well sealed this kind of agreement, a practice acknowledged in the establishment of pacts and political relations.22 These two cases illustrate the existence of local negotiators. It is however difficult to perceive who these local negotiators were; if it was the communities as a whole or their elites representing the community and/or themselves. One of the signs of ability for negotiation shown by rural communities appears at a later date, when they obtain the king’s consent to maintain their customs. This is explicit in the royal privilege charters granted to communities in the region in the twelfth and the thirteenth centuries. It is beyond the scope of the present text to discuss whether or not the granting of the earliest royal privilege charters mark the end of the peasants’ freedom, but it will evaluate how these early charts 23 reflect the presence of complex local communities that had not been subjected to landlord pressure. These documents show that these frontier communities already existed and had their own organization long before the royal privilege charter.24 We believe that these communities to whom kings awarded the first royal privilege charters must have originated in communities similar to those identified at Penedos dos Mouros, Soida, and Senhora do Barrocal. Despite the destruction of their settlements, these human groups resisted in the region and possessed, decades later, the ability to negotiate again with the dominant power, in this case, the power of the king.25

22 W. Davies, ‘When gift is sale: reciprocities and commodities in tenth-century Christian Iberia’, in W. Davies, P. Fouracre (eds), The Language of Gift in the Early Middle Ages, Cambridge, 2012, pp. 217–37. It is not strange that the Lafões region is explicitly mentioned in Islamic sources. Indeed, at the time of Alfonso III of Asturias some count elites were established in the area. See M. L. Real, ‘A dinâmica cultural em ‘Portucale’ e ‘Colimbriae’ nos séculos VIII a XI’, in A. De Man, C. Tente (eds), Estudos de Cerâmica…cit., pp. 110–39. And also M. L. Real, ‘O papel das elites na definição progressiva do território: sua presença na senhorialização da ‘Fronteira Beirã’ (Séculos IX–XII)’, in C. Tente ed., Do Império ao Reino. Viseu e o território entre os séculos IV a XII, Viseu, pp. 297–344. In the Lafões area there was a considerable number of churches and monasteries before 1100 and their building must be related with these elites. It is not by chance that this density of religious buildings sharply contrasts with their scarcity in the remaining territory between the Douro and the Mondego river valleys, in particular where the presence of these elites is not documented by other means. 23 We can mention those granted by Fernando I in the area of the Douro River, more specifically to Linhares, Ansiães, São João da Pesqueira, Paredes and Penela da Beira between 1057 and 1065, of which we do not know the original texts but only later confirmations by the first Portuguese kings. 24 J. Mattoso, Identificação de um País. Oposição, Composição, Lisbon, 2015 (first edition 1985). 25 In the royal privilege charter of Alfaiates, a village located in the Portuguese–Leonese frontier, issued in the thirteenth century, which was studied by J. Mattoso it is possible to see a very cohesive community, internally stratified, with the ability to defend its own common goods and itself from exterior powers, including the religious authorities. Originally, the king imposed himself only as an external authority against possible abuses; he interfered in the political structure of the community only superficially. But in the course of time, the monarchy would absorb all the autonomous political formations including the former municipalities. See J. Mattoso, ‘Da comunidade primitiva ao município: o exemplo de Alfaiates’, in Fragmentos de uma composição Medieval, Lisboa, 1987.

Social Complexity in Local Communities During the Tenth Century

…and Opposition ‘Opposition’ is the first part of the reference book on the formation of Portugal by J. Mattoso, entitled “Identificação de Um País”.26 This term is particularly well chosen to show how opposition between North and South, mountains and plains, and different social and cultural groups that were formed during the Christian conquest, were crucial for the formation of the Portuguese Kingdom. This idea is now retrieved to argue that negotiation was not always the dominant option and because the archaeological record also shows evidence for opposition. The fact that settlements located in this frontier region were set on fire and abandoned may be the material testimony of the use of force in a twofold process — the submission of resistant groups and the advancement of more dominant powers. The only case of re-occupation is attested at Senhora do Barrocal, where the wall around the settlement was built immediately after the fire. This is a crucial observation because it testifies that the site was not abandoned after that event, thus revealing a life history contrasting with observations obtained elsewhere. But who re-built it? It is not impossible they were the very same people that lived there at the time of the fire. However, another possibility seems likely. As all the evidence indicates, the event that determined the site’s abandonment was probably an attack, and the walls may well have been built by the attackers themselves. This hypothesis is supported by the observation that the wall had a military purpose, not residential as before.27 We do not know where the populations from the burned villages would have been re-located. Some of them were probably displaced to found the modern-day villages, where castles would be built during the second half of the eleventh century.28 But the question remains: Who were the attackers? It is my opinion that these could not be the (decadent) count elites, the Caliphate, the Taifa kingdoms, nor even the Asturian powers, because these were struggling with increasing difficulties in the political control of the area, as we have seen. It is possible these attacks resulted from social re-organization of local communities. In this regard, Reyna Pastor29

26 J. Mattoso, Identificação de um País…cit. 27 C. Tente, S. Prata, S. Brookes, M. Moreno-García, G. Souza, F. Cuésta, ‘Povoamento e modos de vida… cit.’. 28 In order to test this hypothesis, we will need to excavate selected areas of the modern villages, which has not yet been possible. 29 R Pastor de Togneri, Resitencias y luchas campesinas en la época del crecimiento y consolidación de la formación feudal. Castilla y León, siglos X–XIII, 2nd Edicion, Madrid, 1990, pp. 37–43. The author claims that these groups may have arisen out of the needs of war. However, it is not difficult to imagine that social stratification in peasant communities may have been influenced by other factors, such as the possession of the most productive plots, the ability to negotiate with individuals external to the communities, personal/familial achievements, etc. Some specialized activities, such as the one documented at São Gens, could have been a factor of social differentiations between peasant groups. Unfortunately, archaeology will hardly be able to identify which factors of social differentiation between members of the same community were at play in this cultural context.

161

162

cata r i n a te nt e

states that the peasant communities of Old Castile were already stratified by the tenth century, with some social groups beginning to stand out. Processes of internal hierarchy taking place within some peasant communities, the military instability in the region in the time period between the end of the ninth and the middle of the eleventh centuries, and the absence of dominant powers, design the perfect scenario for the emergence of internal conflicts. Political ambitions of some emerging groups must have originated situations of stress between themselves and local communities opposing their expansion. As highlighted by Mattoso, the rise of the landlord system must have been a slow and individualized process in most cases.30 But it was not always so, and written sources also describe some cases in which violence was used to submit communities.31 In sum, violence was not the rule but it was not an unknown socio-political resource.

Conclusions I must admit that some years ago I thought most of the rural communities in this sector of the Iberian Peninsula were living in autarchy.32 However, the enlargement of the excavated areas in settlements, together with the excavation of recently-discovered sites, shows a much more complex picture for the tenth century in this area. The social and political scenario we are starting to uncover shows that, as argued by C. Wickham33 for North-Western Iberia, peasant communities had strong identities. Moreover, they also had specific political agendas, as demonstrated clearly in the recent book edited by J. A. Quirós Castillo.34 The (at least apparent) high degree of economic specialization observed at São Gens and Soida seems to suggest some form of economic complementarity between sites. At the local level, subsistence practices ensured the survival of community members but, at the same time, they engaged in more specialized activities allowing them to integrate complex networks external to the community.35

30 The author identifies three forms of absorption of free peasant communities: incommunicatio; confiscation of lands; and estate purchase. See J. Mattoso, Identificação de um País…cit., pp. 310–12. 31 An example must have been the episode of violence by landlords over resisting peasants in the Trásos-Montes region described in the 1258 general inquiries, as mentioned by R. Durand, ‘Communautés villageoises et seigneurie au Portugal (x–xiii siècles)’, in Estudos de História de Portugal. Homenagem a A. H. de Oliveira Marques, Lisbon, 1982, pp. 121–36. 32 C. Tente, ‘Viver em autarcia. O Alto Mondego entre os séculos V e XI’, in I. Martín Viso ed., Tiempos oscuros? Territorio y sociedad en el centro de la Península Ibérica (siglos VI–X), Salamanca, 2009, pp. 137–57. 33 C. Wickham, Framing the Early Middle Ages. Europe and the Mediterranean, 400–800, Oxford, 2005. 34 J. A. Quirós Castillo ed., Archaeology of Social Inequality in Northwestern Iberia during the Early Medieval period, Oxford, 2016. 35 J. A. Quirós Castillo, ‘Inequality and social complexity in peasant societies. Some approaches to early medieval north-westen Iberia’, in J. A. Quirós Castillo ed., Archaeology of Social Inequality in Northwestern Iberia during the Early Medieval period, Oxford, 2016, p. 7.

Social Complexity in Local Communities During the Tenth Century

Before the excavations carried out at Senhora do Barrocal, it was very difficult to identify markers of social differentiation. However, the finds of roof tiles, prestige pottery productions imported from the Muslim world, and an iron spur, and the capacity to mobilize labour and means to build a church, are indicative of the striking social feature of its inhabitants. Their daily lives would not differ greatly from their fellow, coeval neighbours, but it is beyond any reasonable doubt that this was a rather distinct community that used a large spectrum of social markers to express power. Though it is unquestionable that the inhabitants of Senhora do Barrocal were operating on differentiated social levels, distinct from the remaining local communities, it also true that their role in the region cannot be compared with that of Aistra, in the Alava region, where the archaeological record indicates it was a place for elites, absorbing incomes obtained from a dense socio-economic network,36 and this does not seem to have taken place at Senhora do Barrocal. It is still difficult to evaluate the role played by these local elites in the management of the region as a whole. The overall image is much more complex than usually acknowledged, not only in strict archaeological terms but also in the social and economic structures of Medieval communities. We are now able to identify and understand more profoundly how power was represented in its various material manifestations during the Early Middle Ages. This is particularly visible in the building of churches and monasteries but gains further relevance if cases such as the single site of Senhora do Barrocal or the Lafões territory are put in context. Neither Asturian kings nor Islamic powers seem to have controlled the region until the conquest of Coimbra by Fernando I in 1064. The absence of structuring powers, and consequently of central places, reinforced the role of local communities in this territory. As mentioned by C. Wickham,37 this was the time when the growth of local communities helped to establish community ties, which were probably reinforced by the complex political scenario in the region between the tenth and eleventh centuries. Some sectors within the communities must have stood out. These must have become important enough to be able to act outside the local space and establish lasting relations with both the Caliphal and Asturian dominant powers. The changes in settlement, directly correlated with the fires identified in the sites, are probably the consequence of the struggle between local elites to control territories and communities.

36 J. A. Quirós Castillo, ‘Ser y parecer. Arqueología del poder en el noroeste peninsular’, in C. Tente ed., Do Império ao Reino. Viseu e o território entre os séculos IV a XII, Viseu, 2018, pp. 23–44. 37 C. Wickham, 2002, ‘Asentamientos rurales en el Mediterráneo Occidental en la Alta Edad Media’, in C. Trillo ed., Asentamientos rurales y territorio en el Mediterráneo Medieval, Granada, 2002, pp. 11–29.

163

Richard Hodges

The Primitivism of the Early Medieval Peasant in Italy?

The Early Middle Ages would be a poorer and simpler world than the Imperial past because networks of warrior aristocrats, and their kin, could provide no effective substitute for the structures of the state. The losers were not principally aristocratic households, or even the merchants who thronged the jetties of Commacchio or Dorestad, but the millions of peasants for whom the fall of the imperial system ushered in a duller and more uncomfortable age.1

‘Truffling Hunting’ with Chris Wickham, 1978–1985 Chris Wickham and I first actively pondered together the material nature of the Italian early Medieval peasant during the excavations at Santa Maria in Civita in the Biferno Valley, Molise in 1978. We assumed that this hilltop site was a failed precursor of the castello of Guadialfiera nearby, and that therefore its conspicuous material culture provided a remarkable and important index of wealth in the otherwise unknown eighth to ninth centuries. Santa Maria in Civita was exceptionally rich in terms of red-painted pottery, glass, a little metalwork as well as faunal and palaeobotanical evidence. Our assumptions, as I later showed, were fallacious. The palaeobotanical and settlement form were to show it was an incipient seigneurial centre of the later ninth century not a village of peasant dwellings. For some reason the hilltop was summarily abandoned with its storage pits of cereals still full.2 Our assumptions in 1978 were reinforced because we assumed the small hilltop settlement was a failed version of the better-known deserted hilltop villages in

1 M. Whittow, ‘How much trade was local, regional and inter-regional?’ in L. Lavan (ed.) Local Economies. Production and Exchange of Inland Regions in Late Antiquity X (I), Leiden, 2013, pp. 133–66. 2 R. Hodges, G. Barker and K. Wade, ‘Excavations at D85 (Santa Maria in Civita): an early medieval hilltop settlement in Molise’, Papers of the British School at Rome, 48, 1980, pp. 70–124; M. Van Der Veen, ‘An Early Medieval hilltop settlement in Molise: The plant remains’, in Papers of the British Richard Hodges  •  American University of Rome, Italy Social Inequality in Early Medieval Europe: Local Societies and Beyond, ed. by Juan Antonio Quirós Castillo, Turnhout : Brepols, 2020 (HAMA 39), pp. 165–174 © FHG10.1484/M.HAMA-EB.5.118450

166

r icha r d ho d g e s

the Biferno valley like Vetrana (near Guglionesi) known from the tenth–eleventh centuries.3 Two years later with these assumptions in mind we began the project at San Vincenzo al Volturno, based upon the Benedictine monastery then principally known from the twelfth-century Chronicon Vulturnense. In framing the archaeological survey of the terra of San Vincenzo, Chris Wickham provided detailed topographic notes on each of the monastery’s small settlements, many of which later became chartered castelli. With the ceramic sequence from the Biferno Valley in mind, we set out to pinpoint these places and seek other, unknown sites.4 Most memorable was Chris Wickham’s lone visit to the site of an eighth- to tenth- century conduma, Casa S. Lorenzo. Casa S. Lorenzo is mentioned briefly in the Chronicon Vulturnense5 when its tenant in ad 973, then living in Castel San Vincenzo, had his lease confirmed. The lease is for a small and quantifiable property (arable of 100 × 90 passus, vineyard of 20 × 16). The tenant in question was a recent immigrant from the nearby Valle Trita. It was tempting to speculate that this was an older farm, one of several dotted about the Rocchetta Plain (on which San Vincenzo is located) which this man acquired in the tenth century, possibly after its original eighth- to ninth-century tenants, perhaps of servile status, had disappeared as a result of the sack of San Vincenzo al Volturno in ad 881. The position of the land is recorded on modern cadastral maps as a terrace about 1 km south of San Vincenzo al Volturno. The Casa S. Lorenzo terrace is about 400 m long and 80 m wide. On its east side is a sheer drop to the gorge of the river Volturno below; its west side is formed by a slope which is steep only towards the top, rising to the plateau about 70 m above. The terrace is used for cereals today, while the slope is overgrown. Nonetheless, remains of a Post-Medieval (nineteenth-century) farm and terraces on the slope show that it has been utilised in recent times. On that September day in 1980 Chris Wickham found a few sherds of Roman date on the terrace itself, but no traces of any diagnostic early Medieval pottery.6 When he arrived back at San Vincenzo al Volturno later School at Rome, 53, 1985, pp. 211–24; R. Hodges and C. Wickham, ‘The evolution of hilltop villages (ad 600–1500)’, in G. Barker (ed.), A Mediterranean Valley, Leicester, 1995, pp. 254–85; K. Bowes and R. Hodges, ‘Santa Maria in Civita revisited’, Papers of the British School at Rome, 70, 2002, pp. 359–61. 3 R. Hodges and C. J. Wickham, ‘Vetrana: un villaggio abbandonato altomedievale presso Guglionisi nella valle del Biferno (Molise)’, Archeologia Medievale VIII, 1981, pp. 492–502. 4 R. Hodges, ‘Introducing Medieval Archaeology to Molise in 1977–80’, Post-Classical Archaeology, 1, 2011, pp. 481–92. 5 C. J. Wickham, Il Problema dell’Incastellamento dell’Italia Centrale: L’Esempio di San Vincenzo al Volturno, Florence, 1985, p. 27; See also F. Marazzi, San Vincenzo al Volturno. L’Abbazia e il suo Territorium fra VIII e XII secolo, Montecassino, 2012, p. 81. 6 In 1986 the terrace was examined again. Deep-ploughing made it clear that the Roman site was situated somewhere towards the centre of the terrace, immediately below the post-medieval farm on the slope, but not on the terrace itself. A deep pipe-trench was also cut through the terrace from east to west from which came a few further sherds of Roman pottery supporting this hypothesis. No Medieval pottery was found. As early Medieval pottery has been identified at San Vincenzo al Volturno we cannot doubt that, if it existed, it would have been recognised here. The available evidence, therefore, suggests that the early Medieval conduma was small, and perhaps located where the ruins of the post-Medieval farm now lie, close to or on top of the small Roman site.

the p r i m i t i vi s m o f t h e e ar ly m ed i e val pe asant i n i taly ?

that September day he was hot and bothered. A place he had envisaged to be possibly as rich as Santa Maria in Civita, close to the celebrated monastery, was to all intents and purposes materially invisible. Soon afterwards we identified the hilltop site of Colle Castellano close to the village of Montaquila in the San Vincenzo field survey. This had diagnostic later ninth-, tenth- and eleventh-century ceramic red-painted and coarse wares.7 Chris Wickham hypothesized that it was the precursor of the Norman-period castello of Montaquila, a failed village known as Olivella.8 Colle Castellano being more than 20 kms from the monastery and given the uncertainty about its historical name, it was decided to delay its investigation and focus instead upon a castello called Vacchereccia, 4 kms. from San Vincenzo al Volturno in order to evaluate the monastery’s economic impact upon its rural settlements. Encouraged by Chris Wickham and a detailed topographical assessment, we embarked upon a full-scale evaluation in 1982.9 First, we excavated dozens of test-pits on the steep slope of this thickly wooded hilltop site. Where traces of occupation occurred, as at the top and on the mid-slope, we opened up trenches. The result of this bold venture was to discover clear indications of a twelfth-to fourteenth-century village on the hilltop, and the most vestigial traces of an early Medieval dwelling and its rubbish on a destroyed terrace on the mid slope. The evidence was truly striking for its minimalist footprint, little more than dry-stone walling deployed to terrace the front of a probable post-built structure. The material culture consisted of a diagnostic brooch, some poor quality cooking-wares and re-used tiles, some from San Vincenzo al Volturno’s early ninth-century tile kilns. The underwhelming material record of Vacchereccia was extraordinary. This conclusion was soon accentuated by the later realization that the diagnostic pottery from Vacchereccia’s mid-slope dwelling in fact probably dated to the tenth-eleventh centuries, and that the distinctive wares found in later excavations at Colle Castellano — Olivella — were associated with a small stone tower, presumably belonging to a family of a minor landowner, not a peasant family. In sum, the celebrated ninth-century monastery of San Vincenzo al Volturno, with its astonishing array of architectural ambition occupied a world where peasants were virtually invisible.10 There was only a minimal ‘trickle down’ relationship between the monastery at its European zenith in the ninth century and its peasants (if this

7 R. Hodges, G. Clark, S. Coccia and H. Patterson, ‘Excavations at Colle Castellano’, in K. Bowes, K. Francis, and R. Hodges, ( eds), Between text and territory. Survey and Excavations in the Terra of San Vincenzo al Volturno. London, 2006, pp. 187–223. 8 For contrarian reasons the interpretation of Colle Castellano as the lost village of Olivella has been challenged by Federico Marazzi, op. cit. p. 58, fn. 100. 9 R. Hodges, P. Grierson, P. Herring, P., V. Higgins, J. Nowakowski, H. Patterson, and C. Wickham, ‘Excavations at Vacchereccia (Rocchetta Nuova): a later Roman and early medieval settlement in the Volturno valley, Molise’ Papers of the British School at Rome 52, 1984, pp. 148–94; R. Hodges, ‘Vacchereccia reconsidered’, In K. Bowes, K. Francis and R. Hodges, (eds), Between text and territory. Survey and Excavations in the Terra of San Vincenzo al Volturno, London, 2006, pp. 263–67; Marazzi, op. cit., pp. 80–81. 10 R. Hodges, ‘Landscape and Society: The Making of San Vincenzo’s Mediterranean Valley’, in K. Boyle, R. Rabett & C. Hunt (ed.), Living in the Landscape, Cambridge, 2014, pp. 267–85.

1 67

168

r icha r d ho d g e s

relationship even existed). Puzzled because the archaeology defied the presumed logic of the historical economy, Chris Wickham two decades later arrived at the following conclusion (my italics): Unhelpfully, the single most unambiguous sign of renewed economic activity after 750 is the huge wealth and artisanal sophistication found in S. Vincenzo al Volturno in the early ninth century; but S. Vincenzo was one of the remotest rural monasteries in Italy.11 But San Vincenzo was not remote. It belonged to the Principality of Benevento, a duchy that by any Mediterranean standards was flourishing in cultural terms in the ninth-century because it was a buffer zone between Byzantium and an emergent Latin Christendom. More to the point nor was San Vincenzo al Volturno exceptional in the ninth-century. It belonged to a Benedictine world which was international in character. Its architectural history imitated the new rhetoric of building in early ninth-century Rome in this period12 and elements of the architectural formula — crypt, passageways, prestige goods workshops — have been found at other Benedictine monasteries in Italy, notably Farfa, S. Pietro a Palazzuolo and Nonantola.13 At issue was the inherent assumptions about trickledown economics from elite centres such as monasteries to peasant communities in the transformation of the Roman world and the beginnings of the Middle Ages. The challenge to these trickledown assumptions became crystal clear in the excavations of the Tuscan deserted Medieval village of Montarrenti in September 1985. In the third season of the Sheffield-Siena University excavations of this hilltop village west of Siena in Area 2000 on a terrace on the mid-slope deep beneath a twelfth- to fourteenth-century ruined house remains of the earliest occupation were found.14 Very simply a small sunken rock-cut building with traces of post-holes at its angles and the burnt shadows of household fires was discovered beneath a well-built eleventh–twelfth-century successor dwelling. The modest simplicity of the earliest structure was at first astonishing.15 No less astonishing was the pitiful quantities of associated material culture. Montarrenti, now a type site for the evolution of Italian hilltop villages, evolved from unalloyed material impoverishment to be a place where, by the twelfth century, urban-style towers on the summit were accompanied by large ashlar-built dwellings around the mid-slope girth of the hill. The Studio Inklink reconstructions of the village in different periods vividly bring to life the extraordinary transformation of the village, even if their detail is perhaps sometimes fanciful.16 ‘Truffle-hunting’, as I colourfully described it at the time of the excavations, had

11 C. Wickham, Framing the Middle Ages, Oxford University Press, Oxford, 2005, p. 651. 12 C. Goodson The Rome of Pope Paschal I. Papal power, urban renovation, church building and relic translation, 817–24, Cambridge, 2010, pp. 260–65. 13 R. Hodges, S. Leppard and J. Mitchell, San Vincenzo Maggiore and its Workshops, London, 2011, pp. 433–59. 14 V. Bartolini ‘Relazione finale di scavo area 2000’, Archeologia Medievale XIII, 1986, pp. 259–66. F. Cantini, Il castello di Montarrenti, All’Insegna del Giglio, Florence, 2003, pp. 27–28 15 See R. Hodges, Travels with an Archaeologist, London, 2017, pp. 32–35. 16 Cantini, Montarrenti, pp. 232–40.

the p r i m i t i vi s m o f t h e e ar ly m ed i e val pe asant i n i taly ?

pinpointed what ‘parachutists’ could not: the early Medieval Italian peasant had lived in highly limited economic circumstances until an explosive change occurred in their economic and thus their material circumstances between the later ninth and twelfth centuries. Their material footprint by any prehistoric, Roman or post-Roman indices suggested a level of primitivism, to use a relative and pejorative term (like aboriginal or poor). I use this word in the following pages with reluctance in the absence of a less socially charged term.17 The under-developed culture of peasant life between the seventh and ninth centuries has been affirmed by many excavations in Tuscany at Miranduolo,18 close to Montarrenti, and at Rocchette in the Colline Metallifere19 as well as by investigations of several dispersed — single habitation — sites in the Maremma.20 Comparison with the material-rich fifth-century peasant farmstead of Pievina near Cinigiano and the modest dwelling at Colle Massaro close by from the eighth- to ninth centuries reinforces the striking material poverty of the latter.21 Tuscany is not unique: similar sites have been found in many other regions of Italy. The little sites of San Donato in the Sabina and Nepi in northern Lazio superficially possess the same characteristics.22 In most cases the dwellings sometimes were on low terraces with simple drains around the outside, but most were sunken into the place with rock-cut or earthen floors on which hearths were set. The wooden posts of the structures were generally irregular and often wedged with stones. The carpentry on the present evidence was as simple and expedient as the overall architectural concept. Associated storage facilities have so far not been detected. Commodities other than a limited range of pottery were not a feature of this period. So, no jewelry, metalwork, tiles or other objects tend to be associated with these places. In material terms, these settlements resemble those of

17 Cf. B. Ward-Perkins, ‘Continuists, catastrophists, and the towns of post-Roman northern Italy’, Papers of the British School at Rome 65, 1997, pp. 157–76 at p. 167;R. Hodges, ‘Parachutists and truffle-hunters: at the frontier of archaeology and histor’,, in M. Aston, D. Austin & C. Dyer (eds), Rural Settlement of Medieval Britain, Blackwell, Oxford, pp.287-305. 18 M. Valenti (ed.), Miranduolo in alta val di Merse (Chiusdino — SI), Florence, 2008, pp. 84–112. 19 G. Bianchi, ‘Building, inhabiting and ‘perceiving’ private houses in early medieval Italy’, Arqueologia de la Arquitectura 9, 2012, pp. 195–212. 20 E. Vaccaro, Sites and Pots: Settlement and Economy in Southern Tuscany (AD300–900), Oxford, 2011, pp. 168–231. 21 Compare the later Roman site at Pievina - M. Ghisleni, E. Vaccaro and K. Bowes, ‘Excavating the Roman peasant I: excavations at Pievina’, Papers of the British School at Rome 79, 2011, pp. 95–145 with the early Medieval site of Colle Massari: E. Vaccaro, ‘Ceramic production and trade in Tuscany (third-mid ninth c. ad): new evidence from the south-west’, in E. Cirelli, F. Diosono and H. Patterson, (eds), Le forme della crisi. Produzioni ceramiche e commerci nell’Italia centrale tra Romani e Longobardi (III–VIII d.c.), Bologna, 2015, pp. 221–27 at p. 223; 225. 22 H. Patterson and A. Rovelli, ‘Ceramics and coins in the Middle Tiber Valley from the fifth to the tenth centuries ad’, in H. Patterson (ed.), Bridging the Tiber. Approaches to regional archaeology in the Middle Tiber Valley, London, 2004, pp. 267–84. H. Patterson, ‘Rural settlement and the economy in the Middle Tiber Valley: ad 300–1000’, Archeologia Medievale XXXVII, 2010, pp. 143–61.

169

170

r icha r d ho d g e s

the Neolithic age before the so-called Secondary Products Revolution.23 The sizes of the dwellings do not differ much from some of the diverse types of Roman peasant dwellings. However, after the material-rich age of late antiquity, it is as if, to quote the British novelist Lawrence Durrell, an ice age had settled on the Roman Empire.24 These dwellings each appear to be the home of a family, although the human form of the household is not clear. Was this a duller and more uncomfortable life, as Mark Whittow asked in his review of Chris Wickham’s Framing the Middle Ages25? We cannot yet know. The human dimension is missing in the archaeological record. Cemetery data is now needed to proffer observations about qualities of life, and such data currently do not exist. The apparent subsistence character of the excavated remains suggests that taxation at peasant level was minimal, to judge from the materialism. However, it is both wrong to imagine this primitivism was a consequence of poverty — a social condition — as it would be wrong to uphold today the assumptions we had at Santa Maria in Civita in the Biferno valley (Molise), based on our own tacit ‘Galbreithian’ (1950s) views of progressive affuence of a post-classical society (see below).26 The archaeological evidence not only challenges the texts but also the inherent assumptions we have had in reading those texts.

A Resistance to the Archaeological Evidence? It is not difficult to grasp why many (though not all) historians27 have resisted this discovery of primitive settlements in the first centuries of early Medieval Italy. The village of the historians is invariably seen through the lens of their observers, the authors of their texts. But why have archaeologists resisted the meaning of this evidence of primitivism? Herein lies an interesting problem, or rather a package of problems. First, to acknowledge the primitive settlement and material evidence affirms a massive decline of Edward Gibbon-esque proportions — including demographic decline — of some kind at the end of late antiquity. Second, to acknowledge this primitive rural world is to effectively acknowledge the minimal social relations between all levels of society in this world. The archaeology suggests virtually no material relations existed between different levels of society before some point in the mid to later ninth century.

23 A. Sherratt, ‘Plough and pastoralism: aspects of the secondary products revolution’, in I. Hodder, G. Isaac and N. Hammond (eds), Pattern of the Past: Studies in Honour of David Clarke, Cambridge, 1981, pp. 261–305. I owe this idea to my wife, Dr Kim Bowes. 24 L. Durrell, Prospero’s Cell, London, 1945, p. 75. 25 M. Whittow, op. cit., p. 159. 26 See J. K. Galbreith’s The Affluent Society, New York, 1958: also R. Hodges, Dark Age Economics: A New Audit, London, 2012, pp. 41–42; Ward-Perkins, op. cit., p. 172. 27 See, for example: V. Fumagalli, Città e campagna nell’Italia medievale. Il Centro-Nord. Secoli VI–XIII, Bologna, 1979; B. Andreolli; P. Galetti; T. Lazzari; M. Montanari (ed.), Il Medioevo di Vito Fumagalli, Spoleto, 2010; M. Montanari, Campagne medievali, Turin, 1984.

the p r i m i t i vi s m o f t h e e ar ly m ed i e val pe asant i n i taly ?

Third, to acknowledge this primitive rural society begs the question of how it was persuaded to change, and how did it become the seigneurial world now well documented as evolving through the tenth and eleventh centuries. Let me amplify briefly each point: (i) The Shock of Decline?

The extraordinarily full archaeological evidence now shows the collapse of the Roman countryside from the later sixth to the earlier seventh centuries is now indisputable.28 There is no need to describe the evidence in terms of pessimism or optimism:29 the evidence speaks for itself even when archaeologists attempt to stretch the date ranges of their ceramics and thus speciously extend their dated levels into the eighth century. Collapse occurred. Household production of artisanal goods that had effectively replaced factory production in the fifth or early sixth centuries largely disappeared. Of course, this does not mean that rural society disappeared but it came close to it. The natural question is to ask why this collapse happened. The scientific evidence for the Justinianic pandemic30 and significant climate change31 is becoming increasingly compelling. The combination of catastrophes coupled with social changes hindered (if not prevented) the capacity for peasants to store foodstuffs, making peasant life much more challenging than we have previously appreciated. Along with social and economic factors — long since known — it is no longer difficult to model both demographic collapse on a huge scale and also the concomitant conditions which led to the end of market systems in much of Italy, and from this to the primitive rural circumstances evident in our excavations. But why are archaeologists so averse to collapse? Why, with the evidence at hand, is the village of the historians, a textual version that invariably emphasizes continuities, more attractive to archaeologists? What does this tell us about the first generation of Medieval archaeologists and their own worldviews?32 It would be trite to ascribe this to nationalism or localism or even a resistance to any alternatives to progress. More commonly historians of knowledge ascribe such disparities to the narrowing of a discipline as a result of the academic fragmentation of our research so that, in this case, archaeologists focus on their own region and only their time-period and have little understanding of the accelerated methodologies used in the affiliated

28 E. Cirelli, F. Diosono and H. Patterson, (eds), Le forme della crisi. Produzioni ceramiche e commercio nell’Italia centrale tra Romani e Longobardi (III–VIII d.c.), Bologna, 2015. 29 Ward-Perkins, op. cit.. 30 L. K. Little (ed.), Plague and the End of Antiquity. The pandemic of 541–740, Cambridge, 2008. 31 M. McCormick et al., Climate change during and after the Roman Empire: reconstructing the past from scientific and historical evidence, Journal of Interdisciplinary History 43, 2012, pp. 169–220: U. Büntgen et al., ‘Cooling and societal change during the Late Antique Little Ice Age from 536 to around 660 ce’, Nature Geoscience 9, 2016, DOI: 10.1038/NGEO2652. 32 R. Hodges, ‘Medieval archaeology and civic society: celebrating 40 years of Archeologia Medievale ’, in S. Gelichi (ed.), Quarant’Anni di Archeologia Medievale in Italia, Florence, 2014, pp. 177–83.

17 1

172

r icha r d ho d g e s

sciences that they themselves deploy.33 Whatever the preferred explanation, the question must be asked as we attempt to explain why the obvious archaeological evidence has been ignored. (ii) Changing Social Engagement

Urban based society — ‘cityness’ as Chris Wickham called it34 — still existed in discrete polyfocal elite units with minimal material resources.35 Thirty years after Wickham argued in a celebrated conference in 1987 that the material remains did not necessarily derive from economic and demographic decline, because the peasantry kept more surplus while the aristocracy spent on perishables rather than bricks and mortar,36 the archaeology of the eighth century, especially in erstwhile urban contexts, possesses far great clarity. Look at Rome in the early eighth-century with its primitive nummi, and limited imported amphorae from Sicily as well as limited range of ceramic wares. Moneyers might have been virtually extinct but there were still painters trained in the Byzantine traditions to decorate old buildings like S. Maria Antiqua to high standards. These small spotlights on the age tend to distract us — as do the texts — from an urban primitivism not least because towns figure prominently in the texts and of course they frame our contemporary view of the world.37 Rome is not unique. We possess similar pinpricks of elite consumption rhetoric — for example, at Cividale, at the Po fluvial port of Comacchio, at the Tempietto di Clitunno and at the duke’s palace in Salerno. Yet by contemporary Anglo-Saxon or Merovingian (let alone Danish) standards this urban consumption was miniscule, and the urban capacity to produce or distribute was highly limited in a way that society in North-West Europe at this time was not.38 Meanwhile, as the expanding palynological record in Italy shows much of the peninsula had become a green ocean as in other parts of Latin Christendom.39 Was this a golden age for peasants — as Chris Wickham and I once believed — in our Galbrethian desire for reifying the peasant liberated from social (especially tax) constraints? It is not at all clear as yet. We do not yet have the quantitative data in Italy as archaeologists to answer this question but with refined scientific tools a new

33 P. Burke, A Social History of Knowledge, Oxford, 2000; J. Moreland, Archaeology, theory and the Middle Ages, London, 2010, pp. 276–77. 34 Wickham, Framing the Middle Ages, p.595. 35 R. Hodges, ‘The idea of the polyfocal ‘town’? Archaeology and the origins of medieval urbanism in Italy’, in S. Gelichi and R. Hodges, (eds), New Directions in European Medieval Archaeology. Essays for Riccardo Francovich, Turnhout, 2015, pp. 267–84. 36 See C. Wickham, ‘Italy and the early Middle Ages’, in K. Randsborg (ed.) The Birth of Europe. Archaeology and Social Development in the First Millennium ad, Rome, 1988, pp. 140–51 at p. 148: ‘The overall level and sophistication of medieval culture declined because the peasantry was better off. I expect they just had to eat more.’; see also Ward-Perkins, op. cit., p. 172. 37 Hodges, Polyfocal, op. cit. 38 C. Loveluck, Northwest Europe in the Early Middle Ages, c. ad 600–1150. A comparative history, Cambridge, 2013. 39 J-P. Devroey, Economie rurale et société dans l’Europe franque (vie-ixe siècles), Paris, 2003, pp. 28–32.

the p r i m i t i vi s m o f t h e e ar ly m ed i e val pe asant i n i taly ?

generation should be able to resolve the quality and standards of peasant life. To imagine peasants were ’better off ’ than in Roman or High Medieval times, though, is surely far-fetched. (iii) Commodification of the Countryside

Plainly we know much more about the archaeology of the later ninth and tenth centuries40 as diverse non-perishable commodities — though not coins — reached the hands of peasants living in primitive conditions. This resembles the so-called Secondary Products Revolution, known from prehistory. Andrew Sherratt’s model of a secondary products revolution involved a widespread and broadly contemporaneous set of innovations in Old World farming. The use of domestic animals for primary carcass products was broadened from the fourth-third millennia bc to include exploitation for renewable ‘secondary’ products such as milk, wool, traction (the use of animals to drag ploughs in agriculture), riding and pack transport.41 Added to this, we can readily see the transformation of the countryside in the subsequent generations as new crops (and presumably convertible agriculture — rotation) were introduced, clearly for consumption by peasants as well as elites. Coinage may still have been absent,42 and coin-dated sites are still precious and few, but there is no mistaking the dramatic changes to rural society and its economic relations that had taken place in Italy by ad 1000: new house forms, storage facilities, new crops, a variety of ceramic forms, metal tools, and almost certainly a wide range of perishable commodities that archaeologists do not find. Simultaneously the texts multiply in number describing all manner of landed properties. Incastellamento in the light of the later seventh- to ninth century primitivism begs a simple question of us as archaeologists? This was well put by Frans Theuws in a similar transformation in the Low Countries: ‘the aristocrat’s and the dwellers’ landscapes came to overlap (in the ninth century) more than they did in the seventh and early eighth century. Devroey expects that a ‘ruralization’ of the elite took place in this period. I would say that at the same time, an elitization of the rural landscape and world took place. … However strange this may sound, the local dwellers may have (happily?) cooperated in creating (new landscapes) because of perceived opportunities …in relation to their own farmsteads. [T]he transformation of the rural world in Carolingian times may not have been a one-way, top-down elite operation’.43 This observation, of course, challenges us to interpret the texts with care because the underlying presumption is

40 See A. Molinari (ed.) ‘Mondi rurali d’Italia: insediamenti, struttura sociale, economia, secoli X–XIII’, Archeologia Medievale XXXVII, 2010, pp. 11–281. 41 Sherratt, op. cit. 42 A. Rovelli, ‘Coins and trade in early medieval Italy’, Early Medieval Europe 17, 2009, pp. 45–76. 43 F. Theuws, ‘Settlement excavations and the process of ‘manorialization’ in northern Austrasia’’, in S. Gasparri (ed.) 774: Ipotesi su una Transizione, Turnhout, 2008, pp. 199–220 at p. 220.

17 3

1 74

r icha r d ho d g e s

that rural society was persuaded (to use Max Weber’s thesis in other contexts44) to change their working and living practices by commodification — materialism — that is difficult to detect in the written accounts of these times. This would suggest not so much economic trickle-down in ‘Reaganomic terms’ as a massive social upheaval involving a revolution in rural behavior from culinary practices to a new ethos also encompassing working and reproduction. *** Forty years ago when Chris Wickham and I began the excavations at Santa Maria in Civita we could not have imagined this nuanced anthropological perspective of a major historical episode in Mediterranean history. It is a tribute to Italian Medieval archaeology that the clash between views of village evolution through the material and textual sources is provoking major new questions that with contemporary scientific analyses will lead to interpretations that assist us all to ask why written accounts took the form they did. In framing the Middle Ages what more can we ask of our disciplines than to be certain that paradigms are questioned, altered and evolved. In working with archaeologists in this adventure, Chris Wickham’s role in unearthing the material primitivism of Italy’s early Medieval peasant, whether this appeals to him or not, is now a part of history.

44 M. Weber, Economy and Society. An outline of Interpretive Sociology, Berkeley, CA, 1978, 212–16 and C. Wickham, ‘Systactic structures: social theory for historians’, Past and Present 152, 1991, pp. 188–203 at p. 191.

3

‘Small World’ and Social Inequalities

Jean-Pierre Devroey & Nicolas Schroeder

Land, Oxen, and Brooches Local Societies, Inequality, and Large Estates in the Early Medieval Ardennes (c. 850–c. 900)

Back then, my family was not poor. In the stable, there were a horse, a cow, sheep, and pigs. Years passed, we grew up. Soon, my brother and my sister could help with the work on the farm. This material comfort was not long lasting, though: poverty slowly installed itself in the household. We were struck by all sorts of misfortunes and setbacks. The time of receptions and feasts was over.1

Ivan Iakovlevitch Stoliaroff grew up at the end of the nineteenth century in Karatchoune, a small rural town in the province of Voronej, Russia. In his early years, his parents decided to supplement their incomes — hitherto based on mixed farming and pottery — by selling various items such as salt, tar, petrol, oil, salted fish, and biscuits. This commercial enterprise did not turn out well: too many customers asked for credit that they were incapable of repaying; jealous neighbours started gossiping against the family; local officials turned up, trading their Camorra-style support in exchange of small ‘presents’ such as tea, eggs, and vodka. The household struggled to get through the great famine of 1891–2: they had to kill their horse and cow, being left without a draft animal and milk. Ivan’s brother and sister left the house, shifting the economic balance of the household: there were fewer stomachs to fill, but also fewer arms helping on the farm. Moreover, these marriages came with expensive feasts, presents to various members of the families-in-law, and a dowry. As Ivan’s parents could not afford to pay the land tax anymore, local officials who had not been adequately bribed decided to seize the household’s arable land and lease it out to another farmer. The family attempted to oppose this decision and talked to the leaseholder. He was moved by their situation. He agreed to cede them the harvest he had sown on their fields in exchange for a relatively small amount of money. Yet, Ivan’s mother eventually convinced her husband that they should not pay anything, 1 I. Stoliaroff, Un village russe, V. Stoliaroff (trad.), Paris, 1992 (Terre humaine), pp. 59–60. Jean-Pierre Devroey  •  Académie royale de Belgique, Université libre de Bruxelles (ULB) Nicolas Schroeder  •  Université libre de Bruxelles (ULB) Social Inequality in Early Medieval Europe: Local Societies and Beyond, ed. by Juan Antonio Quirós Castillo, Turnhout : Brepols, 2020 (HAMA 39), pp. 177–202 © FHG10.1484/M.HAMA-EB.5.118451

1 78

j e a n -p ie r r e d e vro e y & n i co l as s c h ro e de r

arguing that the trouble they had endured deserved financial compensation. The case was taken to court and Ivan’s parents were now deprived of some of their movable goods. Only the family’s samovar and their last pig, which had both been carefully hidden, could be saved from the local authorities. Ivan’s father borrowed money from the deacon of a neighbouring village at quite a bad rate. This allowed them to get rid of their creditors and to temporarily stabilise their financial situation — at a much lower level than before, though: they now were poor.2 Several points deserve attention in this tragic story of the economic and social decline of a household. Firstly, the logic which underpins Stoliaroff ’s narrative of his parents’ misfortune is that in nineteenth-century Karatchoune, families could be richer or poorer; these differences were expressed through various material or ritual markers such as the amount of livestock or the celebration of feasts. Secondly, households could fail or succeed: starting with a well-equipped farmstead and a small capital earned during the father’s military service, the household was not the richest of Karatchoune from the outset, but it was not the poorest either. It could have succeeded in the long run. An important element which structurally constrained the household was its changing composition: the departure of Stoliaroff ’s siblings accentuated economic difficulties. Beyond this shift, which is part of the normal ‘life-cycle’ of peasant households,3 the downward trajectory of the family was the result of a series of misfortunes, choices that only revealed themselves as bad in retrospect, and pressures from authorities and neighbours. Indeed, the decline of the household was accompanied throughout by members of the local society: each episode of this dramatic scenario involves relatives, neighbours, local officials, or representatives of the Church, who offered help or contributed to worsen the family’s situation. More often than not, these interactions are ambiguous. At some point, the family ought to stop granting credit to customers, but this would surely have increased the general resentment against the family. One might argue that the farmer who took the household’s land in lease ultimately acted with solidarity and was badly rewarded for it. Yet, in the eyes of Stoliaroff ’s mother, he should never have agreed to the lease in the first place, as he knew very well whose land he was taking over and whose misery he was exploiting. The loan obtained by Stoliaroff ’s father helped the family at a particularly difficult moment, but the deacon obviously also acted out of economic interest. Solidarity was not non-existent in this society, but it was often forced by social pressure, mitigated by self-interest and selfishness. This leads us to a final observation: in this narrative, poverty emerges as a result of both structural conditions (the general architecture of household economies and social relations in the village) and contingencies. With its richness in detail and great temporal resolution, Stoliaroff ’s account is equally precious and dangerous as a guide for the historian of rural societies interested in inequality and economic stratification. It does indeed offer valuable information about the sociology of rural societies, that can be abstracted and theorized in order to articulate the scarce information provided, for example, by early medieval written



2 This paragraph synthesises Stoliaroff, Un village russe (cited n. 1), pp. 43–70. 3 This concept is discussed further below, pp. 190-192.

l and, oxe n, and b ro o che s

evidence. The dangers and limitations of such an approach are obvious: sociological abstraction can hide differences and peculiarities behind a fog-screen of theory and generalizations; the animation of scanty evidence through imaginative reconstruction quickly takes the historian into the realm of over-interpretation. Keeping these points of caution in mind, Stoliaroff ’s narrative nevertheless has some interest for historians who focus on social inequality and dynamics in early medieval peasant societies. The subject is not a new one: it has been approached directly or indirectly in various empirical studies and synthetic discussions.4 For historians working with written evidence, the difficulties of the task are numerous. Above all, evidence is scarce and the available data only sheds light on particular aspects of wealth and its distribution in local societies. The nature of the evidence heavily orients what can be known about these questions. Narrative evidence, especially hagiographic texts, can provide lively anecdotes and reveal some of the social dynamics within local societies.5 Yet, because of their symbolic and moral purpose as well as their literary logic, they are hard to interpret in the perspective of a rural sociology. Charters are certainly the best type of documents available to build up ‘a rural sociology of people dealing with other people’.6 In series, charters help to shed light on the trajectories of particular individuals and families.7 This material allows us to gain insight into the making and unmaking of fortunes, alliances, and clienteles; it provides a dynamic approach to particular individuals and groups within local societies. However, the picture is often fragmentary as lower status individuals, who do not engage in land transactions, are only documented indirectly, if at all. At first sight, estate records might appear to give a more comprehensive picture of local societies. However, this is not necessarily the case. Firstly, it is never certain that they record the entire population of a settlement.8 Secondly, they often only provide a decontextualized snapshot; whether indices of inequality such as differences in access to land reveal structural conditions or temporary





4 A short selection of references: R. Hilton, ‘Reasons for inequality among medieval peasants’, in The Journal of Peasant Studies, 5/3, 1978, pp. 271–84; L. Kuchenbuch, Bäuerliche Gesellschaft und Klosterherrschaft im 9. Jahrhundert. Studien zur Sozialstruktur der familia der Abtei Prüm, Wiesbaden, 1978, pp. 383–98; W. Davies, Small Worlds. The village community in Early medieval Brittany, London, 1988, pp. 86–104; Ch. Wickham, The Mountains and the City. The Tuscan Apennines in the early Middle Ages, Clarendon Press, Oxford, 1988, pp. 40–67; Id., Framing the Early Middle Ages. Europe and the Mediterranean, 400–800, Oxford, 2005, pp. 383–588; Id., ‘Rural Society in Carolingian Europe’, R. McKitterick, (ed.), The New Cambridge Medieval History, vol. 2 (c. 700–c. 900), Cambridge, 2008, pp. 510–37; P. Bonnassie, La Catalogne au tournant de l’an Mil. Croissance et mutations d’une société, Paris, 1990, pp. 97–154; J.-P. Devroey, Économie rurale et société dans l’Europe franque (vie–ixe siècles). 1. Fondements matériels, échange et lien social, Paris, 2003, pp. 297–309; Th. Kohl, Lokale Gesellschaften. Formen der Gemeinschaft in Bayern vom 8. bis zum 10. Jahrhundert, Ostfildern, 2010, pp. 122–56. 5 See, for example, Ch. West, ‘Visions in a ninth-century village: an early medieval microhistory’, History Workshop Journal, 81/1, 2016, pp. 1–16. 6 Wickham, ‘Rural Society’ (cited n. 4), p. 511. 7 See, for example, Wickham, The Mountains and the City (cited n. 4), pp. 40–67 and L. Feller, A. Gramain, F. Weber, La fortune de Karol. Marché de la terre et liens personnels dans les Abruzzes au haut Moyen Âge, Rome, 2005, pp. 93–129. 8 See below, p. 185.

17 9

1 80

j e a n -p ie r r e d e vro e y & n i co l as s c h ro e de r

situations is not always clear.9 Moreover, even the most comprehensive descriptions leave out various elements that are crucial to a study of social inequality.10 Finally, the systematic sense that estate records might convey at first sight is counterbalanced by their allusive formulae and, on occasion, cryptic vocabulary. In the face of these difficulties, some scholars have attempted to combine a close reading of the available evidence with a comparative and sociological perspective.11 From this point of view, the anthropology and sociology of peasant societies provide frameworks to interpret the available evidence (but also, sometimes, the absence of evidence12) and to compare social formations. This paper is a small contribution to this wider project. We will discuss the problem of social inequality and economic stratification in the estate of Villance, which is located in the Ardennes (fig. 1), mainly using an estate record, but also a few charters and hagiographic evidence. The focus of this case-study is admittedly narrow. One of the points of this approach is precisely to suggest that this is the appropriate scale at which to tackle the problem of social inequality. This small-scale discussion also allows us to flag many difficulties in the interpretation of evidence, which tend to get lost in more general and synthetic perspectives. This is part of our objectives, as historians who contribute to a book focussed on the archaeology of social inequality. We have tried to point out the potential of written evidence, as well as acknowledge the difficulties and limitations of its interpretation. In our experience, knowing about our colleagues’ methodological strengths and weaknesses helps historians and archaeologists to combine the different skills of the hedgehog and the fox.13 Chris Wickham certainly is one of those who understand best what our knowledge of early medieval local societies can gain from this alliance.

The Ardennes and the Estate of Villance in the Ninth Century The Ardennes are a low mountain range, which stretches across contemporary southern Belgium, Northern France, and Luxembourg (fig 1). The region consists

9 This point is discussed further below, pp. 190-192. 10 Kuchenbuch, Bäuerliche Gesellschaft (cited n. 4), pp. 249–60; also see below, p. 201. 11 On the importance and limitations of comparison, see Ch. Wickham, ‘Problems of comparing rural societies in early medieval Western Europe’, Transactions of the Royal Historical Society, 2, 1992, pp. 221–46. The input of sociology and anthropology is discussed by Kuchenbuch, Bäuerliche Gesellschaft (cited n. 4), p. 10; Wickham, Framing the Early Middle Ages (cited n. 4), pp. 535–50 and J.-P. Devroey, Puissants et misérables. Système social et monde paysan dans l’Europe des Francs (vie-ixe siècles), Brussels, 2006 (Mémoires de la Classe des Lettres de l’Académie royale de Belgique. Collection in-8°, 3e série, xl), pp. 9–14. 12 See Wickham, Framing the Early Middle Ages (cited n. 4), pp. 535–50. 13 We borrow this metaphor from Stephen Jay Gould, who suggests that productive interdisciplinarity relies upon the ability to combine the attitude of the hedgehog (‘a clear goal pursued relentlessly and without compromise’) and the fox (‘the flexibility of a wide range of clever and distinct strategies’). St J. Gould, The hedgehog, the fox, and the magister’s pox. Mending the gap between science and the humanities, Cambridge, Mass./London, 2011, p. 262.

l and, oxe n, and b ro o che s

Fig. 1. The Ardennes, Villance, and the localities mentioned in the text.

of a massif that rises steeply from the surrounding lowlands in the north and more progressively in the south. It is bounded by the river Meuse in the west and the Eifel massif in the east. The region combines a central plateau, with relatively flat surfaces and rolling hills, several strongly dissected river systems, and high plateaux covered with moorland. Nowadays, the region is dominated by woodland and pasture. In pre-industrial times, small areas of arable land and extensive wasteland were common features in the landscape. With its numerous fiscal estates, royal monasteries and forests, the proximity of Aachen, and its situation between the Meuse and Mosel valleys, the Ardennes were one of the Carolingian ‘heartlands’.14 The region had a high political, symbolic, and even economic value for them, despite its geographical marginality. Traditionally, this situation has been explained as follows: because of its repulsive geography, the 14 See F. Rousseau, ‘Les Carolingiens et l’Ardenne’, in Bulletin de la Classe des Lettres et des Sciences Morales et Politiques de l’Académie royale de Belgique. Ser. 5, 48, 1962, pp. 187–221; E. Ewig, ‘Les Ardennes au haut Moyen Âge’, in Anciens pays et assemblées d’États, 38, 1963, pp. 1–38; H. MüllerKehlen, Die Ardennen im Frühmittelalter. Untersuchungen zum Königsgut in einem karolingischen Kernland, Vandenhoeck and Ruprecht, Göttingen, 1973 (Veröffentlichungen des Max-Planck-Instituts für Geschichte, 38); R. Hennebicque, ‘Espaces sauvages et chasses royales dans le Nord de la Francie, viième-ixème siècles’, in Revue du Nord, 62/244, 1980 [= Actes des congrès de la Société des historiens médiévistes de l’enseignement supérieur public, 10e congrès, Lille, 1979. Le paysage rural: réalités et représentations], pp. 35–57; Ch. Wickham, ‘European forests in the Early Middle Ages: landscape and land clearance’, in L’ambiente vegetale nell’alto medioevo, Spoleto, 1990 (Settimane di studio del Centro italiano di studi sull’alto medioevo, 37), pp. 479–548.

181

1 82

j e a n -p ie r r e d e vro e y & n i co l as s c h ro e de r

massif was almost entirely abandoned during the crises of the late Roman period. Following Roman law, this ‘abandoned’ land became ‘public’ and therefore royal property. The Carolingian kings could then organize the colonization of the region through the foundation of royal monasteries and estates. In recent years, however, this model has come under criticism: it has been observed that on the central plateau of the Ardennes, landowners of varying status were established within and alongside royal and monastic estates, forming a much more complex society than accepted hitherto.15 Using charters which reveal disputes about the use of woodland in the late eighth and ninth centuries, it has been argued that the monasteries and royal estates of the Ardennes were not (only) created through woodland clearances, but (also) through the incorporation of already established farmsteads within increasingly extractive estate structures.16 As we will see, this re-evaluation of the region’s history could have some significance for the analysis of social inequality in the estate of Villance. The village of Villance is located in a landscape of gently rolling hills on the north-west of the central plateau of the Ardennes (fig. 1). It is mentioned for the first time in a diploma issued by Louis the Pious on 26 June 839.17 This document shows that sometime before 833, the royal villa of Villance was granted out to a dignitary at the court of Louis the Pious named Richard.18 He betrayed the king, who confiscated Villance in 834/5. In 839, Louis had pardoned Richard and the estate was given to him again with several unfree individuals (mancipia and servi). At that time, a church existed in the villa, as well as mills. Interestingly, some of the dependants of the villa were not living in it, but in other locations.19 Richard died shortly after the restitution of the villa. Following his will and with the consent of Lothar I, his brothers donated Villance to the monastery of Prüm on the twelfth of November 842.20 According to a notice transcribed in Prüm’s tenth century cartulary, the transfer of property took

15 J.-P. Devroey and N. Schroeder, ‘Beyond Royal Estates and Monasteries: landownership in the early medieval Ardennes’, in Early Medieval Europe, 20, 2012, pp. 39–69 and N. Schroeder, ‘L’Ardenne: appropriation, exploitation et paysages du haut Moyen Âge à 1300’, in M. Pauly and H. Pettiau, (eds), La forêt en Lotharingie médiévale. Actes des 18es Journées Lotharingiennes, 30–31 octobre 2014. Université du Luxembourg, Luxembourg, 2016 (Publications de la section historique de l’Institut Grand-Ducal, 127/ Publications du CLUDEM, 43), pp. 163–92. 16 Schroeder, ‘L’Ardenne’ (cited n. 15), pp. 168–79. 17 MGH DD LdF 401 (26 June 839). 18 About Richard, see P. Depreux, Prosopographie de l’entourage de Louis le Pieux (781–840), Sigmaringen, 1997, n. 45, pp. 131–32; about the episode: Kuchenbuch, Bäuerliche Gesellschaft (cited n. 4), p. 47; E. Renard, ‘La gestion des domaines d’abbaye aux viiie-xe siècles. Notions de base et conseils pour une meilleure compréhension des sources écrites’, De la Meuse à l’Ardenne, 29, 1999, pp. 117–18; J.-P. Devroey, ‘La hiérarchisation des pôles habités et l’espace rural. Autour des possessions de l’abbaye de Prüm (893) en Ardenne belge’, M. Gaillard, M. Margue, A. Dierkens, H. Pettiau (eds), De la mer du Nord à la Méditerranée. Francia Media. Une région au cœur de l’Europe (c. 840–c. 1050). Actes du colloque international (Metz, Luxembourg, Trèves, 8–11 février 2006), Luxembourg, 2011 (Publications du Centre luxembourgeois de documentation et d’études médiévales, 25), p. 181, n. 22. 19 Mancipiis desuper conmanentibus vel ad eam in quocumque loco iure aspicientibus vel pertinentibus. MGH DD LdF 401. 20 MGH DD Lo I 68 (12 November 842).

l and, oxe n, and b ro o che s

place in Villance, where Folrad, one of Lothar’s vassals, three envoys of Prüm’s abbot Markward, the monastery’s advocate Teotfred, and several other people gathered.21 A serf named Germinanus participated in the transaction: Folrad gave him to the monastery’s envoys as a representative of the entire group of dependents.22 In 893, the monks produced an inventory of their properties: several teams toured their numerous estates and collected information about them, which was then compiled into a polyptych.23 A thirteenth-century annotated copy of this document is preserved.24 The chapter devoted to Prüm’s properties and rights in the estate of Villance has attracted much attention and the literature that discusses the text is abundant.25 Yet, with exception of Georges Despy and Ludolf Kuchenbuch, not much has been written about the tenants of the estate: most commentators have focussed on estate structures and management, economic organization, and how these elements relate to demography and settlement history.26 Of course, by nature, the polyptych is not a description of local societies and there are important limitations to what can be said using this evidence. However, its analysis allows us to make a few points about the stratification of local societies in the early medieval Ardennes.

Prüm’s Estate in Villance: Infrastructures and Individuals The chapter about the estate of Villance provides a short inventory of the seigniorial infrastructures (two mills and three breweries — there might have been further buildings, but they are not described) and demesne land (a scatter of seven blocks of land of varying size, a hay meadow, and a forest).27 The document then mentions the

21 Urkunden- und Quellenbuch zur Geschichte der altluxemburgishen Territorien bis zur Burgundischen Zeit, C. Wampach (ed.), 1 (bis zum Friedensvertrag von Dinant 1199), Luxembourg, 1935, n. 83, pp. 77–78. 22 De omnibus vero mancipiis per servum unum nomine Germinanum similer eosdem missos revestivit. Urkunden- und Quellenbuch (cited n. 21), n. 83, p. 78. 23 Das Prümer Urbar, I. Schwab (ed.), Düsseldorf, 1983 (Rheinische Urbare, 5). 24 About this copy and the copyist’s commentaries, see L. Kuchenbuch, ‘Die Achtung vor dem alten Buch und die Furcht vor dem neuen. Cesarius von Milendonk erstellt 1222 eine Abschrift des Prümer Urbars von 893’, Historische Anthropologie, 3, 1995, pp. 175–202 (with reference to earlier litterature). 25 Ch.-E. Perrin, Recherches sur la seigneurie rurale en Lorraine d’après les plus anciens censiers (ixe-xiie siècles), Paris, 1935, pp. 640–42; G. Despy, ‘Villes et campagnes aux ixe et xe siècles: l’exemple du pays mosan’, Revue du Nord, 50, 1968, pp. 154–62; G. Duby, L’économie rurale et la vie des campagnes dans l’Occident médiéval, vol. 1, Paris, 1961, pp. 67 and 116–17; Ch. Dupont, ‘Un moment d’histoire domaniale: Villance au ixe siècle’, Revue du Nord, 58, 1976, p. 151; J.-P. Devroey, ‘Mansi absi: Indices de crise ou de croissance de l’économie du haut Moyen Âge?’, Le Moyen Âge, 82/3–4, 1976, pp. 428–35; Renard, ‘La gestion des domaines d’abbaye’ (cited n. 18); Devroey, ‘La hiérarchisation des pôles habités’ (cited n. 18). 26 Despy, ‘Villes et campagnes’ (cited n. 25), pp. 154–62 has formulated various observations about the economy of tenants (see below, p. 193); Kuchenbuch, Bäuerliche Gesellschaft (cited n. 4) discusses the organization and stratification of local societies gathering information from the entire polyptych. Several points are made about Villance in particular, but the estate is not discussed systematically. 27 An analysis of the topographic organization of these elements is given by Devroey, ‘La hiérarchisation des pôles habités’ (cited n. 18), pp. 181–88.

183

1 84

j e a n -p ie r r e d e vro e y & n i co l as s c h ro e de r

Fig. 2. Prüm’s estate of Villance, c. 893.

existence of forty-seven ‘free’ holdings (mansa ingenuales). This part of the description is structured by localities, as the tenures are scattered across nine neighbouring places (fig. 2).28 Sixteen holdings are referenced in the villa Libin,29 two in Oussi (Libin),30 five in À Frofaÿ (Anloy),31 eight in Anloy,32 seven in Lesse (Redu),33 thirteen in Transinne,34

28 Discussed by Devroey, ‘La hiérarchisation des pôles habités’ (cited n. 18), pp. 181–88. 29 Nine full mansi, three half mansi, three quarter mansi, and one homestead (sacium). 30 Two half mansi. 31 Two full mansi, one half mansus, and two quarter mansi. 32 Six mansi, a quarter mansus, and half a quarter mansus. 33 A mansus and a quarter, four full mansi, and two quarter mansi. 34 Eight full mansi, two half mansi, and three quarter mansi.

l and, oxe n, and b ro o che s

Gender

Personal dependence

Territorial lordship

Tenancy

M M F M F

member of Prüm’s familia member of Prüm’s familia member of Prüm’s familia not a member of the familia not a member of the familia

in the potestas outside the potestas / in the potestas in the potestas

no holding no holding no holding / /

Fig. 3. Criteria which determine the personal tribute paid by individuals in Prüm’s estate of Villance (c. 900?).

three in Derrière Mussy (Transinne),35 three in La Pierre afin (Villance),36 and seven in Hoÿmont (Villance).37 For each of these holdings, the names of the tenants are given, as well as the rent and services that the monastery expects to receive from them. These duties are presented as a standardized set of obligations, which varies depending on the number of co-tenants on a holding. The next section of the estate-description enumerates seven unfree holdings (mansi servili), without locating them precisely (it is likely that they were in Villance itself); they consist of six mansi and one quarter mansus. At the end of this section, a holding of two jornales is mentioned. The three last sections of the chapter might be additions made to the polyptych in the early tenth century.38 Under the title ‘About the fiefs of the ministeriales’, three holdings are enumerated. The first one is described as a beneficium. It consists of two mansi and eight jornales of land held by Amulbertus. The second one is a mansus and a half held by Seihardus. The last one is three full mansi and a chapel held by the priest Hartbertus. The next section describes the rent men and women belonging to the estate have to pay to the monastery, following four criteria: gender, personal dependence, residence within Prüm’s territorial lordship (potestas), and tenancy (fig. 3).39 The last part of the chapter is a sum, which recapitulates the properties and incomes of the abbey. What does this description reveal about local societies and inequalities in them? First of all, it is important to stress that the document implicitly reveals that it does not provide a full survey of land and population in the mentioned settlements: individuals live within Prüm’s territorial lordship (potestas) without being tenants.40 They can be personally dependent on the monastery (members of the familia), on another lord, or free. As they do not hold land from Prüm, they are not nominally recorded. These observations also imply that not all land in and around Villance was owned by the monastery: Villance is not a block-estate, but an ensemble of properties

35 36 37 38 39 40

One full mansus, half a mansus, and a quarter mansus. One full mansus and two half mansi. Three full mansi and four quarter mansi. Devroey, ‘La hiérarchisation des pôles habités’ (cited n. 18), p. 184. The notion of potestas is discussed by Kuchenbuch, Bäuerliche Gesellschaft (cited n. 4), pp. 267–68. Renard, ‘La gestion des domaines d’abbaye’ (cited n. 18), pp. 124–25; Devroey, ‘La hiérarchisation des pôles habités’ (cited n. 18), p. 187.

185

186

j e a n -p ie r r e d e vro e y & n i co l as s c h ro e de r

and rights scattered over several settlements (fig. 2).41 Some of Prüm’s tenants might own land in property or hold land from another landowner that is, of course, not recorded in the polyptych. The document allows us to distinguish three types of holdings: free holdings, unfree holdings, and benefices. The following discussion is structured along these categories.

Free-Unfree: A First Line of Division? The free-unfree divide relates to the legal status of individuals and holdings. We have already seen that a servus was given to Prüm in 842 to symbolize the transfer of the dependants (mancipia) of the estate. Moreover, the polyptych distinguishes between femine de familia (women who belong to the monastery) and ancille (female slaves).42 These elements suggest that individuals with various legal statuses were living on the estate.43 However, the chapter which describes Prüm’s Villance estate does not provide information about the status of individuals: it only distinguishes between free and unfree holdings.44 The relationship between the status of tenants and holdings is complex: the analysis of polyptychs that provide information about the legal status of holdings and individuals shows that they do not necessarily match (meaning a free individual can hold an unfree tenure and, although this situation is much less common, vice-versa).45 Applied to holdings, the free-unfree divide essentially structures differences in the size of tenures and in the nature and quantity of rent and labour services expected from tenants. On average, within the same estate, servile tenures tend to be smaller in size and associated with heavier labour services than free holdings.46 The description of Prüm’s Villance estate does not allow us to discuss the first aspect, as the surface of holdings is not recorded. There is more clarity about rent and labour services, however: tenants of free holdings have to perform a defined set of services throughout the year (15 nights twice a year, cultivation of 1/2 jornales of land, sowing of 4 jornales in autumn and spring, 2 [days (?) of] ploughing on request a year, cutting wood, transportation services in May and December); they can be asked to serve up to three days a week. Tenants of servile holdings have to ‘perform any service that is requested’.47 This is a first important factor of differentiation within the local society: in Prüm’s estate, seven tenant households out of one hundred and fifty-five were potentially

41 This point is discussed in the broader regional context in Devroey & Schroeder, ‘Beyond Royal Estates and Monasteries’ (cited n. 15), pp. 39–69. 42 Das Prümer Urbar (cited n. 23), pp. 207–08. 43 See the discussions in Kuchenbuch, Bäuerliche Gesellschaft (cited n. 4), pp. 360 and 380. 44 The distinction between femine de familia and ancille however suggests that, within the estate, the personal status of women mattered, probably because it determines the status of children. 45 J.-P. Devroey, Puissants et misérables (cited n. 11), p. 300; 529–30; A. Rio, Slavery After Rome, 500–1100, Oxford, 2017, pp. 186–99. 46 Devroey, Puissants et misérables (cited n. 11), pp. 529–30. 47 reliquum tempus facit omne servitium, quicquid ei iubetur. Das Prümer Urbar (cited n. 23), p. 207.

l and, oxe n, and b ro o che s

submitted to unlimited burdens. They had to serve at their lord’s pleasure, while tenants of free holdings were submitted to a defined set of duties. In all likelihood, the fact that the monastery captured this level of discrepancy in writing is indicative of a certain degree of socio-economic disadvantage for those who live on unfree tenures (even if they are legally free); once the clear cut categorization between free and unfree holdings had been imposed (and this might only have happened when the polyptych was written up), it certainly contributed to maintaining, or even reinforcing, this disadvantage.48 Beyond the material differences visible in rent and service, esteem and respectability were at play. The numerous disputes from the second half of the ninth and the tenth centuries during which individuals claimed that they had been unjustly categorized as unfree or requested to perform servile services show that legal status was much more than a label with no importance attached.49 Within local societies, being categorised as unfree or having to perform services associated with servile status (be it personal or based on tenancy) is not necessarily a factor of social depreciation: the existence of mixed (free-unfree) marriages on the estates of Saint-Germain-des-Prés, for example, suggests ‘a level of solidarity between the two status-groups’.50 Yet, in several disputes about legal status, claimants are declared unfree because their free neighbours testify against them, revealing that legal status, despite being an instrument of the lord, could also become a cornerstone around which the internal tensions of local societies crystallised and were played out.51

Free Holdings I. Mansus Taxonomy, Co-Tenancy, and Wealth The bulk of holdings in Prüm’s Villance estate (65) are categorized as free. The polyptych provides information about two factors which allow us to differentiate these holdings among themselves: their size and the number of tenants per holding (fig. 4).

48 We herewith follow the stimulating line of argument developed by Rio, Slavery After Rome (cited n. 45), p. 198. 49 Well-known cases from regions surrounding the Ardennes: Remiremont (825), Courtisols (847), Mitry (861). See respectively: J. Barbier, ‘Fisc et ban à Remiremont: le fisc à l’origine des ‘bans’ romarimontains?’, M. Parisse, J.-P. Rothiot, P. Heili (eds), Le Pays de Remiremont des origines à nos jours. Actes des journées d’études vosgiennes. Remiremont, 2, 3 et 4 novembre 2000, Remiremont, 2001, 9–19; J.-P. Devroey, ‘Libres et non-libres sur les terres de Saint-Remi de Reims: la notice judiciaire de Courtisols (13 mai 847) et le polyptyque d’Hincmar’, Journal des savants, 2006 (1), pp. 65–103 and J. Barbier, ‘‘The praetor does concern himself with trifles’: Hincmar, the polyptych of Saint-Remi of Rheims and the slaves of Courtisols’, R. Stone & Ch. West (eds), Hincmar of Rheims: life and work, Manchester, 2015, pp. 211–27; J. Nelson, ‘Dispute settlement in Carolingian West Francia’, W. Davies and P. J. Fouracre (eds), The settlement of disputes in early medieval Europe, Cambridge, 1986, pp. 51–52. Similar cases (involving individuals rather than groups) are still recorded in the first half of the tenth century. For example: G. Despy, ‘Serfs ou libres? Sur une notice judiciaire cambrésienne de 941’, Revue belge de philologie et d’histoire, 39/4, 1961, pp. 1127–43. 50 Wickham, Framing the Early Middle Ages (cited n. 4), p. 405. 51 Rio, Slavery After Rome (cited n. 45), pp. 196–99; Devroey, Puissants et misérables (cited n. 11), p. 302.

187

188

j e a n -p ie r r e d e vro e y & n i co l as s c h ro e de r

1/4